WETF-wind Power PDF

VOLUME 1 - TECHNOLOGY
INTRODUCTION TO VOLUME 1 3
CHAPTER 1 TURBINE TECHNOLOGY 7
1.1 Evolution of Commercial Wind Technology 7

1.1.1 Achievements 7
1.1.2 The Challenge 8
1.1.3 A Unique Technology 8
1.1.4 Run up to Commercial Technology 9
1.1.5 Design Styles 10
1.1.6 Design Drivers for Modern Technology 13
1.1.7 Growth of Wind Turbine Size 14
1.1.8 Architecture of a Modern Wind Turbine 16
1.1.9 Erection of Large Wind Turbines 17
1.2 Technology Trends 19

1.2.1 Larger Diameters 19
1.2.2 Tip Speed - Offshore and Land Based Designs 20
1.2.3 Pitch versus Stall 21
1.2.4 Speed Variation 21
1.2.5 Hub Height 22
1.2.6 Rotor Mass 22
1.3 Recent Developments 23

1.3.1 Direct Drive Generators 23
1.3.2 Hybrid – Single Stage of Gears and Multi-Pole Generator 25
1.3.3 Rotor Blade Developments 26
1.3.4 Single Bearing Arrangement 28
1.3.5 Offshore Technology 28
1.4 Technology Status 32

1.4.1 Overall Design Trends 32
WIND ENERGY - THE FACTS - TABLE OF CONTENTS

1.4.2 Size Limitations 33
1.4.3 The Success of Wind Technology 36
1.5 Concluding Remarks and Future R&D 37
CHAPTER 2 WIND RESOURCE ESTIMATION 38
2.1 Introduction 38
2.2 Regional Wind Resources 38
2.3 Wind Atlases 39

2.3.1 Onshore 39
2.3.2 Offshore 41
2.4 Energy Estimates 41

2.4.1 Onshore 41
2.4.2 Offshore 44 XIX
2.4.3 Updating Resource Potential 44
2.4.4 Concluding Remarks 45
2.5 Local Wind Resource Assessment and Energy Analysis 45
2.5.1 Introduction 45
2.5.2 The Importance of the Wind Resource 47
2.5.3 Best Practice for Accurate Wind Speed Measurements 47
2.5.4 The Annual Variability of Wind Speed 49
2.5.5 Analytical Methods for the Prediction of the Long Term Wind Regime at a Site 53
2.5.6 The Prediction of the Energy Production of a Wind Farm 55
2.5.7 Definition of Uncertainty in Predicted Energy Production 57
2.6 Offshore Wind Farm Design and Resource Estimation 58

2.6.1 Fundamentals 58
2.6.2 Measurement Offshore 59
2.6.3 Wind Analysis Offshore 59
2.6.4 Energy Prediction 59
2.6.5 Other Effects to Consider Offshore 59
2.7 Forecasting 60
2.7.1 Overview of the Method 61
2.7.2 Improvement over Persistence 62
2.7.3 Power Output 62
2.7.4 Portfolio Effects 63
2.7.5 Conclusions 63
2.8 Future Developments 63
CHAPTER 3 WIND FARM DESIGN 65
3.1 Introduction 65
3.2 Preliminary Layout Design 65
3.3 Detailed Layout Design 65

3.4 The Infrastructure 66

3.4.1. Civil Works 66
3.4.2. Electrical Works 67
3.4.3. SCADA and Instruments 67
3.5 Construction 67
3.6 Costs 68
3.7 Commissioning and Operation 68
3.8 Concluding Remarks 69
CHAPTER 4 TRANSMISSION AND DISTRIBUTION NETWORKS 70
4.1 Introduction 70
4.2 Setting the Scene 70

XX 4.2.1 Large Interconnected Networks 70
4.2.2 Small Isolated Networks 71
4.3 Electricity Networks 72
4.4 Considerations for Wind Energy 74

4.4.1 Connections 74
4.4.2 Operation 75
4.4.3 Strategic Planning Considerations 77
4.5 Issues for High Wind Penetration in Europe 80
CHAPTER 5 RESEARCH AND DEVELOPMENT 84
5.1 Wind Industry Research and Development Overview 84
5.2 Socio-Economic, Policy and Market Issues 85

5.2.1 Transparency 85
5.2.2 Increasing the Value of Wind Power 86
5.2.3 Education and Human Resources Development 86
5.3 Environmental and Social Impacts 86
5.4 Turbine and Component Design Issues 87
5.5 Testing, Standardisation, Certification and Safety 88
5.6 Grid Integration, Energy Systems, and Resource Prediction 88
5.7 Operation and Maintenance 89
5.8 Location of Wind Farms 90
5.9 Offshore Wind 90
5.10 Multi-Megawatt Turbines 91

5.11 Summary of R&D Objectives 91
XXI
VOLUME 2 - COSTS & PRICES
CHAPTER 1 INTRODUCTION 96
CHAPTER 2 COST AND INVESTMENT STRUCTURES 97
CHAPTER 3 OPERATION AND MAINTENANCE COSTS OF WIND POWER 100
CHAPTER 4 THE COST OF ENERGY GENERATED BY WIND POWER 103
CHAPTER 5 DEVELOPMENT OF THE COST OF WIND POWER 105
CHAPTER 6 FUTURE DEVELOPMENT OF THE COST OF WIND POWER 106
CHAPTER 7 COSTS OF CONVENTIONAL POWER PRODUCTION 108
CHAPTER 8 EXTERNAL COSTS OF POWER PRODUCTION 110

XXII
VOLUME 3 - INDUSTRY & EMPLOYMENT
CHAPTER 1 INDUSTRY STATUS 114
1.1 Introduction 114

1.1.1 New Developments 114
1.1.2 Market Changes 114
1.2 Demand Drivers 115
1.2.1 Market Types 115
1.2.2 Political Risk 116
1.3 The Onshore Market 116
1.3.1 Market Status (Cumulative Market) 116
1.3.2 Market Growth (Annual Market) 119
1.4 Offshore Market 121
1.5 Domestic Market 123
1.6 Export Market 124
1.7 Industry Segments 125

1.7.1 Manufacturers - Megawatt Scale 125
1.7.2 Manufacturers - Small Turbines 125
1.8 Wind Farm Developers 126
CHAPTER 2 EMPLOYMENT IN THE WIND TURBINE SECTOR 127
2.1 Direct and Indirect Employment Effect 127

2.2 Direct Employment in Wind Turbine Manufacturing 127
2.3 Direct Employment in Wind Turbine Installation 128
2.4 Direct Employment in Maintenance Activities 129
2.5 Indirect Employment in Manufacturing 130

2.6 Indirect Employment in Wind Turbine Installation 130
2.7 Total Direct and Indirect Employment (Manufacturing, Installation and Maintenance in Europe) 131
CHAPTER 3 EMPLOYMENT PREDICTION AND METHODOLOGY 133
3.1 Direct and Indirect Employment 133

3.2 Input-Output Methodology 133
3.3 Projection/Prediction Parameters 135
3.4 Sensitivity – Main Parameters 135
CHAPTER 4 SCENARIOS FOR EMPLOYMENT IN THE WIND TURBINE SECTOR 137
4.1 Projection of Employment Based on Wind Energy Installation in Europe and Globally 137
CHAPTER 5 DEVELOPMENT AND INNOVATION 139
XXIII
VOLUME 4 - ENVIRONMENT
CHAPTER 1 EXTERNALITIES 147

1.1 Introduction to Externalities 147
1.2 Definition and Classification 147
1.3 Importance of Externalities 147
1.4 Externalities and Electricity Production 149
1.5 Impacts of Wind Energy and Other Technologies. 152
1.6 Externalities of Wind Energy 154
1.7 Benefits of Wind Energy 155
1.8 Present State of Knowledge 156
CHAPTER 2 ENVIRONMENTAL BENEFITS OF WIND ENERGY 158

2.1 Background 158
2.2 Electricity Generation and Emissions in EU-25 and other European Countries 159
2.2.1 Electricity Generation Sector and Environmental Policy Framework 160
2.2.2 Emission Data 163
2.3 The Calculation of External Costs with the EcoSense Model 165
2.3.1 Software Description 165
2.3.2 Input Data to the Model 166
2.4 Benefits of Wind Energy - Results 167

