Вы находитесь на странице: 1из 100
Steady State Analysis of the Interconnection of Offshore Energy Parks Miguel Jorge da Rocha Barros
Steady State Analysis of the Interconnection of Offshore Energy Parks Miguel Jorge da Rocha Barros

Steady State Analysis of the Interconnection of Offshore Energy Parks

Miguel Jorge da Rocha Barros Marques

Dissertação para obtenção do Grau de Mestre em

Engenharia Electrotécnica e de Computadores

Júri

Presidente:

Prof. Doutor Paulo José da Costa Branco

Orientador:

Prof. Doutor Rui Manuel Gameiro de Castro

Co-orientador:

Profª. Doutora Maria Eduarda de Sampaio Pinto de Almeida Pedro

Vogal:

Profª. Doutora Sónia Maria Nunes dos Santos Paulo Ferreira Pinto

Setembro de 2010

Agradecimentos

Em primeiro lugar quero agradecer ao Professor Rui Castro pela confiança depositada em mim desde o princípio e pelo apoio prestado em todos os momentos da elaboração desta tese. O meu obrigado também à Professora Maria Eduarda Pedro, como co-orientadora, pela disponibilidade ao esclarecimento de dúvidas que surgiram. Uma palavra de agradecimento ainda ao Professor Ferreira de Jesus pela paciência e disponibilidade quase totais e que em muito contribuíram para a boa conclusão deste trabalho.

Uma palavra ainda de forte agradecimento aos meus colegas de curso – alguns dos quais se tornaram verdadeiros amigos – por toda a camaradagem e espírito de entreajuda, essenciais ao longo de todo o curso.

Uma palavra muito importante de gratidão à minha família. Aos meus pais, não só por terem viabilizado economicamente os meus estudos mas, sobretudo, pela força e carinho que me deram. Os seus percursos pessoais e académicos foram sempre um exemplo e uma fonte de inspiração ao longo do meu trajecto académico. À minha irmã, cujo apoio entusiasmado e companheirismo, ainda que por vezes à distância, foi também essencial.

Um agradecimento muito especial à minha namorada, por ter estado ao meu lado em todos os momentos, mesmo nos mais difíceis, e me ter sempre apoiado com ânimo, compreensão e amor, tornando definitivamente mais fácil o meu percurso.

i

Resumo

Entre as fontes renováveis de energia, a energia eólica tem ganho relevância, tornando-se uma alternativa viável às fontes convencionais. As elevadas velocidades do vento e as vastas áreas disponíveis no mar aumentaram recentemente o interesse global em parques eólicos offshore.

À medida que os parques eólicos offshore se tornam maiores e são instalados mais longe da

costa, a transmissão de energia para a rede em terra torna-se numa característica essencial. As opções disponíveis são HVAC, HVDC-LCC e HVDC-VSC.

O objectivo deste trabalho é comparar a performance em regime estacionário e transitório de

dois sistemas de transmissão alternativos: HVAC e HVDC-LCC. Alguns esquemas de compensação do excesso de energia reactiva para o HVAC são também avaliados, tal como as vantagens do STATCOM em ambas as tecnologias. O software PSS/E é usado para simulação de todos os casos.

Os resultados obtidos mostram algumas diferenças importantes entre as tecnologias. Em regime permanente, o comportamento do HVDC-LCC e do HVAC com compensação onshore e offshore têm algumas semelhanças, nomeadamente no factor de potência e nas perdas. Em regime transitório, a resposta da ligação com HVAC a um defeito na rede é muito diferente da resposta da ligação com HVDC-LCC. A ligação com HVAC permite o “fault ride through” do parque eólico offshore enquanto a ligação com HVDC-LCC é bloqueada, interrompendo a transmissão de energia durante o defeito.

Palavras-Chave: Parques Eólicos Offshore, HVAC, HVDC-LCC, STATCOM, PSS/E.

ii

Abstract

Among the renewable energy sources, wind energy has gained relevance, becoming a viable alternative to conventional sources. High wind speeds and wide available area at sea have recently increased the global interest on offshore wind farms.

As offshore wind farms become larger and are placed further from the shore, the power transmission to the onshore grid becomes a key feature. The available options are HVAC, HVDC-LCC and HVDC-VSC.

The objective of this work is to compare the performance in steady-state and the transient behaviour of two alternative transmission systems: HVAC and HVDC-LCC. Compensation schemes for excessive reactive power for HVAC are also evaluated, as well as the benefits of the STATCOM for both technologies. The PSS/E software is used for simulation of all cases.

The results obtained show some important differences for the technologies. In steady-state, the behaviour of HVDC-LCC and HVAC with compensation both onshore and offshore hold some similarities, namely in the power factor and the power losses. In transient regime, the response of the HVAC link to a fault in the grid is very different to the response of HVDC-LCC link. The HVAC link allows the “fault ride through” of the offshore wind farm while the HVDC link is blocked, interrupting the power transmission during the fault.

Keywords: Offshore Wind Farms, HVAC, HVDC-LCC, STATCOM, PSS/E.

iii

List of Figures

Figure 1 – EU power capacity mix: a) 2000 data; b) 2009 data

1

Figure 2 – Global Cumulative Installed Wind Capacity (1996-2009)

2

Figure 3 – Annual and cumulative installed capacity (in MW) of offshore wind power in Europe (1991 –

2009) [1]

3

Figure 4 – Europe wind resources over open sea (Offshore Wind Atlas) [6]

8

Figure 5 – Cumulative share of installed capacity by country (end 2009)

10

Figure 6 – Map of operational offshore wind farms in Europe (January 2010) (Adapted from [8])

11

Figure 7 – Offshore foundations and general characteristics [9]

12

Figure 8 – Offshore wind turbine size evolution

12

Figure 9 – General layout of an offshore wind farm

16

Figure 10 – Layout of an HVAC wind farm[16]

18

Figure 11 – Three-core XLPE submarine cable

19

Figure 12 – Equivalent π-model for a submarine

20

Figure 13 – Maximum transmission lengths with inductive shunt compensation in both ends [17]

21

Figure 14 – (a) Schematic Diagram of an SVC; (b) an SVC in Radsted, Denmark

22

Figure 15 – (a) Schematic Diagram of a STATCOM; (b) a STATCOM installation

23

Figure 16 – Loading of an open-end cable due to capacitive charging current for different schemes of

24

Figure 17 – Basic Configuration of a wind farm using an HVDC-LCC [7] Note: F=filter; HFF=High-

reactive power compensation [19]

frequency

26

Figure 18 – Available HVDC Cables: a) OF cable; b) MI cable; c) XLPE cable

28

Figure 19 – Mass Impregnated HVDC Cable [22]

28

Figure 20 – Six-pulse Converter Bridge

29

Figure 21 – Diagram of a 12-pulse converter [23]

31

Figure 22 – HVDC configurations: a) Monopole; b)

32

Figure 23 – Basic Configuration of a wind farm using an HVDC-VSC [11]

34

Figure 24 – Single line diagram of the test grid used (with HVAC transmission, compensations at both

ends)

38

Figure 25 – Layout of the wind turbines in the Lillgrund wind farm

40

Figure 26 – GE 1,5 MW voltage protection

43

Figure 27 – PSS/E model of the offshore wind

44

Figure 28 – HVAC transmission layout in

45

Figure 29 – HVDC-LCC transmission model in

47

Figure 30 – STATCOM device in PSS/E

49

Figure 31 – Legend of the values presented in the single-line diagrams

50

Figure 32 – Power Flow result for HVAC with offshore

52

Figure 33 – Power Flow result for HVAC with onshore

53

Figure 34 – Power Flow result for HVAC with offshore and onshore

54

iv

Figure 35 – Power Flow result for HVAC with STATCOM compensation

55

Figure 36 – Power Flow result for HVDC-LCC

58

Figure 37 – Power Flow result for HVDC+STATCOM

59

Figure 38 – Single line diagram of the grid used for the dynamic simulations. Note: Bus 3005, where

the fault occurs is marked in the orange

61

Figure 39 – Frequency variation of offshore buses for the Case 1

63

Figure 40 – Voltage variation of offshore buses for the Case 1

63

Figure 41 – Offshore wind turbine speed variation for the Case 1 fault

63

Figure 42 – Frequency variation of onshore buses for the Case 1

63

Figure 43 – Voltage variation of onshore buses for the Case 1

63

Figure 44 – Speed variation of machines on the onshore grid for the Case 1 fault

63

Figure 45 – Frequency variation of offshore buses for the Case 2

64

Figure 46 – Voltage variation of offshore buses for the Case 2

64

Figure 47 – Offshore wind turbine speed variation for the Case 2 fault

64

Figure 48 – Frequency variation of onshore buses for the Case 2

64

Figure 49 – Voltage variation of onshore buses for the Case 2

64

Figure 50 – Speed variation of machines on the onshore grid for the Case 2 fault

64

Figure 51– Frequency variation of offshore buses for the Case 3

65

Figure 52– Voltage variation of offshore buses for the Case 3

65

Figure 53– Offshore wind turbine speed variation for the Case 3 fault

65

Figure 54– Frequency variation of onshore buses for the Case 3

65

Figure 55– Voltage variation of onshore buses for the Case 3

65

Figure 56– Speed variation of machines on the onshore grid for the Case 3

65

Figure 57– Frequency variation of offshore buses for the Case 4

66

Figure 58– Voltage variation of offshore buses for the Case 4

66

Figure 59– Offshore wind turbine speed variation for the Case 4 fault

66

Figure 60– Frequency variation of onshore buses for the Case 4

66

Figure 61– Voltage variation of onshore buses for the Case 4

66

Figure 62– Speed variation of machines on the onshore grid for the Case 4

66

Figure 63 – Reactive Power injected/absorbed by the STATCOM on bus

66

Figure 64– Frequency variation of offshore buses with HVDC-LCC

69

Figure 65– Voltage variation of offshore buses with HVDC-LCC

69

Figure 66– Offshore wind turbine speed variation with HVDC-LCC

69

Figure 67– Frequency variation of onshore buses with HVDC-LCC

69

Figure 68– Voltage variation of onshore buses with HVDC-LCC

69

Figure 69– Speed variation of machines on the onshore grid with HVDC-LCC

69

Figure 70 – Active Power in the converters, for the HVDC-LCC

70

Figure 71 – Reactive Power in the converters, for the HVDC-LCC

70

Figure 72– Frequency variation of offshore buses with HVDC-LCC+STATCOM transmission

71

Figure 73– Voltage variation of offshore buses with HVDC-LCC+STATCOM

71

v

Figure 74– Offshore wind turbine speed variation with HVDC-LCC+STATCOM

71

Figure 75– Frequency variation of onshore buses with HVDC-LCC+STATCOM transmission

71

Figure 76– Voltage variation of onshore buses with HVDC-LCC+STATCOM

71

Figure 77– Speed variation of machines on the onshore grid with HVDC-LCC+STATCOM

 

71

Figure 78 –Active Power in the converters, for HVDC-LCC+STATCOM

72

Figure 79 – Reactive Power in the converters, for HVDC-LCC+STATCOM

72

Figure 80 – Reactive Power injected/absorbed by the STATCOM on bus

72

vi

List of Tables

Table 1: Colour correspondence for each voltage level in Figure 24

39

Table 2: Components of the GE 1,5 MW Wind Turbine Model in

42

Table 3: 150 kV cable

45

Table 4: Onshore and Offshore transformer data

46

Table 5: Equations for Rectifier and Inverter

47

Table 6: Reactive power compensation alternatives

51

Table 7: HVAC selected results for different compensation

56

Table 8: HVDC-LCC and HVDC-LCC+STATCOM selected

60

Table 9: HVAC Dynamic Cases

62

Table 10: Reactance and Susceptance of the lines

80

Table 11: 400 kV line calculated

80

Table 12: Two-Winding Transformers

81

Table 13: Load

81

Table 14: Shunt

81

Table 15: Wind Turbine aggregate Power Flow

82

Table 16: Unit transformer of one single wind turbine

82

Table 17: 33 kV Horns Rev cable parameters

83

Table 18: 33 kV AC Cable parameters used in the modelled wind

83

Table 19: DC Line Power Flow

83

Table 20: DC Link Converters Power Flow

84

Table 21: “CDC4T” model parameters for the DC Link

84

Table 22: STATCOM Power Flow parameters

85

Table 23: “CSTCNT” model parameters for the

86

vii

Abbreviations

 

A

Ampère

A

Rotor Swept Area

ABB

Swiss-Swedish High Tech Engineering Multinational

AC

Alternating Current

 

B

Susceptance

DC

Direct Current

EEA

European Environment Agency

EU

European Union

EWEA

European Wind Energy Association

 

F

Farad

FACTS

Flexible Alternating Current Transmission System

FP

Power Factor

GE

General Electric

GW

Gigawatt

 

H

Henry

HPFF

High Pressure Fluid Filled

HPGF

High Pressure Gas Filled

HVAC

High Voltage Alternating Current

HVDC

High Voltage Direct Current

Hz

Hertz

I

c

Charging current

I

d

Output DC current

IEEE

Institute of Electrical and Electronics Engineers

IGBT

Insulated Gate Bipolar Transistor

kA

Kiloampère

km

Kilometre

kV

Kilovolt

kW

Kilowatt

 

L

Inductance

LCC

Line Commutated Converters

LPFF

Low Pressure Fluid Filled

LPOF

Low Pressure Oil Filled

MI

Mass Impregnated

MVA

Mega Volt Ampère

Mvar

Mega Volt Ampère Reactive

MW

Megawatt

OF

Oil Filled

PCC

Point of Common Coupling

PSS/E

Power System Simulator for Engineering

p.u.