2.4.1 Avoidable Emissions by the Use of Wind Energy 167
2.4.2 Avoidable External Costs by the Use of Wind Energy 168
CHAPTER 3 STANDARD METHODOLOGY FOR CALCULATION OF EMISSION REDUCTIONS 172

CHAPTER 4 ANALYSIS OF EMISSION REDUCTIONS 173

4.1 Avoidable Specific Emissions Through Wind Energy 173
4.2 Avoidable Total Emissions Through Wind Energy 174
4.3 Avoidable External Costs Through Wind Energy 177
CHAPTER 5 PUBLIC ACCEPTANCE ANALYSIS 179

5.1 Environmental Impacts of Wind Energy 179
5.1.1 Visual Impact 179
5.1.2 Noise 180
5.1.3 Land Use 182
5.1.4 Impact on Birds 182
5.1.5 Impacts of Construction on Terrestrial Ecosystems 184
XXIV 5.1.6 Electromagnetic Interference (EMI) 184
5.1.7 Flickering 185
5.1.8 Consumption of Energy (Energy Balance) 185
5.2 Environmental Impacts of Offshore Wind Energy 186

5.3 Factors Affecting Public Acceptance of Wind Energy 188
CHAPTER 6 PUBLIC ACCEPTANCE IN THE EU 190

6.1 Attitudes of EU Citizens to Energy and Energy Technology Issues 190
6.2 Public Acceptance in Spain 192
6.3 Public Acceptance in the United Kingdom 194
6.4 Public Acceptance in Denmark 197
6.5 Public Acceptance in Germany 197
6.6 Public Acceptance in Sweden 198
6.7 Public Acceptance in Austria 198
6.8 Public Acceptance in Belgium 199
6.9 Conclusions 200
XXV
VOLUME 5 - MARKET DEVELOPMENT
CHAPTER 1 MARKET INCENTIVES 205
1.2 Environmental Taxes 208
1.3 Payment Mechanisms 209

1.3.1 Voluntary Systems and Green Marketing 210
1.3.2 Fixed Price Systems 213
1.3.3 Investment Subsidies 213
1.3.4 Fixed Feed-in Tariffs 214
1.3.5 Fixed Premium Systems 215
1.3.6 Tax Credits 216
1.4 Fixed Quantity Systems 216

1.4.1 Tendering Systems 217
1.4.2 Tradable Green Certificates 219
1.5 Renewables in the New Member States 225
1.6 The EU Legal and Political Framework 225

1.6.1 The Electricity Directive 226
1.6.2 The Renewables Directive 226
CHAPTER 2 FUTURE MARKETS 230
2.2 Conventional Scenario 230

2.2.1 Conventional Scenario Results 231
2.3 Other Forecasts 232

2.4 Wind Force 12 - the Advanced Scenario 232

2.4.1 Methodology 232
2.4.2 Regional Update of Wind Force 12 Advanced Scenario 233
2.5 Overview of Non European Markets 234

2.5.1 North America 234
2.5.2 Central and South America 234
2.5.3 Australia and New Zealand 235
2.5.4 Asia 236
2.5.5 Africa 237
2.5.6 The East 238
XXVI
CHAPTER 3 NEW TARGETS 239
3.1 EWEA Targets – Onshore and Offshore 239

3.1.1 European Commission – Historical Targets 240
3.1.2 European Commission – New Targets 240
3.2 Increasing Wind Power Targets for Europe 241
3.3 Targets for the EU-15 in 2010 243

3.3.1 How Much Electricity Will This Provide? 244
3.3.2 What Proportion of Total EU Electricity from Wind? 245
3.3.3 What Share Will Wind Have of Total New Capacity Installed? 245
3.4 International Energy Agency Scenarios 247

3.4.1 OECD Alternative Policy Scenario 247
XXVII
APPENDICES
A ONSHORE WIND MAPS 253
B OFFSHORE WIND MAPS 262
C WORKED EXAMPLE FOR CUILLIAGH MOUNTAIN WIND FARM, IRELAND 268
D DETAILED DESCRIPTION OF CORRELATION TECHNIQUES 281
E CONNECTION CONSIDERATIONS 283
F POSSIBLE GRID CODE REQUIREMENTS 287
G CALCULATION OF SPECIFIC EMISSIONS OF STANDARD AIR POLLUTANTS FROM

FOSSIL FUEL ELECTRICITY GENERATION (METHODOLOGY) 290
H ASSUMPTIONS FOR THE CALCULATIONS OF EXTERNAL COSTS WITH ECOSENSE 292
I MARGINAL COST OF CO2 EMISSIONS 293
J CALCULATED EMPLOYMENT IN EU LEVEL SECTORS 294
K EASTERN PROMISE, WIND DIRECTIONS, EWEA, MARCH 2003 295
L EWEA MEMBERS 301
M DIRECTIVE 2001/77/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

OF 27 SEPTEMBER 2001 ON THE PROMOTION OF ELECTRICITY PRODUCED FROM
RENEWABLE ENERGY SOURCES IN THE INTERNAL ELECTRICITY MARKET 305
GLOSSARY
313
REFERENCES 325
XXVIII
WIND ENERGY - THE FACTS
VOLUME 1
TECHNOLOGY
HOOFDTITEL
Acknowledgments
This volume was compiled by Paul Gardner, Andrew

Garrad, Peter Jamieson, Helen Snodin and Andrew Tindal
of Garrad Hassan and Partners, United Kingdom. Special
thanks also to Airtricity for their permission to case study
Cuilliagh Wind Farm. Our thanks also to the national wind
associations around Europe for their contributions of
data, and to the other project partners for their inputs.
WIND ENERGY - THE FACTS - TECHNOLOGY
2
INTRODUCTION TO VOLUME 1 - TECHNOLOGY
1
VOLUME
Electricity can be generated in many ways. In each case, short-term behaviour. The latter can be successfully fore-
a fuel is used to turn a turbine which drives a generator cast to allow wind energy to participate in electricity mar-
which feeds the grid. The turbines are designed to suit the kets.
particular characteristics of the fuel. Wind generated elec-
tricity is no different. The wind is the fuel – unlike fossil In order for wind to live up to its raw potential promise,
fuels it is both free and clean, but otherwise it is just the individual turbines must be assembled into wind farms or
same. It drives the turbine which generates electricity into wind power stations. The design and operation of the dif-
a grid. ferent types of wind farms are discussed and examples
provided.
The politics and economics of wind energy have played an
important role in the development of the industry and con- Finally, the key strategic issue for the future is addressed:
tributed to its present success. Engineering is, however, How can the windy parts of Europe, both onshore and off-
pivotal. As the wind industry has become better estab- shore, deliver power for the industrial loads and popula-
lished, the central place of engineering has become over- tion centres. This goal is achieved through the local,
shadowed by other issues. This is a tribute to the success national and international grids. The way in which the grid
of the engineers and their turbines. This volume address- is used and constrained is a key political and technical
es the key engineering issues: issue. The technical and regulatory constraints are
• The turbines – their past achievements and future chal- described and some challenges for the future are raised.
lenges – a remarkable tale of technical endeavour and
entrepreneurship. This volume explores how this new, vibrant and rapidly
• The wind – its characteristics and reliability – how can expanding industry exploits one of nature’s most copious
it be measured, quantified and harnessed? sources of energy – the wind.
• The wind farms – an assembly of individual turbines
into wind power stations or wind farms – their optimi-
sation and development.
• The grid – transporting the energy from remote loca-
tions with plentiful wind energy to the loads – the key
technical and strategic challenges.
This volume provides an historical overview of turbine

development, describes the present status and considers
future challenges. This is a remarkable story starting in
the nineteenth century and then accelerating through the
last two decades of the twentieth century on a course very
similar to the early days of aeronautics. The story is far
from finished but it has certainly started with a
vengeance.
Wind must be treated with great respect. The speed of the

wind on a site has a very powerful effect on the econom-
ics of a wind farm; it provides both the fuel to generate
electricity and the loads to destroy the turbine. This vol-
ume describes how it can be quantified, harnessed and
3
put to work in an economic and predictable manner. The
long-term behaviour of the wind is described as well as its
VOLUME 1 - TECHNOLOGY: TABLE OF CONTENTS
CHAPTER 1 TURBINE TECHNOLOGY 7
1.1 Evolution of Commercial Wind Technology 7