Per unit

 

Q

Reactive Power

R

Resistance

Rc

Commutating Resistance

viii

rpm

S b

SCFF

SCLF

SMES

STATCOM

SVC

TWh

V

V

b

V

LN

VSC

u

U d

X

XLPE

XT

Z

0

Z

b

α

β

γ

ρ

µ

ω

Revolutions per minute

Base Apparent Power

Self Contained Fluid Filled

Self Contained Liquid Filled

Superconducting Magnetic Energy Storage

Static Synchronous Compensator

Static VAr Compensator

Terawatt hours

Volt

Base Voltage

Line to Neutral rms Voltage

Voltage Source Converters

Wind Speed

DC output voltage

Reactance

Cross Linked Polyethylene

Commutating Reactance

Surface Roughness

Base Impedance

Firing Angle

Advance Angle

Extinction Angle

Air density

Delay Angle

Ohm

Angular Speed

Table of Contents

1. Introduction

1

1.1. Motivations

1

1.2. Thesis Objectives and Outlook

3

1.3. Thesis Structure

4

2. Wind Energy Offshore

6

2.1. Introduction

6

2.2. The Wind Resource

6

 

2.2.1.

Wind Offshore

7

2.3.

Offshore Wind Farms

9

2.3.1. Current Status

10

2.3.2. Technologies and Future Trends

11

2.4.

Grid Integration

12

2.4.1. Transmission Technologies

13

2.4.2. Grid Connection Requirements

14

3. Transmission Technologies for Offshore Wind Farms

16

3.1. Introduction

16

3.2. HVAC Transmission

17

3.2.1. General Aspects

17

3.2.2. Topology and Main Components

17

3.2.3. AC Submarine Cables

18

3.2.4. Reactive Power Compensation

21

 

3.2.4.1. FACTS Devices

22

3.2.4.2. Distribution of the Compensation

23

3.3.

HVDC-LCC Transmission

25

3.3.1. General Aspects

25

3.3.2. Topology and Main Components

25

3.3.3. DC Submarine Cables

27

3.3.4. Converter Technology

29

 

3.3.4.1. 6-pulse Bridge Converter

29

3.3.4.2. 12-Pulse Bridge Converter

31

3.3.4.3. HVDC Schemes

31

3.3.5.

HVDC-LCC + STATCOM

32

3.4.

HVDC-VSC Transmission

34

x

4.

PSS/E Modelling

35

4.1. Introduction

 

35

4.2. Grid Structure

35

4.2.1.

Lines

36

 

4.2.1.1.

400 kV Lines

36

4.2.2. Transformers

37

4.2.3. Loads and Shunt Compensators

37

4.2.4. Diagram of the Grid

38

4.3.

Offshore Wind Farm Model

40

4.3.1.

Wind Turbines

41

 

4.3.1.1. Steady-State Model

41

4.3.1.2. Transient Model

41

4.3.1.3. Fault Ride Through Capability

42

4.3.2. Step-Up Transformers

43

4.3.3. 33 kV Cables

43

4.3.4. Layout of the offshore wind farm

43

4.4.

HVAC Transmission

45

4.4.1. AC Submarine Power Cable

45

4.4.2. Onshore and Offshore Transformers

46

4.5.

HVDC-LCC Transmission

47

4.5.1. Steady-State Model

47

4.5.2. Transient Model

48

4.6.

STATCOM

48

4.6.1. Steady-State Model

49

4.6.2. Transient Model

49

5. Power Flow Results

50

5.1. Introduction

50

5.2. HVAC Power Flow

50

5.2.1. Offshore Compensation

52

5.2.2. Onshore Compensation

53

5.2.3. Offshore and Onshore Compensation

54

5.2.4. STATCOM Compensation

55

5.2.5. Discussion

56

5.3.

HVDC Power Flow

58

5.3.1. HVDC Configuration

58

5.3.2. HVDC+STATCOM Configuration

59

5.3.3. Discussion

60

xi

6.

Dynamic Results

61

6.1. Introduction

61

6.2. HVAC Dynamic Behaviour

62

6.2.1. Case 1 results

63

6.2.2. Case 2 results

64

6.2.3. Case 3 results

65

6.2.4. Case 4 results

66

6.2.5. Discussion

67

6.3.

HVDC Dynamic Behaviour

69

6.3.1. HVDC Configuration

69

6.3.2. HVDC + STATCOM Configuration

71

6.3.3. Discussion

73

7. Conclusions

75

7.1. Final Remarks

75

7.2. Future Work

76

References

78

Appendix A – Grid Parameters

80

Appendix B – GE 1,5 MW Wind turbine aggregate “dyr” record file

87

xii

1. Introduction

This chapter presents an overview of some general aspects regarding wind energy and, more specifically, offshore wind energy. The motivation that led to the execution of this thesis and the main objectives of the work are also presented.

1.1. Motivations

For a long time, conventional energy sources such as coal, oil, natural gas, hydro power and nuclear power were used as main fuels for the production of electricity. The energy crisis that affected the world in the 1970s and 1980s led to major changes in the hitherto energy paradigm. Reducing the world’s dependency of fossil fuels became a priority, as the finiteness of these resources was found to be truly concerning. The need to guarantee diversity and reliability of energy sources, together with the fast growing concern for environmental issues also contributed to a greater interest in alternative energy sources.

The need for a long-lasting solution to fulfil the growing worldwide energy demand, a solution that is both environmentally friendly and economically safe was found in renewable energies. In fact, renewable energy sources such as wind, solar photovoltaic and hydro power, to name a few, have suffered significant increases in recent years. The installed power over the last ten years has increased in such way that renewables account for more than 50% of new installations in the EU, cementing a rising trend initiated over a decade ago [1]. Such data can be verified in Figure 1, where the share of total installed power in the EU in 2000 and in 2009 is presented.

installed power in the EU in 2000 and in 2009 is presented. Figure 1 – EU

Figure 1 – EU power capacity mix: a) 2000 data; b) 2009 data [1].

In Figure 1, wind power clearly stands out as the renewable source that experienced the largest increase in installed power: wind power’s share of total installed capacity has increased from 2% in 2000 to 9% in 2009. These facts reflect a global trend on new power installations, as 2009 was the

1

second year running where more wind power was installed than any other generating technology in the EU [1]. Wind energy is considered one of the most promising renewable energy sources, benefiting from a mature technology, developed especially in Europe and in the USA. It is now a totally established technology that has increased remarkably in recent years. Wind power is, in fact, the fastest growing renewable energy: since 1996 the installed wind capacity increased from 6100 MW to 158505 MW as it is shown in Figure 2.

from 6100 MW to 158505 MW as it is shown in Figure 2. Figure 2 –

Figure 2 – Global Cumulative Installed Wind Capacity (1996-2009) [2].

The fast development of wind energy throughout the world has opened up new frontiers in wind energy generation in the form of offshore wind farms. Placing wind turbines at sea became a possibility, as many of the windiest locations onshore are already occupied. Higher wind speeds at marine locations and wider installation areas, together with the possibility of moving wind turbines away from population – to which wind turbines may cause some discomfort – were the main reasons driving the investment on offshore wind energy.

The Vindeby wind farm in the Baltic Sea, off the coast of Denmark, was the first offshore wind farm in the world. It was built in 1992 and it consists of eleven 450 kW wind turbines. Since then, and in particular in the last 7/8 years, offshore wind power has experienced an enormous growth, with the installation of several new offshore wind farms. A graph representing the development of the offshore wind power installed capacity from 1991 to 2009 can be seen in Figure 3, in which the referred growth can be observed.

2

Figure 3 – Annual and cumulative installed capacity (in MW) of offshore wind power in

Figure 3 – Annual and cumulative installed capacity (in MW) of offshore wind power in Europe (1991 – 2009) [1].

The current trends for offshore wind farms indicate new offshore wind farms will be higher rated (as a result of larger wind turbines and a higher number of wind turbines) and placed further from the sea. The power transmission of large amounts of power over long distances, from the wind farm to the onshore grid, makes the transmission system a key feature of the offshore wind farm installation.

Currently available technologies for the transmission system to shore are high voltage alternating current (HVAC) and high voltage direct current (HVDC). For HVDC connections, there are two technical options: line commutated converter based HVDC (known as HVDC-LCC) and voltage source converter based HVDC (HVDC-VSC). The HVDC alternatives differ in the technology used for the power converters.

The technology used for the interconnection of offshore wind farms influences in a very significant manner the whole operation of the system. As so, the main characteristics of operation for each transmission system have to be assessed and thoroughly studied, including steady-state analysis and dynamic behaviour analysis of the link (when disturbances occur), making this an important field of investigation in the near future.

1.2. Thesis Objectives and Outlook

The main objective of this thesis is to compare the performance in steady-state and the transient behaviour of two alternative transmission systems for offshore wind farms: HVAC technology and HVDC-LCC technology.

In steady-state, the power flow analysis of a grid with an integrated offshore wind farm is performed, for both transmission systems. Special attention is paid to both the power factor and the reactive power flow in the point of connection of the wind farm to the onshore grid. For the HVAC technology, compensation schemes for the reactive power generated by the submarine cable are also

3

evaluated. Compensation with shunt reactors placed onshore, offshore and at both ends of the cable and also the use of a STATCOM are subject to analysis based on simulations. For HVDC-LCC technology, the configuration with a STATCOM as a way of compensating reactive power consumption of the converters is also analysed.

The transient behaviour of the wind farm, and particularly of the connection (AC or DC Link), is investigated. The dynamic behaviour of the system, as a response to onshore grid faults, is tested through simulation as different faults are applied to the grid. The influence of the STATCOM is also assessed.

All

the

analysis

are

done

considering

simulation

results

obtained

after

careful

design,

dimensioning and testing of the offshore grid performed in PSS/E software.

1.3. Thesis Structure

The content of this thesis is divided in seven chapters.

The present chapter, chapter one, is a general introduction to the dissertation, providing information on the current status of offshore wind energy and its integration among the global energy panorama. Motivations, objectives and structure of the work developed in the course of the thesis are also outlined.

Chapter two presents an overview of the most important issues regarding offshore wind farms. The main characteristics of offshore wind farms, main advantages and constraints associated with these wind farms, current status and grid integration issues are also approached.

Chapter three provides a closer look on the transmission system used for the connection of offshore wind farms with the onshore grid. The chapter presents a detailed description of the three available options: High Voltage Alternating Current Transmission (HVAC); High Voltage Direct Current based on Line Commutated Converter Transmission (HVDC-LCC) and High Voltage Direct Current based on Voltage-Source Converter Transmission (HVDC-VSC). Main features, operating issues and limitations of each technology are discussed, with special focus on the HVAC and the HVDC-LCC alternatives.

Chapter four focuses on the modelling carried out in the PSS/E software of the offshore wind farm studied. The models used and the parameters selected for each model, for both the AC and the DC transmission, are described with sufficient detail. Differences in modelling for steady-state analysis and for transient behaviour analysis of the grid are also summarized.

Chapter five presents the results of the power flow simulation of the offshore wind farm. For HVAC transmission, some alternatives for reactive power compensation are reviewed. For HVDC transmission, the possible benefits of a STATCOM are also studied. A comparative analysis of the transmission alternatives in steady-state is also performed.