1.1.1 Achievements 7
1.1.2 The Challenge 8
1.1.3 A Unique Technology 8
1.1.4 Run up to Commercial Technology 9
1.1.5 Design Styles 10
1.1.6 Design Drivers for Modern Technology 13
1.1.7 Growth of Wind Turbine Size 14
1.1.8 Architecture of a Modern Wind Turbine 16
1.1.9 Erection of Large Wind Turbines 17
1.2 Technology Trends 19

1.2.1 Larger Diameters 19
1.2.2 Tip Speed - Offshore and Land Based Designs 20
1.2.3 Pitch versus Stall 21
1.2.4 Speed Variation 21
1.2.5 Hub Height 22
1.2.6 Rotor Mass 22
1.3 Recent Developments 23

1.3.1 Direct Drive Generators 23
1.3.2 Hybrid – Single Stage of Gears and Multi-Pole Generator 25
1.3.3 Rotor Blade Developments 26
1.3.4 Single Bearing Arrangement 28
1.3.5 Offshore Technology 28
1.4 Technology Status 32

1.4.1 Overall Design Trends 32
1.4.2 Size Limitations 33
1.4.3 The Success of Wind Technology 36
1.5 Concluding Remarks and Future R&D 37
CHAPTER 2 WIND RESOURCE ESTIMATION 38
2.1 Introduction 38
2.2 Regional Wind Resources 38
2.3 Wind Atlases 39

2.3.1 Onshore 39
2.3.2 Offshore 41
2.4 Energy Estimates 41

2.4.1 Onshore 41
4 2.4.2 Offshore 44
2.4.3 Updating Resource Potential 44
2.4.4 Concluding Remarks 45
1
VOLUME
2.5 Local Wind Resource Assessment and Energy Analysis 45
2.5.1 Introduction 45
2.5.2 The Importance of the Wind Resource 47
2.5.3 Best Practice for Accurate Wind Speed Measurements 47
2.5.4 The Annual Variability of Wind Speed 49
2.5.5 Analytical Methods for the Prediction of the Long Term Wind Regime at a Site 53
2.5.6 The Prediction of the Energy Production of a Wind Farm 55
2.5.7 Definition of Uncertainty in Predicted Energy Production 57
2.6 Offshore Wind Farm Design and Resource Estimation 58

2.6.1 Fundamentals 58
2.6.2 Measurement Offshore 59
2.6.3 Wind Analysis Offshore 59
2.6.4 Energy Prediction 59
2.6.5 Other Effects to Consider Offshore 59
2.7 Forecasting 60
2.7.1 Overview of the Method 61
2.7.2 Improvement over Persistence 62
2.7.3 Power Output 62
2.7.4 Portfolio Effects 63
2.7.5 Conclusions 63
2.8 Future Developments 63
CHAPTER 3 WIND FARM DESIGN 65
3.1 Introduction 65
3.2 Preliminary Layout Design 65
3.3 Detailed Layout Design 65
3.4 The Infrastructure 66

3.4.1. Civil Works 66
3.4.2. Electrical Works 67
3.4.3. SCADA and Instruments 67
3.5 Construction 67
3.6 Costs 68
3.7 Commissioning and Operation 68
CHAPTER 4 TRANSMISSION AND DISTRIBUTION NETWORKS 70
4.1 Introduction 70
4.2 Setting the Scene 70

4.2.1 Large Interconnected Networks 70 5
4.2.2 Small Isolated Networks 71
4.3 Electricity Networks 72
4.4 Considerations for Wind Energy 74

4.4.1 Connections 74
4.4.2 Operation 75
4.4.3 Strategic Planning Considerations 77
4.5 Issues for High Wind Penetration in Europe 80
CHAPTER 5 RESEARCH AND DEVELOPMENT 84
5.1 Wind Industry Research and Development Overview 84
5.2 Socio-Economic, Policy and Market Issues 85

5.2.1 Transparency 85
5.2.2 Increasing the Value of Wind Power 86
5.2.3 Education and Human Resources Development 86
5.3 Environmental and Social Impacts 86
5.4 Turbine and Component Design Issues 87
5.5 Testing, Standardisation, Certification and Safety 88
5.6 Grid Integration, Energy Systems, and Resource Prediction 88
5.7 Operation and Maintenance 89
5.8 Location of Wind Farms 90
5.9 Offshore Wind 90
5.10 Multi-Megawatt Turbines 91
5.11 Summary of R&D Objectives 91

6
1 TURBINE TECHNOLOGY
1
VOLUME
1.1 Evolution of Commercial Wind Initially, none of these issues could be dismissed lightly,
Technology but gradually all have been addressed.
The extent of Europe’s wind resource will be established in • In larger projects with proven medium sized turbines,
the following chapter. The engineering challenge to the availability of 98% is consistently achieved. The latest
wind industry is to harness that energy and turn it into elec- large machines are also approaching that level of
tricity – to design an efficient wind turbine (WT). In this availability.
chapter the evolution of WT technology is discussed, its • Some of the early turbines were noisy – both aerody-
present status described, and future challenges identified. namically and mechanically – and noise was a problem.
Today, mechanical noise is practically eliminated and
The evolution of modern turbines is a remarkable story of aerodynamic noise has been vastly reduced.
engineering and scientific skill, coupled with a strong entre- • WTs are now highly efficient with less than 10% ther-
preneurial spirit. In the last 20 years turbines have mal losses in the system transmission. The aerody-
increased in power by a factor of 100, the cost of energy namic efficiency of turbines has gradually risen from
has reduced, and the industry has moved from an idealistic the early 1980s with the coefficient of performance
fringe activity to the edge of conventional power generation. rising from 0.44 to about 0.50 for state-of-the-art tech-
At the same time, the engineering base and computational nology. The value of 0.5 is near to the practical limit
tools have developed to match machine size and volume. dictated by the drag of aerofoils and compares with a
theoretical limit of 0.59 (known as the Betz limit).
This is a remarkable story, and it is far from finished. • It was often suggested that there would be major prob-
Many technical challenges remain and even more spec- lems of grid stability with penetrations of wind energy
tacular achievements will result. Serious investment is above 10%. Now, a much more complex picture has
needed to maximise potential through R&D. emerged. Benefits of capacity credits, local reinforce-
ment of grids and the ability of variable speed turbines
The use of technical jargon in this section has been kept to to contribute to grid stability counteract concerns about
a minimum but technical terms inevitably arise. These are variability of supply, mismatch with demand and the
explained in the glossary provided at the end of the book. need for storage in the electrical system. In typical grid
systems there may be an adverse economic impact for
1.1.1 ACHIEVEMENTS penetration levels above 20%, but there is no overrid-
ing technical difficulty that would limit wind energy
Modern commercial wind energy started in earnest in the penetration to very low values.
early 1980s following the oil crises of the 1970s when • Visual and environmental impacts require sensitive
issues of security and diversity of energy supply and, to a treatment but, Europe-wide, public reaction to opera-
lesser extent, long-term sustainability, generated interest tional wind farms is generally positive.
in renewable energy sources. • A dismissive view of the possibility of nationally signif-
icant wind energy contributions was prevalent in the
However, wind power sceptics raised questions about: 1980s. With penetration levels of over 17% in
• reliability Denmark, and around 5% in both Germany and Spain,
• noise this view is belied. Moreover, growth of the offshore
• efficiency market, a resource large enough to supply all of
• grid impact Europe’s electricity, will further reinforce the signifi-
• visual and general environmental impact cance of wind energy in the European energy supply.
• potential for serious contribution to a national • Costs of turbines per unit capacity have reduced greatly
7
energy supply since the 1980s. This cost reduction has been
• cost achieved through both technical improvements and
also through volume. Wind energy is now sometimes Figure 1.1: Traditional “Dutch” Windmill
commercially competitive with new coal or gas power
plant on good, windy sites.
1.1.2 THE CHALLENGE
The concept of a wind driven rotor is ancient, and electric