4

Chapter six presents the results obtained for the dynamic simulation of the wind farm. Dynamic behaviour of the HVAC or HVDC-LCC transmissions to different grid faults on the onshore grid is analysed. The influence of the STATCOM on the dynamic behaviour of the link is studied.

In chapter seven, the main conclusions of the work are presented, together with a reference to future studies that may offer important future contributions to the theme of this thesis.

5

2. Wind Energy Offshore

This chapter will present an overview of some general aspects of wind farms offshore: reasons and motivations for building offshore, their advantages and disadvantages, grid integration and transmissions options to shore.

2.1. Introduction

The increasing requirement to produce electrical power from renewable and clean energy sources and, particularly, the major investment in wind energy made over the last 15 years, led to a growing interest in harnessing the wind power that lies off the coast of many countries. Offshore wind power became, therefore, the next evolution in wind power technology. A number of reasons gave rise to this bulk investment. The offshore wind characteristics lead to a larger produced energy but the difficult conditions at sea, the integration of the wind farm and other technical requirements have to be seriously considered. It is however relevant to notice that after years at the starting block, the offshore wind energy is taking off and becoming a viable alternative for energy production.

2.2. The Wind Resource

Winds are caused by pressure differences across the earth’s surface due to differential solar heating of the earth’s atmosphere. Therefore, wind energy is strongly influenced by solar energy: the amount of solar radiation absorbed at the earth’s surface is greater at the equator then at the poles, resulting in faster heating of the air in the equator [3]. This warm air, heated in the equatorian regions, rises to high altitudes and then flows toward the poles. At about 30ºN and 30ºS, the air begins to cool and sink and so a return flow of air takes place in the lower layers of the atmosphere. Zones of high pressure are created by descending air, as zones of low pressure are formed where air is ascending. This horizontal gradient drives the flow of air from high to low pressure. The greater the pressure gradient, the greater is the force on the air and the higher is the wind speed [4].

Stronger, constant and more persistent winds occur at about 10 km above the earth’s surface. However, as placing turbines at such high altitudes is unfeasible, the area of interest is limited to only a few tens of meters high [3].

In addition to this global phenomenon resulting in wind, there are also local effects to be considered. The nature of the terrain has an important effect. Wind turbines utilise wind energy from a lower area known as boundary layer. In this region, wind speed is retarded by frictional forces on the earth’s surface, which means that wind speed increases with height (this gradient is known as wind shear) [4].

6

2.2.1.

Wind Offshore

As stated previously, the nature of the surface influences the wind profile at a given region. Offshore, the smoothness of the surface of the ocean results in a low surface roughness (a parameter referred to as Z 0 ). The surface roughness length is dependent on the sea state, increasing with the local wave conditions which are, in turn, influenced by wind itself. As a result of a low surface roughness there is also low turbulence and wind shear on a marine environment. Offshore locations are in fact windier and the winds are more persistent than those at an onshore location. These effects increase with distance from shore in the downwind direction, making it more appealing to locate offshore wind farms further from land [5].

Data collected from offshore meteorological stations has been compiled to create the wind offshore atlas of Europe, which is presented in Figure 4. The mean wind speed at each location is given for five different heights: 10, 25, 50, 100 and 200 m from sea level. The highest values of wind speed are for the 200 m height, making it the most economically interesting solution for placing offshore wind turbines.

The geographical distribution of wind speed offshore, given by Figure 4, shows that, apart from a small region in southern France, the North Sea is the windiest location in Europe. In fact, for this reason (and also for the fact that water depth in the North Sea increases slowly with distance from shore) most of the wind farms in Europe are located in the North Sea, as part of solid investments by English and Danish governments in offshore farms.

7

Figure 4 – Europe wind resources over open sea (Offshore Wind Atlas) [6]. As the

Figure 4 – Europe wind resources over open sea (Offshore Wind Atlas) [6].

As the power that can be extracted by a wind turbine is proportional to the cube of the wind speed (as given by Equation (1)), the extracted power is highly dependent on the wind speed [3].

=

1 2

(1)

Therefore, and given the fact that wind speed offshore is potentially higher than onshore, a 10% increase in wind speed results in an approximate 30% increase in power production, making wind power offshore a very interesting option.

According to the European Environment Agency (EEA) the technical potential of offshore wind by 2020 is around 25000TWh, six to seven times greater than the projected electricity demand [1]. These figures suggest that a significant wind resource lies off the coast of many countries. The data presented refers to European estimations but some studies done elsewhere point out similar offshore potential in other regions of the world.

8

2.3. Offshore Wind Farms

Following the rising interest in the wind resource offshore, a way of harnessing the wind was considered and steps were taken towards creating platforms located offshore, in order to gain experience and test problems and potential benefits of this technology. The first offshore wind turbine was then installed in Sweden in 1991. The following year, the Vindeby wind farm, the first of its kind, was installed in shallow water (2-5 m deep) off the coast of Denmark. It consists of eleven 450 kW machines, located about 3 km from the shore [5].

Small wind farms, consisting of few wind turbines (less than 10) with low ratings (around 500 kW) were erected in other European countries but it wasn’t until the 2000s that serious investment was put into the design and construction of offshore wind farms. Denmark has led the way, building the Horns Rev wind farm in 2002, the largest at the time of construction and a reference for following projects.

Many new offshore wind farms have followed, and the remarkable growth in new installations over the last 7/8 years (as can be seen in Figure 3), denotes that, in fact, offshore wind energy has gain relevance as an alternative to consider for energy production.

The fact that offshore wind farms are located at sea offers several advantages but also poses major challenges for potential investors. Some relevant constraints associated with the offshore wind farms that need to be considered are the following:

Installation and maintenance of a large structure located far from shore is a complex procedure, given the distance and the need for special ships;

Accessibility of turbines may prove difficult, as harsh weather conditions and extreme waves are likely to exist at sea;

As a consequence, materials used for wind turbines and additional equipment need to be resistant to physical impact as well as to salt corrosion (when the wind farm is placed in salt water);

Overall, at present time, the construction and maintenance of offshore wind energy is more expensive than on shore, given the higher costs of foundation, installation and electrical connection of wind farms [4].

However, as the number and size of offshore wind farms increase, technology is likely to advance to deal with these problems and reduce transportation and installation costs. The advantages will outweigh the disadvantages making offshore wind farms a reliable investment [4].

Advantages of offshore wind power are as follows:

Wind speeds are higher offshore than onshore, which means that a large potential for power production lies offshore. In fact, the annual average full-load hours offshore is between 3500-4000 hours, while for onshore this figure is between 1500 and 3000 (the average being around 2000 hours) [7];

Winds at marine locations are more persistent and less turbulent than those on land;

9

Due to the proliferation of wind parks onshore, some countries are running out of suitable onshore locations, as the windier sites are already occupied;

A wider area for installation of wind turbines is also likely to be available at sea, compared to land;

Wind turbines that might be intrusive and cause visual and noise impact on land might be acceptable if sited away from shore, far from population.

2.3.1. Current Status

As of January 2010 all offshore wind farms were located in Europe. According to EWEA data [1], there are 38 operational offshore wind farms in Europe, representing a total capacity of 2055,9 MW of installed capacity. Denmark has been a pioneer and a worldwide reference as far as offshore wind farms are concerned. For the last few years, however, the UK became the number 1 country in offshore installed capacity, with a total of 882,8 MW, representing almost 43% of the European total. Complete division of installed capacity by country is presented in Figure 5.

of installed capacity by country is presented in Figure 5 . Figure 5 – Cumulative share

Figure 5 – Cumulative share of installed capacity by country (end 2009) [1].

As stated previously, the North Sea is the windiest region in Europe. The relatively low water depth (less than 20 m) of the sea has also contributed to the widespread installation of offshore wind farms in the region. Figure 6 shows precisely that, as can be seen that most of the operational offshore wind farms are at the North Sea.

10

Figure 6 – Map of operational offshore wind farms in Europe (January 2010) (Adapted from

Figure 6 – Map of operational offshore wind farms in Europe (January 2010) (Adapted from [8]).

Offshore wind energy is a growing market, and a significant number of offshore wind farms are either under construction or planned for the next 10 years. Around 40 GW of installed capacity are expected to exist by 2020 [1].

2.3.2. Technologies and Future Trends

The trends and perspective for future offshore wind farms are likely to demand new technology, creating new challenges for the industry. As the technology develops and experience is gained, future projects will be higher rated and placed in deeper waters, further from shore. Looking at the wind farms proposed by developers, the industry will go beyond the so-called 20:20 envelope (20m water depth, 20 km from shore) in which current wind farms are included.

Some experience gained from wind farm projects in land is useful but offshore wind farms differ significantly from the onshore ones.

Submarine support structures are a very important part of the offshore project, as they represent

a significant proportion of offshore development costs. The choice of structure is determined

essentially by water depth and consistency of seabed. The currently available options for foundation are represented in Figure 7, the more commonly used being the monopole and gravity foundations. For deepwater application some prototypes of floating foundations are under development, with the

first experience being the Floating Hywind in Norway, at a depth of 220 m. These floating devices will

be important as the wind farms move further from shore.

11

Figure 7 – Offshore foundations and general characteristics [9]. Another issue of concern regarding the

Figure 7 – Offshore foundations and general characteristics [9].

Another issue of concern regarding the technology used in offshore wind farms is the wind turbines used. The wind profile offshore (less turbulent and with higher average and extreme wind speeds than onshore) along with challenging access conditions demand special requirements for the design and manufacture of wind turbines. These include robustness, corrosion protection, high reliability and low maintenance requirements.

Recent trends for wind turbines indicate that future offshore wind turbines will be larger machines, with higher rotor diameter and increasingly higher capacity, as indicated in Figure 8. Planned wind farms are expected to include turbines with rated power as high as 5 MW.

to include turbines with rated power as high as 5 MW. Figure 8 – Offshore wind

Figure 8 – Offshore wind turbine size evolution [9].

2.4. Grid Integration

Integrating an offshore wind farm in an electrical network poses a significant challenge to the grid. The impact varies with the strength of the grid and the size of the wind farm. As the offshore wind capacity grows, grid integration issues may arise, as increasingly large amounts of electricity are fed into networks, either in distribution or transmission systems, at points not specifically designed for

12

such infeeds [4]. As such, the transmission technology – AC or DC – used to link the wind farm to the onshore grid is of key relevance to meet the grid requirements of the network.

An overview of the available technologies is presented as well as a description of the main considerations of connecting an offshore wind to the grid.

2.4.1. Transmission Technologies

The installation of offshore wind farms has given rise to new challenges regarding electrical connections, both for the internal electric system of the wind farm and for its connection to the main grid. Existing wind farms onshore are connected via AC cables or overhead lines to the national grids. An AC network within the farm collects the power production of each wind turbine. The voltage level of generation is typically around 700V, which is then stepped up to a medium voltage level (typically 33 kV) by a transformer installed in the nacelle or the tower base [4].

The same scheme was applied to the first wind farms installed offshore. However, as wind farms tend to be larger, with longer distances between turbines and longer distance to shore, alternative arrangements were studied. Special focus is, then, on the transmission from the wind farm to shore. Currently available technologies for transmission system to shore are high voltage alternating current (HVAC) and high voltage direct current (HVDC). For HVDC connections, there are two technical options: line commutated converter based HVDC (known as HVDC-LCC) and voltage source converter based HVDC (HVDC-VSC).

The AC connection has been used in all offshore wind farms. With the increase in the size of wind farms, the transmission to shore is performed at higher voltage levels than the 33 kV used in collection, usual values being 132 kV (used in the Nysted wind farm, in Denmark) or 150 kV (as is the case of the Horn Rev wind farm). The HVAC solution is indeed the most straightforward technical approach as it features some interesting advantages such as ease of interconnection, installation and maintenance; operational reliability and cost effectiveness. Some disadvantages include the limits for transmitted power, associated to the charging current on AC cables, which will limit the maximum distance of transmission. Some scheme of reactive power compensation must then be used in AC cable systems.

HVDC solutions might prove a reliable option for large wind farms further from shore as this technology has some very promising characteristics. One of the most important is the ability to transmit large amounts of power over long distances with lower losses than HVAC. Since the transmission is done using DC cables, the transmission distance is not affected by cable charging current. Main disadvantages include the complexity and associated cost of installation and maintenance of the HVDC system, which may well be overcome in the future, due to economies of scale.

In the line commutated based HVDC the conversion is performed with thyristor converters. The HVDC-LCC is a totally established technology, as the first HVDC link dates back to 1954 in Gotland,

13

Sweden. Such proven track record has become one of the most important advantages of this option, in addition to the features referred in the previous paragraph.

The HVDC-VSC transmission is a relatively recent technology that uses insulated gate bipolar transistors (IGBT) converters for the inversion/rectifier operations. This technology was made possible by advances in high-power electronics and is used as a viable alternative to thyristor based LCCs used in “Classic HVDC”. Unlike line-commutated converters, VSCs do not need an AC source commutation voltage; the independent control of both active and reactive power and the reduced injection of harmonics also represent important advantages of HVDC-VSC over HVDC-LCC [4].

More detailed description of the working principles of each of the referred technologies – with special emphasis on HVAC and HVDC-LCC – is presented in Chapter 3.

2.4.2. Grid Connection Requirements

Until some years ago wind farms were allowed, or even required, to disconnect from the grid during a disturbance in the grid [10]. This has changed significantly, specially due to the proliferation of large amounts of installed wind power capacity and, predictably in a near future, installed capacity offshore. The disconnection of a large offshore wind farm would result in a significant loss of generation that could cause some stability problems to the network. Transmission system operators require nowadays for offshore wind farms to stay connected under certain disturbances in the grid. These requirements are known as the fault ride through capability of the wind farm and are generally regulated in grid codes. As established in most grid codes, only under certain circumstances shall wind farms be disconnected from the grid following a grid fault, remaining otherwise connected in order to assist in the stabilization of the grid frequency or the voltage during fault, providing voltage back-up.

Apart from the fault ride through capability, other technical requirements must be fulfilled by the wind farm, since the increasing size of offshore wind farms means that the rating of such installations will be comparable to that of traditional generating plants on the grid. These requirements include [11]

[12]:

Control of active and reactive power (operation under a specified range for power factor);

Frequency range (with time durations for extreme conditions, permissible reduction at frequency extremes, if any);

Contribution to network stability;

AC voltage control capability.

As the proliferation of offshore wind power increases, wind farms will be bound to meet these demands, which may prove difficult depending, to great extent, on the transmission system used between the wind farm and shore. The charging currents affecting AC cables represent a limitation for

14

the HVAC option and so some form of compensating the surplus reactive power generated by the cable may prove necessary to met grid requirements. As far as HVDC is concerned, capacitive currents are nonexistent but other issues are of concern, such as the behaviour of the DC link following an onshore grid fault and the power factor at the connection point in HVDC-LCC, as thyristor based converters absorb reactive power. FACTS devices can be used to reduce or eliminate these problems, therefore enhancing the performance of the HVDC system.

Further aspects of the issues concerning connection, both in HVAC and in HVDC-LCC will be discussed in Chapter 3.

15

3. Transmission Technologies for Offshore Wind Farms

This chapter provides an overview of the available transmission systems for offshore wind farms. Several aspects of the connection in either AC or DC are discussed.

3.1. Introduction

One of the most important aspects of the design of offshore wind farms is the connection to the electrical grid located in land. The electrical energy generated in the offshore wind farm requires one or more submarine cables to transmit the power to the onshore utility grid that services the end-users.

Depending on the wind farm size and distance from shore, the connection poses a challenge, not only to the wind farm developer, but also to the onshore network itself, as some grid code requirements must be met.

There are currently three different transmission technologies available:

HVAC – High Voltage Alternate Current;

HVDC-LCC – High Voltage Direct Current with Line Commutated Converters;

HVDC-VSC – High Voltage Direct Current with Voltage Source Converters;

These technologies present significant differences, on the operation mode and the components used. A detailed explanation of the options is presented throughout this text, in 3.2, 3.3 and 3.4 Special focus is given to the HVAC and HVDC-LCC technologies.

The general layout of an offshore wind farm is represented in Figure 9. This scheme can refer to both the HVAC and the HVDC transmission. The components present at the “collecting point” and at the “wind farm grid interface” vary according to the technology used to connect the generation to the PCC (Point of Common Coupling) with the grid.

to the PCC (Point of Common Coupling) with the grid. Figure 9 – General layout of

Figure 9 – General layout of an offshore wind farm [13].

16

As of 2010 all operating offshore wind farms are connected via AC cables. However, as wind farms become larger and are placed further from the coast, the use of HVDC solutions will become technically and economically interesting alternatives.

3.2. HVAC Transmission

3.2.1. General Aspects

Connecting the wind farm to the grid by an AC cable is the most straightforward technical solution, as both the power generated by the wind farm and the onshore transmission grid are AC. In fact, all the operational offshore wind farms use HVAC for the connection link to shore. The HVAC transmission offers some advantages over the DC solutions such as [11] [14]:

Proven and low-cost technology;

Easy to integrate in existing power systems;

Low losses over small distances.

On the other hand there are some constraints of the HVAC system that limit significantly the use of this technology, namely [11] [14]:

There is an excessive amount of reactive power produced in the AC submarines cables;

Increase in the cable length means increase in its capacitance which results in a reactive power increase, resulting in a transmission distance limit for AC systems;

Necessary use of compensation systems (shunt reactors, STATCOMS, SVC, etc) at the ends of the cable;

Load losses are significantly higher for longer distances;

For large wind farms several cables may be necessary, increasing line losses.

3.2.2. Topology and Main Components

A transmission system based on HVAC technology includes the following main components [13]

[15]:

AC based collector system within the wind farm;

Three core XLPE HVAC submarine transmission cable(s);

Offshore transformer(s);

Reactive power compensation (onshore and/or offshore);

17

Onshore transformer;

Figure 10 illustrates the layout of an HVAC based offshore wind farm, with the components mentioned above.

offshore wind farm, with the components mentioned above. Figure 10 – Layout of an HVAC wind

Figure 10 – Layout of an HVAC wind farm[16].

The voltage used in the collecting system in the wind farm is usually medium voltage, in the range of 33-36 kV. For relatively small wind farms, with short distances to shore, the offshore transformer may not be necessary, so the link is done at the same voltage used in the wind farm.

3.2.3. AC Submarine Cables

The submarine cable used in the transmission system is one of the most important components of the system, as the cost of the cable represents a large fraction of the total cost of the offshore wind farm investment. The laying of the cable is a complex process performed with special laying vessels, so as to guarantee the full operation of the system. The potentially severe conditions at sea and other hazards, such as fishing or dropped objects, also demand special attention on the mechanical resistance of the cable. Technically, different cable types have different characteristics that influence the transmission and can even limit the maximum distance of transmission. A submarine cable has typically the following structure [14]:

Conductor core, typically copper or aluminium;

Electrical Insulation, either solid dielectric (XLPE cables) or oil impregnated paper (OIP);

Shielding, a conductor layer of paper or polymer;

Sheathing, metallic layer used to ground the cable and also protect it from water;

Armour, an outer metallic armature, to increase mechanical resistance;

Optic Fibre, for communications;

The link can be achieved with either three single-core cables or one three-core cable (more than one may be used, depending on the transmitted power). The latter is the most common type, as it

18

represents lower cable and installation costs and lower electromagnetic fields and induced current loss, compared with separate cables.

The main types of AC cables differ in the electrical insulation used in the cable. Historically, cables used lapped paper insulation impregnated with insulating oil. These cables are:

High-pressure pipe-type, either fluid-filled (HPFF) or gas filled (HPGF);

Low-pressure oil-filled (LPOF), also referred as self-contained liquid-filled (SCLF) and Low-pressure fluid-filled (LPFF);

The need of a pumping system in both ends of the lines to circulate the fluid and the environmental risk of pipe rupture, resulting in oil spill are serious disadvantages of these cables. As so, these cables are only considered for very short distances.

Cross-linked polyethylene cable (XLPE) cables are currently the most cost-effective and modern alternative. The insulation is made of solid dielectric (also called extruded dielectric or polymeric insulated) and it presents several advantages over the fluid-filled options: higher mechanical resistance, lower weight and, consequently, easier installation process [14]. Furthermore, XLPE cables have lower capacitance and lower dielectric losses than the fluid-filled cables.

One three-core XLPE cable is shown in Figure 11, where the structure described above can be seen.

Figure 11, where the structure described above can be seen. Figure 11 – Three-core XLPE submarine

Figure 11 – Three-core XLPE submarine cable [17].

The major problem concerning the connection of wind farms with AC submarine cables is the fact that cables generate significant amounts of reactive power. This reactive power is produced by the high shunt capacitance of cables (significantly higher than overhead lines). In the AC system, the cable must carry the load current and the reactive current generated by the cable capacitance, which reduces the power rating of the cable.

Considering the cable is modelled by an equivalent π-model, the submarine cable can be represented by the scheme shown in Figure 12. The charging current at each end are represented by and .

19

Figure 12 – Equivalent π -model for a submarine cable. Neglecting the voltage drop across

Figure 12 – Equivalent π-model for a submarine cable.

Neglecting the voltage drop across the resistance and the reactance of the cable, one can consider that = = . As so, the total charging current affecting the cable is the sum of and

(as represented in Figure 12) and is represented by , which is given by Equation (2):

In the equation,

=

is the angular frequency (rad/s),

C is the capacitance (F)

V is the terminal voltage (V).

(2)

From Equation (2) it is clear that, as the cable system voltage is increased to minimize losses, the charging currents also increase, worsening the situation.

The reactive power generated by the shunt capacitance is then given by Equation (3):

= −

(3)

Note that the negative signal means the capacitance supplies reactive power to the grid. From Equation (3) it can be seen that the reactive power generated by the cable is a function of the capacitance and the voltage, being particularly dependent on the voltage (since it is squared).

Since the cable capacitance is distributed along the entire length of the cable, the longer the cable the higher the capacitance and the resulting generated reactive power. As a consequence, the load carrying capability of the cable is reduced, which means that a cable can only transmit a certain amount of power for a given distance. This results in an obvious limitation for the length of AC links. The maximum rating for a three-core submarine cable with a voltage rating of 150-170 kV is limited to about 200MW, with compensation at both ends, and a maximum cable length of 200 km [11].

A comparison of the transmission capacity of different cables operated at certain voltage levels (132, 150, 230 and 400 kV) is presented in Figure 13. The maximum distances for an AC submarine cable system varies according to the rated voltage for the link and the maximum transmitted power decreases with the length of the cable.

20

Figure 13 – Maximum transmission lengths with inductive shunt compensation in both ends [17]. The

Figure 13 – Maximum transmission lengths with inductive shunt compensation in both ends [17].

The transmitting distance represents an obvious limitation for the HVAC system using a submarine cable. Compensating the reactive power produced by the cable can be a solution since it increases significantly the maximum distance of transmission. Reactive power compensation options are discussed in 3.3.4.

The capacitance of AC cables may also lead to other undesirable effects, like overvoltages, high harmonic currents and resonances between the cable’s capacitance and the reactance of the generators. Real power losses within the cable also limit the distance for HVAC cable transmission. These losses in the submarine cables derive from dielectric losses (relatively small) and ohmic losses in the conductors (the higher component of losses), the metallic shield and the steel wire armour [14].

3.2.4. Reactive Power Compensation

The solution for the large amounts of reactive power at the cable is to compensate the reactive power produced by absorbing reactive power, thus reducing the additional losses and increasing the maximum transmitting distance. The compensation is usually realized by fixed or electronically controlled shunt reactors. The fixed shunt reactor is the simplest device but the progress in FACTS (Flexible AC transmission system) devices, such as SVC (Static VAr compensator) or STATCOM (Static Synchronous Compensator), considerably extends the reactive power and voltage control possibilities offered by the switched shunt reactors.

Shunt reactors are the most commonly used devices as they represent simple and robust solutions with low installation costs. They also have the advantage of requiring no transformer for the connection, thus having no additional power losses. One of the disadvantages is that the reactors are

21

designed for a single operational mode, usually to compensate the cable at full load. Another disadvantage is the fact that the reactive power absorbed by the shunt reactor is proportional to the square of the terminal voltage, as demonstrated by Equation (4), where L stands for the inductance of the reactor.