motors were in profusion domestically and commercially in the
latter half of the twentieth century. Making a WT can seem
simple but it is a big challenge to produce a turbine that:
• Meets specifications (frequency, voltage, harmonic
content) for standard electricity generation with
each unit operating as an unattended power station.
• Copes with wind variability (mean wind speeds on
exploitable sites range from 5 m/s to 11 m/s, with
severe turbulence in the Earth’s boundary layer
and extreme gusts up to 70 m/s).
• Competes economically with other energy sources.
The traditional “Dutch” windmill (Figure 1.1) had prolifer-

ated to a peak of around 100,000 machines throughout
Europe by the late nineteenth century. These machines
preceded electricity supply and were indeed “windmills”
used for grinding grain, for example. They were always
attended, perhaps inhabited and, largely, manually con- 1.1.3 A UNIQUE TECHNOLOGY
trolled. They were integrated within the community,
designed for frequent replacement of certain components Stall
and efficiency was of little importance.
WTs have little respect for engineering conventions.
In contrast, the function of a modern power-generating Most aerodynamic devices (aeroplanes, gas turbines,
WT is to generate high quality, network frequency elec- etc.) avoid stall. Stall, from a functional standpoint, is
tricity. Each turbine must function as an automatically the breakdown of the normally powerful lifting force
controlled independent “mini power station”. It is when the angle of flow over an aerofoil (such as a wing
unthinkable for a modern WT to be permanently attend- section) becomes too steep. This is a potentially fatal
ed, and uneconomic for it to need much maintenance. event for a flying machine, whereas WTs can make
The development of the microprocessor has played a cru- purposeful use of stall as a means of limiting power in
cial role in realising this situation, thus enabling cost- high wind speeds.
effective wind technology. A modern WT is required to
work unattended, with low maintenance, continuously for For a further discussion of stall, see sections 1.1.5. and
more than 20 years. 1.2.3.
The design requirements of stall regulation have led to

new aerofoil developments and also the use of stall
8
strips, vortex generators, fences, Gurney flaps and other
devices for fine-tuning rotor blade performance.
1
VOLUME
Fatigue 1970 - 1990
The power train components of a WT are subject to high- In the early 1980s, many issues of rotor blade technology
ly irregular loading input from turbulent wind conditions, were investigated. Steel rotors were tried but rejected as
and the number of fatigue cycles experienced by the too heavy, aluminium was deemed too uncertain in the con-
major structural components can be orders of magnitude text of fatigue endurance, and the wood-epoxy system
greater than for other rotating machines. Consider that a developed by the Gougeon brothers in the US was
modern WT operates for about 13 years in a design life employed in a number of small and large turbines. The
of 20 and is almost always unattended. A motor vehicle, blade manufacturing industry has, however, been dominat-
by comparison, is manned, frequently maintained and its ed by fibreglass polyester construction which evolved from
design life of about 150,000 kilometres is equivalent to a boat building background and became thoroughly consol-
just four months of continuous operation. idated in Denmark in the 1980s.
Thus, in the use rather than avoidance of stall and in the By 1980 in the US, a combination of state and federal,
severity of the fatigue environment, wind technology has a energy and investment tax credits had stimulated a rapidly
unique technical identity and R&D demands. expanding market for wind in California. Over the period
1980-95 about 1,700 MW of wind capacity was installed,
1.1.4 RUN UP TO COMMERCIAL TECHNOLOGY more than half after 1985 when the tax credits had reduced
to about 15%.
An early attempt at large-scale commercial generation of
power from wind was the 53 m diameter, 1.25 MW Smith Tax credits attracted an indiscriminate overpopulation of
Putnam WT erected at Grandpa’s Knob in Vermont, USA various areas of California (San Gorgonio, Tehachapi and
in 1939. This design brought together some of the finest Altamont Pass) with many ill-designed WTs which func-
scientists and engineers of the time (aerodynamic design tioned poorly. However, the tax credits created a major
by von Karman, dynamic analysis by den Hartog). The tur- export market for European, especially Danish, WT manu-
bine operated successfully for longer than some multi-MW facturers who had relatively cost-effective, tried and tested
machines of the 1980s. hardware available. The technically successful operation of
the later, better designed WTs in California did much to
It was a landmark in technological development and pro- establish the foundation on which the modern wind indus-
vided valuable information about quality input to design, try is built. The former, poor quality, turbines conversely
machine dynamics, fatigue, siting sensitivity, etc. created a poor image for the industry which it has taken a
long time to shake off.
The next milestone in WT development was the Gedser
turbine. With assistance from Marshall Plan post-war 1990 - Present
funding, a 200 kW, 24 m diameter WT was installed dur-
ing 1956-57 on the island of Gedser in the south-east of The growth of wind energy in California was not sustained,
Denmark. This machine operated from 1958 to 1967 with but there was striking development in European markets
a capacity factor of around 20%. with an installation rate in Germany of around 200 MW per
annum in the early 1990s. From a technological standpoint,
In the early 1960s, Professor Ulrich Hütter developed high the significant outcome was the appearance of new German
tip speed designs which had a significant influence on WT manufacturers and development of new concepts, with the
research in Germany and the US. introduction of innovative direct drive generator technology
being particularly noteworthy. Subsequently, a huge expan-
9
sion of the Spanish market has occurred, including wind
farm development, new designs and new manufacturers.
There have been gradual, yet significant, new technology they are inherently less efficient (because of the variation
developments in direct drive power trains, in variable speed in aerodynamic torque with a wide range in angle of attack
electrical and control systems, in alternative blade materi- over a rotation of the rotor). In addition, it was not found
als and in other areas. However, the most striking trend in to be feasible to have the gearbox of large vertical axis tur-
recent years has been the development of ever larger WTs bines at ground level because of the weight and cost of
leading to the current commercial generation of MW the transmission shaft.
machines with a new generation of multi-MW offshore tur-
bines now appearing. The vertical axis design also involves a lot of structure per
unit of capacity taking account of cross arms in the H type
1.1.5 DESIGN STYLES design. The Darreius design (Figure 1.2) is more efficient
structurally. The blade shape is a so-called catenary curve
Significant consolidation of design has taken place since and is loaded only in tension, not in bending by the forces
the 1980s, although new types of electrical generators caused as the rotor spins. However, it is evident that
have also introduced further diversification. much of the blade surface is close to the axis. Blade sec-
tions close to the axis rotate more slowly and this results
Vertical Axis in reduced aerodynamic efficiency. These disadvantages
have caused the vertical axis design route to disappear
Figure 1.2: Darreius Type Vertical Axis Wind Turbine from the mainstream commercial market. FlowWind, the
main commercial suppliers of vertical axis turbines,
stopped supplying them over a decade ago.
Number of Blades 1,2,3, Many?
Small-scale multi-bladed turbines are still in use for water

pumping. They are of relatively low aerodynamic efficiency
but, with the large blade area, can provide a high starting
force able to turn the rotor in light winds which suits
pumping duty.
Most modern WTs have three blades although, in the

1980s and early 1990s, some attempt was made to mar-

ket one- and two-bladed designs.
The single-bladed design (Figure 1.3) is the most struc-

turally efficient as all the installed blade surface area is in
a single beam. It is normal to shut down (park) WTs in very
high winds in order to protect them structurally. This is
because they would experience much higher blade loads
and tower loads if they continued to operate. The one-
bladed design allows unique parking strategies – with the
blade acting as a wind vane or downwind behind the tower
– which may minimise storm loading impact. However,
Due to their expected advantages of omni-directionality there are a number of disadvantages with single-blade tur-
10
and having gears and generating equipment at the tower bines, such as added mass to provide a counterweight to
base, vertical axis designs were considered. However, balance the rotor statically, reduced aerodynamic efficien-
1
VOLUME
Figure 1.3: Single-Bladed Wind Turbine Figure 1.4: Two-Bladed Wind Turbines of Carter Wind
Turbines Ltd.
In rotor design, an operating speed or operating speed

range is normally selected first, having regard to issues
such as acoustic noise emission. With the speed chosen
it then follows that there is an optimum total blade area
for maximum rotor efficiency. The number of blades is, in
cy due to the higher tip loss of a low aspect ratio single principle, open, but more blades imply more slender
blade and complex dynamics requiring a blade hinge to blades for the fixed (optimum) total blade area. This sum-
relieve loads. The designs of Riva Calzoni, Messerschmidt marises the broad principles affecting blade numbers.
and others were of too high a tip speed to be acceptable
in the modern European market from an acoustic point of It is a complete misconception that doubling the number
view. High tip speed is not an intrinsic requirement of the of blades will double the power of a rotor. Instead, if the
single-bladed concept, but is required to optimise the rotor is well designed, it will reduce the power.
design.
In the overall cost benefit, it is hard to discriminate
The two-bladed rotor design (Figure 1.4) is technically on between the two- and three-bladed designs. It is generally
a par with the established three-bladed design. For the wrong to suppose that the two-bladed design saves on the
benefit of a potentially simpler and more efficient rotor cost of a blade, as the two blades of a two-bladed rotor
structure with more options for rotor and nacelle erection, do not equate with two blades of a three-bladed rotor.
either higher cyclic loading must be accepted or a teeter Two-bladed rotors generally run at a much higher tip
hinge introduced. The two-bladed rotor is a little less effi- speed than three-bladed rotors so most historic designs
cient aerodynamically than the three-bladed. would consequently have noise problems. There is, how-
ever, no fundamental reason for the higher tip speed, and
In general, there are some small benefits from increasing this should be discounted in an objective technical com-
blade number, relating to minimising losses that take parison of the design merits of two versus three blades.
place at the blade tips. In aggregate, these losses are
11
less for a larger number of narrow blade tips than for Thus, the one-bladed rotor is, perhaps, more problematic
fewer wides ones. technically whilst the two-bladed rotor is basically acceptable
technically. The decisive factor in eliminating the one-blade Variable Speed versus Fixed Speed
rotor design from the commercial market and in almost elim-
inating the two-bladed design has been visual impact. The Initially, most WTs operated at fixed speed when
apparently unsteady passage of the blade(s) through a cycle producing power. In a start-up sequence the rotor may
of rotation has often been found to be objectionable. be parked (held stopped) and on release of the brakes
would be accelerated by the wind until the required fixed
Pitch Versus Stall speed was reached. At this point, a connection to the
electricity grid would be made and the grid would,
The two principal means of limiting rotor power in high through the generator, hold the speed constant. When
operational wind speeds - stall regulation and pitch regu- the wind speed increased beyond the level at which
lation - are now discussed. rated power was generated, power was regulated in
either of the ways previously described by stall or by
Stall was introduced in Section 1.1.3. Stall regulated pitching the blades.
machines require speed regulation. As wind speed
increases, providing the rotor speed is held constant, flow Subsequently, variable speed operation was introduced.
angles over the blade sections steepen. The blades This allowed the rotor speed and wind speed to match so
become increasingly stalled and this limits power to the rotor could maintain the best flow geometry for maxi-
acceptable levels without any additional active control. mum efficiency. The rotor could be connected to the grid
at low speeds in very light winds and would speed up in
For this to work, the speed of the WT rotor must be held proportion to wind speed. As rated power was
essentially constant and this is achieved through the con- approached, and certainly after rated power was being
nection of the electric generator to the grid. In this produced, the rotor would revert, essentially, to constant
respect, the grid behaves like a large flywheel holding the speed operation with the blades being pitched as neces-
speed of the turbine nearly constant irrespective of sary to regulate power. An important difference between
changes in wind speed. this kind of variable speed operation and conventional
fixed speed operation is that moderate speed variations
Stall control is a subtle process both aerodynamically and are still permitted. This reduces loads on the drive train
electrically and is hard to explain in simple terms. Briefly, and reduces the amount of pitch activity required for
a stall regulated WT will run at approximately constant power regulation.
speed in high wind, not producing excessive power and
yet achieving this without any change to rotor geometry. The design issues of pitch versus stall and degree of rotor
speed variation are evidently connected.
The main alternative to stall regulated operation is pitch
regulation. This involves turning the blades about their In the 1980s, the classic Danish, three-bladed fixed
long axis (pitching the blades) to regulate the power speed, stall regulated design was predominant.
extracted by the rotor. In contrast to stall regulation, pitch Aerodynamicists outside the wind industry (helicopter, gas
regulation requires changes to rotor geometry. This turbine, etc.) were shocked by the idea of using stall. Yet
involves an active control system to sense blade position, because of the progressive way in which stall occurs over
measure output power and instruct appropriate changes the WT rotor, it proved to be a thoroughly viable way of
of blade pitch. operation, making use of, rather than avoiding, stall. It is
one of the unique aspects of wind technology.
The objective of pitch regulation is similar to stall regula-
tion, namely to regulate output power in high operational Active pitch control is the term used to describe the con-
12
wind speeds. trol system in which the blades pitch along their axis like
1
VOLUME
a propeller blade. This approach superficially offered bet- Another significant impetus to the application of pitch con-
ter control than stall regulation, but it emerged through trol and, specifically, pitch control with independent pitch-
experience that pitch control of a fixed speed WT in high ing of each blade, is the acceptance by certification
operational wind speeds above rated wind speed (mini- authorities that this allows the rotor to be considered as
mum steady wind speed at which the turbine can produce having two independent braking systems acting on the low
its rated output power) could be quite problematic. The speed shaft. Hence, only a parking brake is required for
reasons are complex but in turbulent (constantly chang- overall machine safety.
ing) wind conditions it is demanding to keep adjusting
pitch to the most appropriate angle; high loads and power 1.1.6 DESIGN DRIVERS FOR MODERN
can result whenever the control system is “caught out” TECHNOLOGY
with the blades in the wrong position.
The main design drivers for current wind technology are:
In view of such difficulties, which were most acute in high • low wind and high wind sites
operational wind speeds (say 15 m/s to 25 m/s), pitch • grid compatibility
control in conjunction with a rigidly fixed speed became • acoustic performance
regarded as a “forbidden” combination. Vestas initially • aerodynamic performance
solved this problem by introducing OptiSlip – essentially a • visual impact
degree of variable speed with about 10% speed variation • offshore expansion
using a high slip induction generator. Speed variation
helps to regulate power and reduces demand for rapid Although the existing offshore market in terms of installed
pitch action. capacity is only 0.4% of the world’s land-based installed
capacity, the latest developments in wind technology are
Variable speed had some attractions but also had costs primarily driven by the emerging offshore market. This
and reliability concerns. It was seen as the future with means that the technology development focus is on the
expected cost reduction and performance improvements most effective ways to make very large turbines. Specific
in variable speed drive technology. To some extent this considerations are:
has been realised. There was never a clear case for vari- • low mass nacelle arrangements
able speed on economic grounds with small energy gains • large rotor technology and advanced composite
being offset by extra costs and also additional losses in engineering
the variable speed drive. The current drive towards vari- • design for offshore foundations, erection and
able speed in new large WTs relates to greater operational maintenance
flexibility and concerns about power quality of stall regu-
lated WTs. Two-speed systems emerged during the 1980s Of the other main drivers, larger rotor diameters (in rela-
and 1990s as a compromise improving energy capture tion to rated output power) have resulted in order to
and noise emission characteristics of stall regulated WTs. enhance exploitation of low wind speed sites. Reinforced
The stall regulated design remains viable, but variable structures, relatively shorter towers and smaller rotor
speed technology offers better output power quality to the diameters in relation to rated power are employed on
grid and this is now driving the design route of the largest extremely high wind speed sites.
machines. Although some experiments are underway with
a combination of variable speed and stall regulation, vari- Grid compatibility issues are inhibiting further develop-
able speed combines naturally with pitch regulation. For ment of large WTs employing stall regulation. Acoustic
reasons related to the methods of power control, an elec- performance regulates tip speed for land-based applica-
trical variable speed system allows pitch control to be tions and requires careful attention to mechanical and
13
effective and not overactive. aerodynamic engineering details. Only small improve-
ments in aerodynamic performance are now possible (rel- ability, are generally well integrated with the environment
ative to theoretical limits), but maximising performance (Figure 1.5) and accepted by the public.
without aggravating loads continues to drive aerodynamic
design developments. Visual impact constrains design In spite of repeated predictions of a levelling off at an
options that may fundamentally be technically viable, as optimum mid-range size, and in spite of the irrefutable
is the case with two-bladed rotors. logic that WTs cannot get larger indefinitely, turbine size
at the centre of commercial production has increased year
1.1.7 GROWTH OF WIND TURBINE SIZE on year (Figures 1.6 and 1.7).
Modern wind technology is available for a range of sites -

low and high wind speeds, for desert and arctic
climates. European wind farms operate with high avail-
Figure 1.5: Modern Wind Technology