3.2.4.1. FACTS Devices

=

(4)

The SVC and the STATCOM are part of the FACTS device family, used for voltage regulation and power system stabilization, based on power electronics. These devices are capable of both generating and absorbing reactive power. The flexibility of use is, therefore, the main advantage of these equipments, since they allow the continuous variable reactive power absorption (or supply). The reactive power is not proportional to the voltage at the bus, also another advantage of these equipments. The FACTS devices can also contribute in the improvement of the voltage stability and the recovery from network faults.

The similarity of the SVC and STATCOM devices led to them being sometimes referred generally as “Static VAr Compensators”. These are, however, different equipments. The SVC (“SVC Classic” for some manufacturers) is based on conventional capacitor banks together with parallel thyristor controlled inductive branches. These inductive branches can either be TCR (Thyristor Controlled Reactor), used for linear injection of reactive power or TSC (Thyristor Switched Capacitor), used for stepwise injection of reactive power. A SVC device is represented in Figure 14, where a linear diagram and an SVC installation (in an offshore wind farm) are represented.

SVC installation (in an offshore wind farm) are represented. Figure 14 – (a) Schematic Diagram of

Figure 14 – (a) Schematic Diagram of an SVC; (b) an SVC in Radsted, Denmark [18].

The STATCOM device uses a power electronic voltage source (VSC). The converter uses semiconductors with turn-off capability, such as Insulated Gate Bipolar Transistors (IGBTs). The benefits of the STATCOM (commercially known as “SVC Light” by ABB or “SVC Plus” by Siemens), compared with the SVC, are the fact that the capacitor banks used are smaller and also there is no

22

need for big air-cored inductors. Further advantages of the STATCOM are also found in the dynamic behaviour (such as faster transient response). A simplified schematic diagram of a STATCOM and an example of an installation is shown in Figure 15.

and an example of an installation is shown in Figure 15. Figure 15 – (a) Schematic

Figure 15 – (a) Schematic Diagram of a STATCOM; (b) a STATCOM installation [18].

3.2.4.2. Distribution of the Compensation

The distribution of the reactive power compensation devices is also an important matter. Figure 16 presents several options for compensation: on one end only (offshore, in the case), on both ends or distributed along the cable. The figure also compares the loading of an open-end cable. It is obvious that the worst case is with compensation installed at one end only. The best solution is to install distributed compensation along the cable, since the charging current generated would flow towards the reactor, not increasing the rating of the cable so significantly. It is however important to notice that the placement of reactive power compensators in the middle of the cable is likely to not be possible, as the installation of these interstitial reactive power compensators would present an added technological and economic challenge.

23

Figure 16 – Loading of an open-end cable due to capacitive charging current for different

Figure 16 – Loading of an open-end cable due to capacitive charging current for different schemes of reactive power compensation [19].

24

3.3. HVDC-LCC Transmission

3.3.1. General Aspects

Current trends in offshore wind farms indicate future installations will be rated at hundreds of MW and further from shore (as far as 200/300 km). These transmission requirements will not be feasible using AC submarine cable transmission, given the limitations mentioned in 3.2. Connecting offshore wind farms through a DC link has then been considered since there is a significant amount of experience in transmitting large amounts of power over long distances through HVDC links.

There are two different HVDC transmission technologies: the line commutated converter HVDC (HVDC-LCC, also referred as “Conventional” or “Classic” HVDC) using thyristors in the converters and the voltage source converter HVDC (HVDC-VSC) which uses IGBTs.

The main advantage of HVDC-LCC over the HVDC-VSC is its proven track record, since there is an accumulated experience of decades for this technology. The first commercial HVDC-LCC connection was installed in 1954 and since then many other conventional HVDC links were installed all over the world.

Some of the most important advantages that HVDC-LCC transmission offers over AC are [20]

[21]:

Asynchronous connection, since sending and receiving end frequencies can differ;

Transmission distance using DC is not limited by cable charging current;

Low cable power losses;

Higher power transmission capability per cable;

Power flow is fully defined and controlled;

HVDC does not transfer short circuit current.

Some of the constraints of HVDC-LCC transmission are the following [20] [21]:

No experience in connecting offshore wind farms;

Production of harmonics in the converter, making the use of filters necessary;

The converters at each end consume reactive power.

3.3.2. Topology and Main Components

The HVDC-LCC transmission system for an offshore wind farm can be represented by Figure 17.

25

Figure 17 – Basic Configuration of a wind farm using an HVDC-LCC [7] Note: F=filter;

Figure 17 – Basic Configuration of a wind farm using an HVDC-LCC [7] Note: F=filter; HFF=High-frequency filter.

An HVDC-LCC transmission system consists of the following main components (represented in Figure 17) [11]:

AC based collector system within the wind farm;

Offshore three-phase two-winding converter transformers;

Auxiliary power set (STATCOM or Diesel generator);

AC and DC Filters;

Smoothing Reactor;

DC submarine cables;

Onshore converter station (with transformer and LCC converter) and filters.

The AC collector system is typically rated at 33 kV, as in the HVAC wind farms. This medium voltage is stepped-up to the necessary transmission line voltage in the offshore transformers. Usually the transformers are star-star and delta-star connected to power the 12-pulse converter, as this configuration cancels several harmonics. The design of the transformers is critical as their insulation must withstand the AC component of the voltage and the DC component from the thyristor valves. Tappings are also included for proper system control.

The LCC power converters (onshore and offshore) are the most important elements in the system, as they perform the AC/DC conversion offshore and the DC/AC conversion onshore. The LCC converters are based in thyristor valves, capable of standing 8 kV and DC currents up to 4 kA. With these characteristics it is possible to convert up to 1000 MW for land connections and 500 MW for submarine transmissions [14]. In the LCC converter the current is always lagging the voltage due to the control angle of the thyristors; hence these converters consume reactive power. For this reason, reactive power compensation is necessary at both ends to provide reactive power to the system.

26

Capacitor banks or STATCOM devices are considered for this effect. Detailed explanation of the functioning of the thyristor converter is provided in 3.3.4.

A line-commutated converter requires an AC voltage source for its commutation, so an auxiliary power set is required at the offshore station to provide the necessary commutation voltage for the HVDC-LCC connection, when there is little or no wind. The auxiliary power set is also used to supply power to other devices in the wind farm, when the wind farm is disconnected from the grid.

The AC filters mentioned are used to absorb harmonic currents generated by the HVDC converters, thus reducing the impact of the harmonics on the AC system. These filters also supply reactive power to the converter station. The DC filters are used to avoid the generation of circulating AC currents in the cable.

The smoothing reactors are large inductances connected in series with each pole of the DC link. They are used to prevent current interruption at minimum load, limit the DC fault current, reduce voltage and current harmonics and also prevent resonance in the DC circuit.

The DC submarine cable is approached in detail in section 3.3.3.

3.3.3. DC Submarine Cables

The submarine cable is one of the most important components of the HVDC-LCC system, since it is the link for power transmission from the offshore wind farm.

The elements comprising a DC cable are the same as the ones in an AC cables. As so, like the HVAC cables, the HVDC cable main technologies available differ in the electrical insulation used:

Oil Filled (OF) cables;

Mass Impregnated (MI) cables;

Cross-linked Polyethylene (XLPE) cables.

These cables are represented in Figure 18.

27

Figure 18 – Available HVDC Cables: a) OF cable; b) MI cable; c) XLPE cable

Figure 18 – Available HVDC Cables: a) OF cable; b) MI cable; c) XLPE cable [22].

In Oil Filled cables (also known as SCFF – Self Contained Fluid Filled) the insulation consists of paper impregnated with low-viscosity oil. The core of the cable is covered by a hollow shaft where oil is circulated by pumps at both ends of the line. Oil Filled cables can be used to DC voltages up to 600 kV and great sea depths. The need for pumping at each end of the system may prove difficult, and may limit significantly the maximum length of the cable to less than 50 km [14]. The danger of oil spill and the need for cable protection are also obvious disadvantages of the Oil Filled cable technology.

Mass Impregnated cables have similar construction but the paper insulation is impregnated in resin and high viscosity oil and no oil circulation system is required. It is the most used cable technology in existing HVDC systems and so the track record and high reliability are some of the advantages of the MI cable.

As can be seen in Figure 19, the MI cable consists of different layers that include, in addition to the conductor and the paper insulation, layers used for mechanical protection of the copper conductor:

the polyethylene and lead sheaths, jackets and steel armour.

conductor: the polyethylene and lead sheaths, jackets and steel armour. Figure 19 – Mass Impregnated HVDC

Figure 19 – Mass Impregnated HVDC Cable [22].

28

The MI cable technology is available for voltages up to 500 kV and transmission power of up to 800 MW. The capacity of the cable is only limited by the conductor temperature of about 50ºC.

In XLPE cables the insulating material is made of solid dielectric, also known as extruded dielectric. It is a relatively new technology, developed to overcome some of the limitations of the previously referred technologies. XLPE cables have all the advantages of the MI cables and additionally can carry nominal current with a cable temperature of 90ºC. Increased bending capability, higher mechanical resistance and lower weight are also advantages of the XLPE cables, since the installation process is easier than for other cables.

3.3.4. Converter Technology

HVDC-LCC employs line commutated, current source converters, with thyristor valves. The use of the thyristor grants this HVDC technology a limited operating flexibility, since the lack of turn-off controllability of the conventional thyristor results in poor power factors and considerable waveform distortion. Thus, some reactive power compensation and filtering must be provided.

3.3.4.1. 6-pulse Bridge Converter

The basic building block used for HVDC conversion is the three-phase, full-wave bridge referred to as a 6-pulse or Graetz bridge, as represented in Figure 20. The term 6-pulse is due to six commutations or switching operations per period resulting in a characteristic harmonic ripple of 6 times the fundamental frequency in the dc output voltage. Each 6-pulse bridge is comprised of 6 controlled switching elements or thyristor valves. Two thyristors are simultaneously conducting at each instant.

thyristors are simultaneously conducting at each instant. Figure 20 – Six-pulse Converter Bridge [23]. The following

Figure 20 – Six-pulse Converter Bridge [23].

The following analysis is done for the rectifier mode of operation, but similar considerations can be made about the inverter mode. In order to conduct, the thyristor valve must be positively biased

29

(with a positive voltage between the anode and the cathode) and also a pulse has to be applied to the

gate. Therefore, the current will only flow through the valve once the firing pulse has been applied.

This enables the possibility of controlling the instant when current begins to flow through a valve, or to

commutate from one valve to another, by postponing the firing pulse. The angle between the instant at

which the thyristor valve becomes forward biased and the start of commutation (given by the firing

pulse) is referred to as the firing delay angle, α. The method of delaying the firing instant allows the

change in the average value of the DC voltage (in Figure 20) and, consequently, of the transmitted

power. As so, the average output voltage of the converter, working as a rectifier, is given by equation

(5) [24].

Where:

=

3√6

∙ cos

is the DC output voltage;

(5)

is the line-to-neutral rms commutating voltage referred to the secondary of the

converter transformer;

is the firing angle;

is the commutating resistance;

is the output DC current.

Note that the equation includes the commutation resistance that accounts for the internal voltage

drop in the converter due to commutation. The commutation delay accounts for the fact that the

commutating operation is not instantaneous, and so there is a small interval where three valves are

conducting at a time. The angle associated with this delay in commutation is called commutation

angle, µ. The value of µ is generally in the range of 15º to 25º [24].

The power factor of the converter can be proved (in [24]) to be given by equation (6):

3

= ∙ cos

(6)

Therefore, it is of great interest to use a reduced value of firing delay angle, in order to have a

high power factor. The value of α is usually between 10º and 20º [24]. As the firing delay angle

increases, the converter absorbs more reactive power from the AC grid.

The inverter performs the DC/AC operation and is located at the onshore converter station.

The analysis carried out for the rectification process is equally applicable to the inverter operation,

although the inverter equations are expressed in terms of the advance angle β (β =180º-α) and the

extinction angle γ (γ= β - α). The advance angle defines the moment at which the valves of the inverter

30

start conducting and the extinction angle refers to the interval between the end of commutation and the next zero crossing of the commutating voltage.

3.3.4.2. 12-Pulse Bridge Converter

The series-connected double bridge converter has become the standard configuration in HVDC transmission. This converter is represented in Figure 21.

transmission. This converter is represented in Figure 21. Figure 21 – Diagram of a 12-pulse converter

Figure 21 – Diagram of a 12-pulse converter [23].

The 12-pulse converter is a series-connection of two 6-pulse converter bridges and it requires two 3-phase transformers, one with a star-star and the other with star-delta (i.e. with 30º phase shift between them).