14
1
VOLUME
Figure 1.6: Growth in Size of Commercial Wind Turbine Designs
Figure 1.7: Growth in Size of Commercial Wind Turbine Designs
2003
114m
2002
104m

2000
80m
1997
54m
1996
43m
1960
24m
15
The size of commercial WTs has steadily increased with huge growth in the Spanish market (Figures 1.8 and 1.9)
units of 20-60 kW appearing in the early wind farms in in recent years which has combined elements of licens-
California of the 1980s, up to the latest multi-MW ing, technology transfer and independent manufacture (as
machines with rotor diameters over 100 m. The photo-col- with Vestas and Gamesa), development of established
lage of Figure 1.7 shows turbines from the Gedser proto- turbine manufacturers (such as Ecotecnia and Made), as
typic Danish design (1958-67) through a set of commer- well as the Spanish aerospace industry (MTorres) and
cial turbines up to 2003. The largest is the Enercon E112 integrated developer/manufacturers (EHN).
a land-based prototype, but of a design targeted at the
developing offshore market. Figure 1.9: Spanish Wind Turbines
In large offshore projects, it is clear that almost all the

balance of plant costs associated with foundations, elec-
trical interconnection, access and maintenance will be
reduced (per installed kW of wind farm capacity) if the unit
capacity in a wind farm is increased and hence the num-
ber of units reduced. Many technological factors indicate
that it will be very challenging to develop economically
viable turbines above 5 MW rating, based on the current
architecture. The single most important factor is that tur-
bine costs intrinsically increase more rapidly with diame-
ter than with energy output. New concepts may emerge to
provide generating units larger than 5 MW capacity for off-
shore projects. This is the latest challenge for the wind
industry.
Figure 1.8: Spanish Wind Turbines

1.1.8 ARCHITECTURE OF A MODERN WIND
TURBINE
Many developments and improvements have taken place

since commercialisation of wind technology in the early

1980s, but the basic architecture of the mainstream design
is little changed. Most WTs have upwind rotors and are
actively yawed to preserve alignment with wind direction.
The three-bladed rotor proliferates and, typically, has a

separate front bearing with a low speed shaft connected
to a gearbox which provides an output speed suitable for
a four-pole generator. This general architecture is evident
in the Nordex N54, for example (Figure 1.10). Commonly,
with the largest WTs, the blade pitch will be varied con-
The growth of machine size has been paralleled by the tinuously under active control to regulate power at the
growth of markets and manufacturers. The German mar- higher operational wind speeds. For future large machines
16
ket is the largest in the world and German WT designs are there appears to be a consensus that pitch regulation will
well represented in this chapter. There has also been a be adopted.
1
VOLUME
Support structures are most commonly tubular steel tow- maturity and growth of the wind industry, where crane
ers tapering in some way, both in metal wall thickness manufacturers are producing designs specially suited to
and in diameter from tower base to tower top. Concrete wind farm installation.
towers, concrete bases with steel upper sections and lat-
tice towers are also used but are much less prevalent. Figure 1.11 shows typical stages in the erection of land-
Tower height is rather site-specific and turbines are com- based wind turbines. Often, complete rotors are lifted on
monly available with three or more tower height options. to nacelles. Sometimes, hub and blades are lifted individ-
ually. As the wind industry becomes more consistent the
Figure 1.10: Typical Nacelle Layout of a Modern Wind Turbine supporting industries will develop to supply its demands.
The advent of large cranes in sufficient numbers to
support developments in Germany is a good example.
The drive train shows the rotor attached to a main shaft

driving the generator through the gearbox. It is in the area
of the gearbox that significant developments in basic
design architecture are now appearing, in the form of
direct drive generators. The gearbox is removed and the
aerodynamic rotor drives the generator directly. Hybrid
arrangements involving a single stage gearbox and multi-
pole generator are also appearing. These developments
are discussed in Section 1.3. It is far from clear which of
the configurations is optimum. The effort to minimise cap-
ital cost and maximise reliability continues – the ultimate
goal is to minimise the cost of electricity generated from
the wind.
1.1.9 ERECTION OF LARGE WIND TURBINES
Erection of wind farms and systems for handling ever

larger components have progressed since the early com-
mercial projects of the 1980s. For a period up to the mid
1990s, the allowable mass of components to be lifted to
17
hub height was determined by available cranes.
Subsequently, there has been a shift, indicative of the
Figure 1.11: Typical Erection Procedures for Land-Based Wind Turbines
Preparation of foundation Tower Assembly Hub Lift
Rotor Lift Rotor Lift

18
Nacelle Lift Hub Lift Rotor Assembly
1
VOLUME
1.2 Technology Trends Table 1.1: Diameters of 1.5 MW Wind Turbines
Year Design D (m)

This section investigates various design trends. As tur-
NEG MICON 64C/1500 64.0
bines have grown larger and larger, the way in which
FUHRLANDER FL MD70 70.0
important design parameters change with size can be
PWE 1570 70.0
demonstrated and used to predict how turbines may
REPOWER MD 70 70.0
develop in the future. For various design parameters
SUDWIND S70 70.0
these trends can be used to establish key challenges for
TORRES TWT 1500 70.0
the industry.
GEWE 1.5s 70.5
NEG MICON 72C/1500 72.0

1.2.1 LARGER DIAMETERS
FUHRLANDER FL MD77 77.0
GEWE 1.5sl 77.0

Diameter in relation to power rating has generally
REPOWER MD 77 77.0
increased in recent years. This is clearly illustrated in
SUDWIND S77 77.0
Table 1.1 which shows the various 1.5 MW turbines which
PWE 1577 77.4
have been available in the market over the last few years.
GAMESA G-80 1500 80.0
A remarkable increase from 65 m to 69 m to almost 74
NEG MICON 82/1500 82.0
m in average diameter of 1.5 MW turbines has taken
2003 AVERAGE 73.6
place for the years 1997, 2000 and 2003 respectively.
NEG MICON 64C/1500 64.0
The diameter, or rather the square of the diameter, deter-
ENERCON E-66/15.66 66.0
mines how much energy a WT can produce. The rating, the
WINDTEC 1566 66.0
maximum power that the rotor is allowed to produce,
JACOBS MD 70 70.0
plays an important part in determining system loads.
SUDWIND S70 70.0
Balancing the diameter and the rating is therefore a key
TACKE TW 1.5s 70.5
task in WT design.
TACKE TW 1.5sl 77.0
2000 AVERAGE 69.1

This is partly due to the optimisation of designs to max-
NTK 1500/64 64.0
imise energy capture on comparatively low wind speed
TACKE TW 1.5 65.0
sites, but there is growing interest in better load manage-
ENERCON E-66/15.66 66.0
ment through more intelligent control systems as a means
1997 AVERAGE 65.0
of realising relatively larger rotors and increasing energy
capture. Understanding, predicting, controlling, and there-
by limiting the loads, is a vital part of WT development.
would be expected for a land-based site with the nominal
Considering the range of wind turbine sizes, the increase wind shear exponent of 1/7. Power would be as D2 at any
in diameter to rating ratio of the latest turbines has been given hub height, but as hub height increases with turbine
a consistent trend. In the early 1990s, rated power typi- size, power can be expected to vary as D(2 + 3x1/7) (i.e.
cally varied as diameter, D2.4. This parameter is important approximately as D2.4).
because as the turbine increases in diameter it also
increases in height. There is a relationship between diam- For WT designs in 2003, the exponent has decreased
eter and rating as wind shear causes wind speed to from 2.4 to nearly 2 (Figure 1.12 and Table 1.2).
increase with height. The exponent of 2.4 is exactly what
19
Figure 1.12: Rating versus Diameter of Presently Available given power, a fast turning turbine exhibits lower torque
Wind Turbine Designs, Pr=0.000195D 2.155
(drive train load) than a low speed turbine and hence has
a lower drive train cost. There is therefore a trade-off to be
made between drive train load and noise. For the onshore
market, noise is the major constraint.
Table 1.3: Tip Speed Trends – Land Based and Offshore Technology
Wind Turbine Design Power Dia Tip Speed Offshore to

(MW) (m) (m/s) Onshore Ratio
BONUS 600kW 0.60 44.0 62.2
BONUS 1MW/54 1.00 54.0 42.6
BONUS 1.3MW/62 1.30 62.0 61.7
BONUS 2MW/76 2.00 76.0 67.7
BONUS 2.3MW/82 2.30 82.4 71.6 1.26
De Wind D4 0.60 48.0 73.4
De Wind D6/1000 1.00 62.0 67.2
The largest designs are intended for offshore where there De Wind D6/62 1.00 62.0 81.8
is reduced wind shear and reduced turbulence. The reduc- De Wind D6/64 1.25 64.0 83.1
tion of the exponent of diameter in relation to rated power De Wind D8/2MW 2.00 80.0 86.7 1.14
fits that context. ENERCON E-58 1.00 58.0 72.9

ENERCON E-66 1.80 70.0 80.7 1.11
Matching the power rating and the diameter is a key cost GEWE 1.5s 1.50 70.5 73.8
determinant. Different combinations will appear on differ- GEWE 1.5sl 1.50 77.0 73.8
ent markets. GEWE 3.6s offshore 3.60 104.0 83.3 1.13

NEG Micon NM 750/48 0.75 48.2 55.5
NEG Micon NM 1000/60 1.00 60.0 56.6
Table 1.2: Diameter Exponent of Rated Power
NEG Micon NM 1500/82 1.50 82.0 61.8
Year Exponent NEG Micon NM 1500C/64 1.50 64.0 58.0
1996 2.320 NEG Micon NM 2000/72 2.00 72.0 67.9
1997 2.290 NEG Micon NM 92/2750 2.75 92.0 75.2 1.23
1998 2.250 Nordex N50 0.80 50.0 62.2

1999 2.122 Nordex N54/1000 1.00 54.0 60.8
2000 2.147 Nordex N60 1.30 60.0 60.3
2001 2.119
Nordex N62 1.30 62.0 60.3
2002 2.115
Nordex N80 2.50 80.0 80.0
2003 2.075
Nordex N90 2.50 90.0 79.6 1.31
Overall 2.073
LAGERWEY LW 52/750 0.75 50.5 71.4
LAGERWEY LW 58/750 0.75 58.0 63.8
LAGERWEY LW 70/1500 1.50 70.6 70.2
1.2.2 TIP SPEED - OFFSHORE AND LAND LAGERWEY LW 72/2000 2.00 71.2 89.5 1.31
BASED DESIGNS Vestas V47/660 Vari slip 0.66 47.0 64.0
Vestas V52/850kW 0.85 52.0 70.8
The tip speed of a turbine is the product of the rotational Vestas V66 1.65 66.0 65.7
speed and the radius of the blade. Noise increases very Vestas V80 2.00 80.0 79.6 1.19
20
sharply with tip speed and hence high tip speed turbines
are very much noisier than slow tip speed turbines. For a
1
VOLUME
Tip speed trends of some larger scale designs are sum- costs are quite similar for each design type, but pitch reg-
marised in Table 1.3. The largest turbines (2 MW rating or ulation offers potentially better output power quality (this
more, highlighted) usually target the offshore market. has been perhaps the most significant factor in the
Comparing the tip speed of the offshore design of each German market), and pitch regulation with independent
manufactures with the average tip speed of the designs operation of each pitch actuator allows the rotor to be
targeted at land based applications, a very significant regarded as two independent braking systems for certifi-
increase in tip speed ranging approximately from 10% to cation purposes. There has been some concern, particu-
30% is evident. larly for the largest machines, about stall induced vibra-
tions (vibrations which occur as the blade enters stall).
This makes sense given that less sensitivity to acoustic There has, in fact, been little evidence of these vibrations
noise is expected for offshore sites, which may be up to occurring on a large scale, although there have been spe-
30 km from land. The increased tip speed reduces the cific problems of edgewise vibration of stall regulated
torque associated with any given power level and allows rotor blades associated with loss of aerodynamic damping
useful mass and cost reductions in the power train. in deep stall. However, this has been addressed by intro-
ducing dampers in the rotor blades.
1.2.3 PITCH VERSUS STALL
1.2.4 SPEED VARIATION
There has been an enduring debate in the wind industry
about the merits of pitch versus stall regulation. This was Operation at variable speed offers the possibility of
discussed in Section 1.1.5. increased “grid friendliness”, load reduction and some
minor energy benefits. It is thus an attractive option. Among
Figure 1.13: Ratio of Pitch Regulated to Stall Regulated wind turbines over 1 MW rating, out of 52 distinct models
Designs of ≥ 1 MW Rating of 20 different manufacturers, only three were fixed speed,
12 had two speed systems and 37 employed variable
speed. This shows that it is almost mandatory for MW-scale
turbines to have some degree of speed variation and that
continuously variable speed is the predominant choice.
Variable speed operation is realised in many ways, each

differing in significant details. Direct drive systems have a
natural capability for a very wide speed range although,
even here, some restriction on minimum speed may
reduce the cost of power electronics. The “conventional”
variable-speed concept using a geared drive train connects
the generator to the network through a power electronic
converter and enables systems that may have wide or nar-
row speed ranges. The electrical energy is generated at
variable frequency – a frequency related to the rotational
speed of the rotor – and then converted, by the converter
Until the advent of MW-scale WTs in the mid 1990s, stall or inverter (both power electronic devices) to the frequen-
regulation predominated. However, pitch regulation is now cy of the grid. There are several possible configurations,
the favoured option for the largest machines (Figure based on both synchronous and induction generators.
1.13). There are now more than twice as many pitch reg-
21
ulated turbines on the market than stall regulated ver- The preferred system now is the DFIG (doubly-fed induc-
sions. This is due to a combination of factors. Overall tion generator), also called the wound rotor induction
generator (WRIG). This provides almost all the benefits of several alternative approaches which are being investigat-
full-range variable speed drives, but only a proportion, per- ed. All rely to some degree on power electronics. The cost
haps a third, of the power passes through the converter. of power electronics is falling as a result of activity uncon-
The power converter is thus approximately a third of the nected with the wind industry, but the industry is profiting
size and cost of a conventional variable speed drive, and directly from it. The result will be a reduction in the cost of
its losses are reduced by a similar proportion. In this con- the inverters and hence in the capital cost of the variable
cept, the stator of the electrical machine is connected speed drives and, finally, in the resulting electricity.
directly to the network, and the rotor circuit is connected
via the power converter. This is a modern version of the 1.2.5 HUB HEIGHT
classical Kramer or Scherbius system. The DFIG has a
more limited speed range than the conventional variable- The choice of hub height is site dependent. There is a
speed drive (approximately 1.5 or 2:1, compared to 2.5:1 trade-off between the benefits of the extra energy which
or more). This speed range, however, is sufficient to pro- may result from placing the rotor in the higher wind speeds
vide the benefits listed above. The conventional option of to be found at higher levels above the ground against the
a power converter with the same rating as the generator extra cost of making the towers larger. Hub height equal to
is unlikely to compete with the DFIG until the cost of diameter is a good description of the average trend of the
power electronic converters falls substantially and their largest turbines. There is always great variation in tower
efficiency improves. There is evidence that this point may height for any given size of rotor, with high towers suiting
have been reached, with some manufacturers moving low wind speed sites. There is generally low wind shear off-
over to fully rated converters. shore and less benefit from high towers not withstanding
the extra costs in materials and WT erection. It is therefore
Other novel generator configurations have been proposed expected that large offshore turbines will have a tower
for WT applications, including the switched reluctance height possibly less than or equal to diameter and set at
(SR) (also known as variable reluctance) machine. All rely a level to provide adequate blade tip clearance in extreme
on full-size power converters, and are therefore also at a wave conditions. The distribution of hub height and diame-
disadvantage relative to the DFIG. The DFIG configuration ter is shown in Figure 1.14.
used at present requires slip-rings to transfer power to
and from the rotor circuit. There is an alternative method Figure 1.14: Hub Height Trends, H=3.8786D - 0.3
which in effect transfers the rotor power magnetically,
called the brushless doubly-fed induction generator (BDIG)
which avoids the use of slip-rings. However, at least one