This arrangement has some advantages when compared to the 6-pulse bridge converter, namely the fact that the converter injects harmonic currents of orders n=12k±1 into the AC system, and so no filtering of lower order harmonics, like the 5 th and 7 th is necessary (like in the 6-pulse converter) [24]. The alternate voltage supplied by the AC network is also more sinusoidal, meaning it has lower harmonic content, requiring less filtering. The voltage on the DC side repeats itself 12 times during the cycle, making it closer to the ideal continuous voltage. In the 6-pulse converter the voltage repeats itself six times during a period at AC grid frequency.

3.3.4.3. HVDC Schemes

There are two basic configurations used in HVDC systems: the monopole and the bipolar configurations, shown in Figure 22.

31

Figure 22 – HVDC configurations: a) Monopole; b) Bipolar. In the monopole configuration one single

Figure 22 – HVDC configurations: a) Monopole; b) Bipolar.

In the monopole configuration one single DC cable is used. The return path is assured by sea electrodes. Alternatively, due to high earth resistivity, metallic structures in the proximity or

environmental constraints, a metallic conductor may be used, either in the form of a low voltage AC or

DC cable or integrated in the cable itself.

In the bipolar scheme, two DC cables with opposite polarities are used and so the return path is guaranteed. This configuration is the most common in HVDC transmission, since it offers some advantages over the monopole option, namely higher transmission capacity and higher availability, since during maintenance or outages of one pole, it is still possible to transmit part of the power. More than 50% of the transmission capacity can be used, limited only by the overload capacity of the pole

[23].

3.3.5. HVDC-LCC + STATCOM

As referred previously, in a LCC transmission systems both converters absorb reactive power, as in the converters the current always lags the line voltage due to the requirements for a positive commutation voltage at the firing of the thyristors. Thus, this type of converter needs reactive power

for operation, which has to be provided by reactive power devices connected to the AC network. As

these converters depend on the line voltage for commutation, a minor disturbance in the AC voltage

might also result in commutation failures in the converters.

Capacitor banks or synchronous compensators may be used for reactive compensation at both ends of the DC link. A more advanced solution consists of a STATCOM. The STATCOM provides the

32

necessary commutation voltage to the HVDC converter, continuous AC voltage control, fast reactive power compensation to the network under transient conditions and removal of possible non- characteristic harmonic interactions [18]. The STATCOM can also provide limited active power support to the network during transient conditions, such as active power changes of the wind farm output. However, the ability to provide active power support depends on the energy storage on the DC side, which in the case of a conventional DC capacitor will be very limited. Larger energy storage could be provided by means of batteries or SMES (Superconductive Magnetic Energy Storage).

The HVDC-LCC transmission system may benefit significantly from the STATCOM technology since it can, in steady state, provide the reactive power consumed by the converter. Offshore, the STATCOM can also work as an auxiliary power set. In AC grid fault conditions, the STATCOM can provide reactive power to the network, thus contributing to a less severe voltage dip off the onshore bus [12]. The working principles of a STATCOM device are mentioned in section 3.2.4.1.

33

3.4. HVDC-VSC Transmission

HVDC with voltage source converters is a fairly recent technology (first installed in 1999) and it is known commercially as “HVDC Light”, for ABB and “HVDC Plus”, for Siemens. The basis of the HVDC-VSC technology is the self commutated converters, with IGBT (Insulated Gate Bipolar Transistor) valves. Figure 23 depicts a general layout of an HVDC-VSC system applied to an offshore wind farm.

of an HVDC-VSC system applied to an offshore wind farm. Figure 23 – Basic Configuration of

Figure 23 – Basic Configuration of a wind farm using an HVDC-VSC [11].

VSC converters are self-commutating, not requiring an external voltage source for its operation. Also, the reactive power flow can be independently controlled at each AC network and the reactive power control is independent of active power control. These features make VSC transmission an interesting option for connection of offshore wind farms. In addition to the referred HVDC advantages, the VSC technology also offers the following main benefits [11] [14]:

Total control of active and reactive power;

Minimum risk of commutation failures;

Smaller size of converters and filters than HVDC-LCC;

Possibility of the converters starting with a dead grid, not needing any start-up mechanism (“Black-start” capability).

However, there are some constraints to the use of VSC technology. The main constraint is the considerably lower experience in HVDC transmission, compared with the LCC option. HVDC-LCC has more than 30 years of service experience, with high availability rates.

Also, VSC transmission has higher system power losses, when compared with an LCC system. Typically, the power loss at full load for the converters is about 4-6% for VSC and 2-3% for LCC transmission [11]. The IGBT valves are also much more expensive than the power thyristors used in LCC technology.

Apart from these disadvantages, some of which easily surpassed once VSC links become widespread and further experience is gained, the voltage source converter technology is a promising option for HVDC in years to come.

34

4. PSS/E Modelling

This chapter describes the models used in the PSS/E software for the simulation of the alternatives for grid connection of an offshore wind farm. The simulations performed in this software are also described.

4.1. Introduction

In order to simulate the behaviour of an offshore wind farm and of the required transmission system, appropriate models of the wind farm, the transmission system (either HVAC or HVDC) and the electrical grid have to be constructed. Only then is it possible to analyse the steady-state and the transient behaviour of the power system.

The software used to model the system and perform the simulations in the work of this thesis is the PSS/E software. PSS/E stands for Power System Simulator/Engineering and it is a software tool provided by Siemens Power Technologies International (PTI). It is used by most utilities in the world to perform power system simulations, as it allows the performance of power flow analysis, dynamic simulations and stability studies, among other features. PSS/E is composed of a comprehensive set of programs for studies of power system transmission network performance in both steady-state and dynamic conditions, obvious reasons for its widespread use by transmission and distribution systems operators.

In the particular case of this thesis, the PSS/E version 30 was used for the features it offers for network simulation: the extensive library of power system components, including wind turbine, DC links and FACTS are extremely useful for the simulations performed in the scope of the thesis. The wide range of possibilities available in PSS/E modelling increases significantly the complexity of the simulation but also add to the interest of this software for network modelling.

4.2. Grid Structure

The electrical grid that will be used for steady-state and transient analysis – and in which the offshore wind farm will be integrated – is the “savnw” network, provided by PSS/E as an example of a relatively large grid, as it has 23 buses, 6 generators and 7 loads. In order to properly use this grid, a few changes were made, namely to the operating frequency and voltage levels. These changes were made taking into account values used in the European electrical grid.

Additional changes to the original “savnw” network were performed since the offshore wind farm added (rated at 110MW) will be placed in a bus where the existing generation is of 750 MW. Therefore, adjustments in active and reactive power of the network are required, which can be made by adapting the power of the loads and the shunt compensators.

As so, the changes made to the original network can be summarised:

35

The frequency was changed to 50 Hz (the frequency in use in most parts of the world, including Portugal) since the frequency of original example grid is 60 Hz (typical of American grids);

The voltage levels were changed from 500 kV, 230 kV and 22 kV to, respectively, 400 kV, 220 kV and 33 kV. These voltage levels are common in most European Grids, including Portugal;

The 110 MW offshore wind farm was added, replacing an existing power plant of 750 MW;

The inductive shunt compensators were removed;

The parameters of the 400 kV lines were adapted;

In order to accommodate for the changes reported above, some network parameters have to be adapted, which are described in the next sections. Most of the parameters of the grid are presented in detail in Appendix A.

4.2.1. Lines

As a result of changing the grid frequency to 50 Hz, the reactance and the susceptance of the branches have to be converted, since these parameters depend on the grid frequency. This conversion is done using Equations (7) and (8).

=

×

= ×

(7)

(8)

The line parameters (except those of the 400 kV lines) used in the test grid are presented in Table 10, in Appendix A.

4.2.1.1. 400 kV Lines

The parameters of the 400 kV lines were changed to values of 400 kV lines existing in the Portuguese grid. This decision was made upon studying the values of the existing 400 kV line parameters and verifying these resemble the ones of a power cable, since significant amounts of reactive power were being produced by the line. This issue does not occur in the 220 kV and 33 kV lines.

Considering

the

base

power

impedance is = = 1600 Ω.

= 100

36

and

the

base

voltage = 400 ,

the

base

The line parameters: resistance, reactance and susceptance in p.u. can, respectively, be calculated with Equations (9), (10) and (11).

Where:

=

=

×

×

= ( × ) ×

is the line resistance in Ω/ ;

is the line reactance in Ω/ ;

is the line susceptance in S/ ;

is the line length in ;

is the base impedance in Ω;

(9)

(10)

(11)

The values for the 400 kV line parameters calculated through the equations above are shown in Table 11, in Appendix A.

4.2.2. Transformers

For the two-winding transformers in the “savnw” grid, the transformer reactance is frequency dependent, so their values also have to be adapted to the 50 Hz frequency, according to Equation (7).

Tap changing was used in the network transformers, changing the voltage magnitude of the transformers, and, thus regulating properly the reactive power flow.

The values of the results of the calculations for the transformers in the grid are shown in Table 12, in Appendix A.

4.2.3. Loads and Shunt Compensators

As mentioned previously, adding the offshore wind farm rated at 110 MW to bus 102 of the grid in the place of the existing 750 MW conventional power plant leads to an imbalance of active and reactive powers. This imbalance can be compensated by reducing the consumed power in the loads. As so, the values of the load powers of the grid are presented in Table 13, in Appendix A.

The shunt compensators are also regulated in order to maintain the voltage in the buses at values within reasonable values (between 0,95 p.u. and 1,05 p.u.). Table 14 (in Appendix A) outlines the values of the shunt compensators of the grid.

37

4.2.4.

Diagram of the Grid

The resulting grid, after the changes performed, is presented in the single-line diagram shown in Figure 24, in which the offshore wind farm is already included. Bus 102, referred above, is marked in the figure. The legend of the colours used in the figure for each voltage value is presented in Table 1.

the figure for each voltage value is presented in Table 1. Figure 24 – Single line

Figure 24 – Single line diagram of the test grid used (with HVAC transmission, compensations at both ends).

38

Table 1: Colour correspondence for each voltage level in Figure 24.

Colour

Rated Voltage (kV)

Red

400

Black

220

Purple

150

Dark Green

33

Blue

20

Orange

13,8

Bordeaux Red

0,7

39

4.3. Offshore Wind Farm Model

The model for the offshore wind farm is based in an existing installation: the Lillgrund Wind Farm. This wind farm is located off the coast of Sweden, at a distance of 9 km from the Point of Connection in the onshore grid and the transmission is achieved by a combination of an AC sea cable (7 km long) and an AC land cable (2 km long). With 48 wind turbines, rated at 2,3 MW each, the total capacity of the wind farm is 110 MW. Figure 25 depicts the layout of the wind turbines in the Lillgrund wind farm, used as a reference.

turbines in the Lillgrund wind farm, used as a reference. Figure 25 – Layout of the

Figure 25 – Layout of the wind turbines in the Lillgrund wind farm [25].

As can be seen in Figure 25, the internal grid of the Lillgrund wind farm consists of 33 kV sea cables divided in five feeders and each of these feeders connects 9 or 10 wind turbines to the offshore substation. The total is 48 wind turbines rated at 2,3 MW.

The modelled offshore wind farm is based on this layout. The turbines are joined in aggregates and then connected to the offshore substation bus via 33 kV AC cables.

40

4.3.1.

Wind Turbines

The wind turbines used are the GE 1,5 MW model, available in the PSS/E Wind package. This model is of a DFIG (Doubly Fed Induction Generator) wind turbine developed by General Electric and released for PSS/E simulation and testing.

In order to match the 110 MW of the reference wind farm, a total of 74 1,5 MW wind turbines were used in the model. These are joined in five aggregates: two aggregates of 24 MW (16 wind turbines for each aggregate) and three of 21 MW (14 wind turbines per aggregate). This actually adds up to 111 MW, which is the value assumed hereby for the offshore wind farm power.

For PSS/E simulation of the wind farm, two distinct models are designed for the wind turbines: the steady-state model (which allows the power flow simulations) and the transient model (used for dynamic simulations).

4.3.1.1. Steady-State Model

The load flow provides initial conditions for dynamic simulations. In the Load Flow parameters of the PSS/E, the wind turbines were modelled as five conventional generators, rated at 24 MW and 21 MW. The values specified on the existing generator record are outlined in Table 15, in Appendix A. Note that the values in Table 15 were calculated considering the values in the Individual WTG Power Flow Data of the PSS/E Wind User Guide [26] and multiplying by the number of lumped elements, as recommended by the PSS/E guide.

Since the model used is of a DFIG wind turbine, both the active and reactive power can be controlled. As so, for the wind turbine aggregates, a fixed power factor of 0,9 (injecting reactive power in the grid) was chosen, resulting in a value for Qgen, which is 11,62 Mvar for the 24 MW aggregates and 10,16 Mvar for the 21 MW aggregates.

The values calculated by this manner were used in the HVDC connection alternative. However, for the HVAC transmission system, the wind turbines were regulated for a unit power factor, which means no reactive power is generated by the wind turbines. The reason is that the AC cable already produces a significant amount of reactive power, so an additional quantity of reactive power generated by wind farm would deteriorate the system behaviour.

4.3.1.2. Transient Model

The transient model of the GE 1,5 MW wind turbine comprises nine models for each of the five aggregates. These models are an accurate and complete representation of the behaviour of the GE Wind turbine, including, for example, the simulation of wind gusts and ramps. Table 2 indicates the models used and their description.

41

Table 2: Components of the GE 1,5 MW Wind Turbine Model in PSS/E.

Model Code

Description

GEWTG1

GE Wind Turbine Generator/Converter

GEWTE1

GE Wind Turbine Electrical Control

GEWTT

Two Mass Shaft

WGUSTC

Wind Gust and Ramp

GEWTA

GE Wind Turbine Aerodynamics

GEWTP

GE Pitch Control

GEWTPT

Plotting Output Variables as VARs

VTGTPA

Under Voltage/Over Voltage Generator Bus Disconnection Relay

FRQTPA

Under Frequency/Over Frequency Generator Bus Disconnection Relay

Values for the parameters of these models were based on typical values given in the PSS/E Wind manual example [26]. The number of aggregated wind turbines was changed in the models to 14 for the 21 MW aggregates and 16 for the 24 MW aggregates.

An example of the “.dyr” files used for one of the aggregates, where the parameters of each models are written, is presented in Appendix B.

4.3.1.3. Fault Ride Through Capability

The generator model includes the Under/Over voltage protective functions (modelled by VTGTPA). These functions are defined such that wind plants must not trip for events that are less severe than the defined thresholds and time durations. So, the wind turbines can “ride through” a fault that causes a voltage dip on the terminal voltage of the generator, as long as the voltage dip and respective duration respect the curves in Figure 26.

42

Figure 26 – GE 1,5 MW voltage protection characteristics. 4.3.2. Step-Up Transformers The 0,69/33 kV

Figure 26 – GE 1,5 MW voltage protection characteristics.

4.3.2. Step-Up Transformers

The 0,69/33 kV transformers that adapt the voltage at the generation buses (690 V) to the voltage of the internal grid of the wind farm (33 kV) have the parameters presented in Table 16, in Appendix A.

As each aggregate contains 14 or 16 wind turbines, the rated power of the transformers is, respectively, 24,5 MVA and 28 MVA.

4.3.3. 33 kV Cables

The cables used to link the five buses where the wind turbines are connected to the offshore bus

are 33 kV AC cables, XLPE insulated. The cable parameters are presented in Table 17, in Appendix

A.

Considering the base power = 100 and the base voltage = 33 , the base impedance

yields: = = 10,89 Ω. Therefore, using Equations (9), (10) and (11), the values in Table 17 – and

bearing in mind that = × – the parameters for the medium voltage cables are presented in Table 18, in Appendix A. The lengths of these cables were attributed considering typical distances in existing offshore wind farms, such as the Lillgrund wind farm.

4.3.4. Layout of the offshore wind farm

Figure 27 illustrates the offshore wind farm designed in PSS/E, where the wind turbines, the step- up transformers and the 33 kV sea cables mentioned above are represented.

43

Figure 27 – PSS/E model of the offshore wind farm. 44

Figure 27 – PSS/E model of the offshore wind farm.

44

4.4. HVAC Transmission

The HVAC transmission is the simplest alternative for the transmission of the electrical power of

the offshore wind farm to the onshore grid. The main components are the submarine cable and the

two transformers: onshore and offshore. Figure 28 illustrates the HVAC transmission scheme, with the

AC cable and both transformers. Note that the 102 bus in Figure 28 corresponds to the 102 bus in

Figure 27. For proper compensation of the reactive power generated in the cable, shunt reactors (or a

STATCOM device) were applied to one or both ends of the cable. They are not, however, represented

and described in detail in this chapter, as they vary according to the case simulated.

this chapter, as they vary according to the case simulated. Figure 28 – HVAC transmission layout

Figure 28 – HVAC transmission layout in PSS/E.

4.4.1. AC Submarine Power Cable

The submarine power cable chosen is a 100 km 150 kV XLPE cable. 150 kV is a typical option for

offshore wind farms, used in, for example, the Horns Rev wind farm, in Denmark. Hence, using the

parameters of the submarine cable in Horns Rev, we have the following set of values: =

0,039 Ω⁄ ; = 0,12 Ω⁄ ; = 0,19 / [10].

Considering

the

base

power

= 100

and

the

base

voltage = 150 ,

the

base

impedance is = = 225 Ω. So, using Equations (9), (10) and (11) the 150 kV cable parameters are

depicted in Table 3.

Table 3: 150 kV cable parameters.

Cable

 

 

 

Length

     

From

To

[

/

]

[

/ ]

[

/ ]

[

]

Bus

Bus

       

20

21

 

0,039

 

0,12

 

0,19

 

100

0,0173

0,053

1,34235

45

4.4.2. Onshore and Offshore Transformers

The offshore transformer adapts the voltage from the 33 kV of the internal grid to the 150 kV of

the submarine cable. As for the onshore transformer, it is used to increase the voltage from the 150 kV

of the cable to the 400 kV of the onshore grid. Table 4 shows the offshore and onshore transformers

data.

Table 4: Onshore and Offshore transformer data [10].

Buses

 

Rated

Rated

Voltage

Rating

[

]

From

Bus

To Bus

Location

Voltage

[kV/kV]

Impedance

[%]

[MVA]

20

102

Offshore

33/150

13,8

160

0,08625

21

151

Onshore

150/400

15

200

0,075

46

4.5. HVDC-LCC Transmission

The Line Commutated Converter based HVDC technology requires a DC cable and two

converters, one onshore and another offshore. In PSS/E there are appropriate models for the

simulation of a transmission with HVDC-LCC for both steady-state and transient situations.

4.5.1. Steady-State Model

The steady-state model enables the power flow analysis of the DC Link and establishes the initial

values for dynamic simulations.

In PSS/E, the data requirements for the “Two-terminal DC Line” (as mentioned in [27]) fall into

three groups: control parameters/set points, converter transformers and DC line characteristics. In the

Load Flow Module of PSS/E 30, the two tabs refer to “DC Line” and the respective “Converters”:

Rectifier and Inverter.

Graphically, the DC line is represented by Figure 29, where the converter is represented by a

diode-like symbol. Note that the rectifier and inverter models include the converter transformer,

despite these transformers not being visible.

transformer, despite these transformers not being visible. Figure 29 – HVDC-LCC transmission model in PSS/E. The

Figure 29 – HVDC-LCC transmission model in PSS/E.

The following values were assumed:

= 110

= 150

=

= 18°

= 20°

18°

The current in the DC line is given by: = = 733 A.

The model parameters were calculated considering [27] and [24]. Some of the equations used for

the rectifier and the inverter are expressed in Table 5.

Table 5: Equations for Rectifier and Inverter Parameters.

Parameter

 

Rectifier

 

Inverter

Converter Power Factor

cos( ) = cos( ) + cos ( + )

2

cos( ) = cos( ) + cos ( + )

2

47

Transformer Nominal Power

 

=

cos( )

AC Converter Voltage

 

= ×

3√2

Converter Transformer Ratio

=

=

   

Commutating Reactance

[Ω] = [p. u. ] × , with =

   

3

Commutating Resistance

= ×

Using values calculated with the equations in Table 5, assuming some values recommended in

[27], the parameters used in the load flow model are reported in Table 19 and Table 20, in Appendix

A.

4.5.2. Transient Model

In order to simulate the dynamic behaviour of the HVDC link, the transient model used is the

“CDC4T”. The parameters of this model were calculated using the equations and the recommended

values in the PSS/E Manuals [28] and [29].

The parameters used in the “CDC4T” model are reported in Table 21, in Appendix A.

4.6. STATCOM

PSS/E provides appropriate models to simulate FACTS devices. A variety of such devices is

available for power flow and dynamic simulations, divided in either shunt devices (such is the case of

STATCOM), series devices and combined devices (with both series and shunt elements).

The STATCOM, as described in detail in Chapter 3 is a FACTS device, used in the network to

absorb or provide reactive power, thus providing voltage support to the network, in either steady-state

or transient situation.

In the work developed in this thesis, the STATCOM is used in the HVDC network, to inject

reactive power in the onshore bus, reducing the voltage dip originated by the fault in the AC grid. In

HVAC transmission, the STATCOM also provides this voltage support in case of a fault and,

additionally in steady-state, it is used for reactive power compensation of the AC cable.

ABB’s “PCS 6000 STATCOM double-outdoor” is used as a reference for the modelling of the

STATCOM device. It has a capacity of 2x32 Mvar.

48

4.6.1.

Steady-State Model

For power flow analysis, the STATCOM is modelled as a FACTS device with the parameters adjusted as to simulate the behaviour of this device. In both cases of the use of the STATCOM (in HVAC and HVDC transmission) it was modelled with the same parameters.

Figure 30 is the graphic representation of a STATCOM in PSS/E.

30 is the graphic representation of a STATCOM in PSS/E. Figure 30 – STATCOM device in

Figure 30 – STATCOM device in PSS/E.

The parameters presented in Table 22 (Appendix A) are the most important values used in power flow analysis of the STATCOM and they are obtained using PSS/E Manual [27] for FACTS devices.

4.6.2. Transient Model

The dynamic behaviour of the STATCOM is simulated using the “CSTCNT” model. The parameters of this model were calculated using the equations and the recommended values in the PSS/E Manuals [28] and [29].

The values of the “CSTCNT” model are presented in Table 23, in Appendix A.

49

5. Power Flow Results

Chapter 5 presents the results of the steady-state simulations performed on the grid with the offshore wind farm, as described in Chapter 4. HVAC and HVDC-LCC transmission are analysed. Discussion of the results is also performed.

5.1. Introduction

The power flow analysis carried out comprehends numerical calculations of active and reactive power flows and node voltages. PSS/E software is used, in the scope of this thesis, for the power flow analysis of transmission power alternatives from offshore wind farms. Special attention is given to the Point of Common Coupling (hereby designated as PCC), i.e., the point of connection of the wind farm with the remaining grid. The voltage at the PCC, the active power injected and the reactive power injected/absorbed in the PCC is analysed as a part of the power flow study. Power losses for each transmission system are also assessed.

The power flow results are presented by the single-line diagrams of the network (from the PSS/E load flow software). The active and reactive flows at each end of the branches and the voltage magnitude and angle at each bus are depicted in each figure of this chapter. Figure 31 depicts how each of the values from the power flow calculations are represented in the power flow result figures.

are represented in the power flow result figures. Figure 31 – Legend of the values presented

Figure 31 – Legend of the values presented in the single-line diagrams.

5.2. HVAC Power Flow

For the HVAC transmission system, as presented in Figure 28, the power flow analysis focuses on the reactive power flow in the submarine 100 km cable. Therefore, with the objective of compensating the reactive power generated in the cable, some reactive power compensation options are studied. The alternatives considered are reviewed in Table 6.

50

Table 6: Reactive power compensation alternatives.

Designation

Onshore (Mvar)

Offshore(Mvar)

Offshore shunt reactor

0

60

Onshore shunt reactor

80

0

Onshore and offshore shunt reactors

40

40

Offshore STATCOM

0

64

The main criteria for the choice of the value of the shunt reactors (defined in terms of reactive power injected, in Mvar) is the power factor at the PCC, which is chosen to be approximately of 0,9, a typical value in grid integration of wind farms [11]. As so, the shunt reactor chosen for offshore compensation only absorbs 60 Mvar, as the 0,9 power factor at the PCC is guaranteed with this value.

An estimate of the reactive power produced by the 150 kV AC cable (data on Table 3) can be made taking into account Equation (3). The approximate amount of reactive power produced by the 100 km cable is given by (12):

= × × × = 2 × 50 × 0,19 × 10 × 100

× (150 × 10 ) = 134,3

(12)

So, for the HVAC transmission, approximately 60 % of compensated by the schemes presented in Table 6.

this value of

reactive power is

Note that, for HVAC power flow, the wind turbine generators supply no reactive power, since there is already an excess of reactive power, as a consequence of the shunt capacitance of the AC cable. Therefore, the capability of the DFIG machines of providing voltage support to the grid, by supplying reactive power, is not considered for the present study.

For the first case analysed (offshore compensation), the presented grid includes buses 3004, 201 and 211, closest to the wind farm. For the following power flow results, however, the results relative to these buses are omitted as they are very similar to the results for the offshore compensation only transmission.

51

5.2.1. Offshore Compensation

In Figure 32 the results of the grid connected by an AC cable with compensation offshore are shown. The 60 Mvar shunt reactor (marked in the figure with the orange rectangle) absorbs 64,6 Mvar of reactive power, given the dependency of the reactive power of the voltage at the bus (as expressed by equation (4)).

All the bus voltages near the AC link (buses 102, 20, 21 and 151) are within acceptable values (i.e. below 1,05 p.u.) and the power factor at PCC is approximately 0,88.

p.u.) and the power factor at PCC is approximately 0,88. Figure 32 – Power Flow result

Figure 32 – Power Flow result for HVAC with offshore compensation.

52

5.2.2. Onshore Compensation

The most important remark regarding Figure 33 is the fact that using onshore compensation only leads to overvoltage on the offshore bus. Other values for the onshore reactor where experimented but led to negligible improvements on the magnitude of the voltage of bus 20. The power factor at the PCC is, however, acceptable, since the shunt reactor (marked in orange rectangle on the figure) absorbs a large portion of the reactive power produced at the cable: 83,4 of the 128,3 Mvar produced by the cable.

at the cable: 83,4 of the 128,3 Mvar produced by the cable. Figure 33 – Power

Figure 33 – Power Flow result for HVAC with onshore compensation.

53

5.2.3. Offshore and Onshore Compensation

Figure 34 shows the power flow result for a scheme of compensating reactive power at both ends of the cable by two 40 Mvar reactors. The compensators are marked in orange on the figure. No overvoltages occur in any bus, onshore or offshore.

No overvoltages occur in any bus, onshore or offshore. Figure 34 – Power Flow result for

Figure 34 – Power Flow result for HVAC with offshore and onshore compensation.

54

5.2.4.

STATCOM Compensation

Figure 35 presents the results for the transmission using an offshore STATCOM. The results are very similar to the use of offshore compensation with a shunt compensator, since the amount of reactive power consumed by the STATCOM and the reactor are identical. The STATCOM is shown inside the orange rectangle on the figure.

STATCOM is shown inside the orange rectangle on the figure. Figure 35 – Power Flow result

Figure 35 – Power Flow result for HVAC with STATCOM compensation.

55

5.2.5. Discussion

The compensation alternatives studied bare some similarities but also important differences

between them.

All the schemes were designed with the purpose of obtaining a power factor of approximately 0,9

(injecting reactive power in the grid) at the PCC. Issues like reactive power circulation, bus voltages,

power rating of the cable and power losses are analysed.

Table 7 shows the results obtained for the power factor at the PCC, the voltages at each end of

the cable and the active and reactive power at each end of the cable for the four configurations

studied. The values are taken from the power results presented above.

Table 7: HVAC selected results for different compensation alternatives.

 

Power

V

V

P at the offshore end (MW)

P at the onshore end (MW)

Q at the offshore end(Mvar)

Q at the onshore end(Mvar)

Factor at

onshore

offshore

the PCC

bus (p.u.)

bus (p.u.)

Offshore

0,88

1,023

1,038

110,8

108,8

76,2

60,3

Onshore

0,93

1,021

1,070

110,8

108,3

10,8

128,3

Offshore and

0,94

1,020

1,046

110,8

108,8

55,1

82,1

Onshore

STATCOM

0,88

1,023

1,037

110,8

108,8

77,9

58,4

offshore

Compensating onshore only is the option that leads to less favourable results. The excess of

reactive power at the onshore bus leads to an excessive voltage magnitude at bus 20, as seen in

Figure 33. The influence of bus voltage amplitude on the transmission of reactive power is thus

demonstrated by the results. Another important issue to consider is the power rating of the AC cable.

In the onshore compensation scheme the power transmitted by the cable reaches its peak near the

onshore bus. This may lead to the overloading of the cable, a direct consequence of the significantly

high amount of reactive power transmitted through the cable.

The other compensation alternatives present more satisfactory results. As far as the power rating

of the cable is concerned, compensating offshore only leads to a more evenly distributed loading of

the cable (as presented in Figure 16) as similar amounts of reactive power flow to each end of the

cable. The bus voltages at each end of the cable (bus 20 offshore and bus 21 onshore) are below 1,05

p.u. (as can be seen in Table 7) so no overvoltage exists, since the excess of reactive power at the

offshore bus no longer exists. Therefore, compensation offshore is proven to be necessary, in order to

maintain the stability of the bus voltages. It is, however, important to notice that the installation of

compensators offshore is likely to be more expensive and technically challenging than the placement

of such devices at the onshore substation.

The STATCOM compensation presents very similar results to the compensation offshore using a

shunt reactor. This option was introduced since the STATCOM device may offer benefits for the

system in its dynamic behaviour, as approached in Chapter 6.

56

The power losses of the system are the same for all compensation schemes, since the devices used for compensation are considered as being lossless. The active power generated by the wind farm is 111 MW and the active power at the PCC is 108,8 MW. Therefore, the losses for the HVAC transmission can be calculated by:

=

111 − 108,8

=

111

= 1,98%

In Table 7, it is obvious that for any of the cases studied the power factor achieved is close to the 0,9 power factor, set as an objective of the compensation. Without compensation of any sort, the reactive power injected in the grid would be very high, leading to a much deteriorated power factor at the PCC. The bus voltages would also be above 1,05 p.u., due to the excess of reactive power.

The solution considered to be the most effective and cost-efficient, given the smaller size of the compensators is the offshore and onshore compensation. This configuration is considered for the dynamic simulations in Chapter 6.

57

5.3. HVDC Power Flow

The HVDC power flow analysis is carried out for two configurations: one without STATCOM and another with a STATCOM at the onshore bus (bus 151). Note that for the studies with HVDC transmission the wind turbines are working with a 0,9 power factor (injecting reactive power in the grid). The objective is once again to guarantee an approximate 0,9 power factor at the PCC.

The active and reactive power flows and the voltages at the buses at each end of the link are compared for these two HVDC transmission options and also taking into account the HVAC transmission power flow results.

5.3.1. HVDC Configuration

Figure 36 depicts the single-line diagram with the power flow results for the HVDC transmission.

with the power flow results for the HVDC transmission. Figure 36 – Power Flow result for

Figure 36 – Power Flow result for HVDC-LCC transmission.

58

5.3.2.

HVDC+STATCOM Configuration

Figure 37 shows the power flow results for the HVDC+STATCOM alternative. The STATCOM can be seen at bus 151, injecting reactive power in steady-state. The introduction of the STATCOM in the system influences the power flows at the nearby lines.

the system influences the power flows at the nearby lines. Figure 37 – Power Flow result

Figure 37 – Power Flow result for HVDC+STATCOM transmission.

59

5.3.3. Discussion

Comparing with the HVAC transmission, the HVDC configuration holds no reactive power issues,

since these issues yield from the shunt capacitance of the AC cable, inexistent in HVDC.

Table 8 presents some results for the two HVDC configurations studied, including power factor at

the PCC, voltage magnitudes on both ends of the transmission scheme and active and reactive

powers circulating on both ends of the submarine cable.

Table 8: HVDC-LCC and HVDC-LCC+STATCOM selected results.

 

Power

V

V

P at the offshore end (MW)

P at the onshore end (MW)

Q at the offshore end(Mvar)

Q at the onshore end(Mvar)

Factor at

onshore

offshore

the PCC

bus

bus

(p.u.)

(p.u.)

       

HVDC-LCC

0,89

1,012

0,978

110

109

44

54

HVDC-

             

LCC+STATCOM

0,89

1,016

0,978

110

109

44

55

From Table 8 it can be noted that the two configurations bare similarities. Offshore, as the wind

farm is the same, the voltage magnitude and the active and reactive powers transmitted in the cable

are equal. Onshore, the influence of the STATCOM – although slight – can be noted, as the voltage is

higher, due to the injection of reactive power by the STATCOM.

As expected, the HVDC-LCC converters consume reactive power, which is a consequence of

having lagging current. Therefore, as can be seen in Table 8, the rectifier (located offshore) consumes

44 Mvar and the inverter (onshore) consumes 54 Mvar and 55 Mvar, for the HVDC and for the HVDC-

STATCOM transmission, respectively. As so, the power factor at the PCC is, for both cases, 0,89

(consuming reactive power).

The use of a STATCOM is justified in steady-state for the injection of reactive power in bus 151,

in order to compensate the reactive power absorbed by the inverter. So, in the power flow analysis,

the STATCOM is working as a capacitor bank, injecting 65 Mvar – its maximum value for the bus

voltage at bus 151. In this case, the use of the STATCOM changes the reactive power that flows to

and from the lines connected to bus 151, because of the behaviour of the LCC converter (consuming

reactive power). For the purpose of this study, in steady-state, capacitor banks could have also been

used but the STATCOM is chosen as to also provide support to the grid in transient situations.

The power losses in the HVDC transmission are very similar to the ones in HVAC (around 2%),

as the generated power is the same (111 MW) and the power at the PCC is similar (109 MW). So the

power losses are given by:

111 − 109

=

111

= 1,8%

In HVDC, the losses are originated in the DC line (characterized by a resistance). The line

reactance and capacitance concepts do not exist and so neither does reactive power in the link.

60

6. Dynamic Results

Chapter 6 presents the results of the dynamic simulations performed on the grid under analysis. The response of the HVAC and HVDC-LCC transmission systems to an AC grid fault is analysed and results are discussed.

6.1. Introduction

The analysis of the dynamic behaviour of the offshore wind farm comprehends the response of the wind farm to voltage and frequency disturbances in the grid.

The study of this work is focused on the response of both the offshore wind farm and the onshore grid to a three-phase fault in an onshore bus, bus 3005, as marked in Figure 38.

fault in an onshore bus, bus 3005, as marked in Figure 38. Figure 38 – Single

Figure 38 – Single line diagram of the grid used for the dynamic simulations. Note: Bus 3005, where the fault occurs is marked in the orange rectangle.

61

The behaviour of some parameters of the grid, namely the voltage and the frequency in each bus,

during and immediately following the disturbance are analysed for the two transmission systems under

analysis: HVAC and HVDC-LCC.

The parameters of the grid that are presented in the results, in 6.2 and 6.3, are the frequency and

voltage variation in the offshore and onshore buses and the speed variation of the offshore and

onshore machines. For the HVDC transmission the active and reactive power in the rectifier and

inverter are also presented.

The analysis is carried out for the two transmission alternatives and for different fault types, with

the objective of assessing the fault ride through capability of the offshore wind farm, i.e. the

requirement for the wind farm to stay connected to the grid during the disturbance, thus contributing to

the reestablishment of the normal operation. The fault ride through capability of the wind turbines in

the offshore wind farm is guaranteed by the under/over voltage disconnection relays of the wind

turbine generators. These devices allow the operation of the wind turbines even when the terminal

voltage decreases. This capacity of “riding through” a fault is limited to defined voltage dips and fault

durations. So, and according to the implemented fault ride through characteristic (as defined in Figure

26), the wind turbine will only trip if the fault that occurs across the terminals of the machine are

outside the defined limits.

6.2. HVAC Dynamic Behaviour

In order to evaluate the behaviour of the offshore wind farm connected to the onshore by an

HVAC link, different dynamic cases are analysed, as presented in Table 9.

Table 9: HVAC Dynamic Cases Analysed.

 

Fault Duration

 

Reactive Compensation Scheme

Case

(ms)

Fault Severity

Case 1

100

Moderate

Shunt reactors at both ends

Case 2

300

Moderate

Shunt reactors at both ends

Case 3

100

Severe

Shunt reactors at both ends

Case 4