generator manufacturer has concluded that such
machines are inherently larger and more expensive than
the slip-ring option. No commercial turbine uses the BDIG.
As the experience of DFIG with slip-rings is good in WTs,
this remains the preferred option. Slip-ring maintenance
intervals of six months are achieved, and may be
stretched to yearly. Of the 37 mentioned variable speed
designs one, the Gamesa 80, is high slip (7%) and all oth-
ers have speed ratio ranges (maximum steady rated
speed/minimum operating speed) from 1.5 to 3.3. 1.2.6 ROTOR MASS
There were four designs with speed ratios above 2.2 and The rotor accounts for approximately 20% of the cost of the
22
24 with speed ratios in the range 1.8 to 2.2. This is an turbine. Blade manufacturers have naturally sought to
area of WT design which is quite complicated. There are reduce material volume and mass, especially in the largest
1
VOLUME
blades. The way in which design principles change with blade 1.3 Recent Developments
size is therefore very important. If blade stress is kept con-
stant as the size increases (a reasonable design assump- 1.3.1 DIRECT DRIVE GENERATORS
tion) then the blade loads and required blade strength will
both scale as the cube of diameter, implying that geometric Direct drive transmission systems for WTs, avoiding the
similar blades are feasible in a given material and that blade gearbox as a cost and maintenance item, are of increas-
mass will then also scale as cube of diameter. ing interest. Historically, gearboxes have presented chal-
lenges; hence their removal through the direct drive con-
As the blade turns it has to support its own weight and cept may seem desirable. It is, however, possible that
thus bending becomes a dominant loading if inappropriate mechanical difficulties are simply replaced by electrical
materials are used. This scaling is then defeated. In that ones. As yet, there is no clear answer, but the issue may
case, the blade bending moment will scale as the diame- prove to be important for the future development of the
ter to the power of four. industry.
Also, the higher tip speeds of large offshore rotors imply Direct drive generators operate at the rotational speed of
reduced solidity (solidity is essentially the ratio of blade the WT rotor and are directly coupled without need for a
projected area to rotor swept area) and hence slimmer gearbox. The expected advantages are reductions in cap-
blades. The reduced blade area will only allow reduced ital cost, drive train losses, downtime and maintenance
blade mass if materials of sufficiently high specific cost. The simplicity of the system and the avoidance of a
strength are available. Again, this fits in with the gearbox as a maintenance item are definite advantages of
increased prominence of carbon fibre reinforcement in the direct drive system. To date, direct drive systems are
large blade design. As designs evolve with increasing generally heavier than conventional drive trains and the
attention to mass reduction, an overall picture of rotor cost comparison is unclear. Enercon has undoubtedly
mass scaling as less than cubic is apparent (Figure 1.15). been the most high profile wind energy company to com-
mit to direct drive technology almost since the start of its
Figure 1.15: Rotor Mass Trends, Mrotor = 0.000486D 2.6
significant market presence in the early 1990s.
Figure 1.16: Wind Farm of WTs at Borssum
It will be a challenge to maintain this trend (of less than

cubic scaling through improved design concepts and
materials) if rotors continue to get larger. If the power
exponent increases as it will naturally tend to do, then the
23
optimum size of the rotor will reduce. It is therefore in this
area that the search for appropriate materials will focus.
The Enercon designs, now in production at 1.8 MW and diameter. Therefore, larger diameter implies higher
with a prototype E112 at 4.5 MW (Figure 1.17 shows the power output per unit mass.
build of the E112 making apparent the large diameter of • To keep the output frequency in a reasonable range for
the direct drive generator), employ direct drive generators normal electrical machine design (approximately 20 Hz
with wound rotors. in Enercon machines), the number of poles on the gen-
erator rotor should be kept high. There are difficulties
Figure 1.17. Assembly of the E112 in making poles with small dimensions, which there-
fore implies that circumference should be kept large if
possible.
Figure 1.18: The Zephyros LW72 Wind Turbine
The Dutch manufacturer, Lagerwey, markets WTs of 52m

and 58m in diameter with direct drive generators. The
LW52 and LW58 are wound rotor synchronous machines
like Enercon’s. A related design, the Zephyros LW 72
(Figure 1.18 shows the first installation at a site in the
Netherlands, south of Rotterdam), uses permanent mag-
net generators and generates at medium voltage (3-4 kV).

These are, in principle, conventional synchronous
machines with rotor slip-rings and a rotor excitation cir- A four quadrant insulated gate bi-polar transistor (IGBT)
cuit. The other direct drive machines, in production and in rectifier is used at 690 V for the LW52 and LW58. The
development, use permanent magnets on the rotor. LW72 works with an integrated gate commutated thyristor
(IGCT) rectifier at 3 kV (18 rpm, 1,500 kW output) and, with
However, the mass and size of direct drive generators are higher rotor speed, at 4 kV (24 rpm, 2,000 kW). The high-
intrinsically large. In particular, the large diameter of the er speed version is intended for offshore use where noise
generator has implications for nacelle layout and trans- is not an issue. An IGCT converter is considered to be
port, especially for onshore turbines. more reliable because it has fewer and more robust ele-
ments. This is the main difference between the LW52 and
The diameter is large for two reasons: LW58, 750 kW turbines and the LW72. The LW58 and
• The power output is proportional to rotor length times LW72 have generator diameters close to 4 m which is
24
rotor diameter squared, while the active mass (electri- favourable for transport costs, whereas the LW52 (an older
cal steel and copper) is related to rotor length times design) has a generator diameter of over 5m diameter.
1
VOLUME
It is clear that the direct drive generator with fully-rated eter WT. A single stage of planetary gearing (ratio 5.7:1)
power converter concept is commercially viable. It may be is coupled with a low speed (40-146 rpm) multi-pole PMG.
particularly suitable for the offshore market where its The nacelle arrangement is very compact. The PMG uses
potential high reliability may be an advantage and its large rare earth magnets and is water cooled. The nacelle struc-
diameter is not a significant transport restriction. Jeumont ture consists of a simple steel cylinder welded to a stub
Industries has also developed a direct drive permanent cylinder abutting at right angles which contains the slew-
magnet generator system, employed in the J48 750 kW ing ring.
turbine.
Figure: 1.19: A Hybrid Design
In addition to the standard drive train at MW-scale with a
three stage gearbox and four- or six-pole generator, or the
gearless drive train with multi-pole direct drive generator,
an intermediate solution has been considered with a sin-
gle stage of gearing and multi-pole generator.
1.3.2 HYBRID – SINGLE STAGE OF GEARS

AND MULTI-POLE GENERATOR
It may be that the optimum WT design will be the one with

the minimum tower head mass since, in mature produc-
tion, mass may broadly equate to cost. The pursuit of min-
imum mass is therefore a common goal.
The term “hybrid” is adopted to describe a new type of

power train in which a gearbox is used to increase speed, The WinWinD design employs variable speed with individ-
but not to the level at which standard generators of up to ual blade pitching. This WT system is of similar mass to
eight-poles can be used. It can thus be considered as a conventional designs and has a simple and compact
compromise between the fully direct drive and the fully nacelle layout (see Figure 1.20).
conventional solution. The generator is multi-pole and
essentially similar to direct drive designs, but is more
compact, being of relatively higher speed and lower
torque. A concept described as “Multibrid” was initially
proposed by the German consultancy company, Aerodyn.
The idea was to have a single stage of gearing (6:1 is
about the ratio limit achievable in a single stage) driving a
medium speed multi-pole generator. The aim was to avoid
the complexity of a multi-stage gearbox but also have a
lower system mass with a more efficient and compact
nacelle arrangement than is possible with a large diame-
ter direct drive generator. The Multibrid design concept is
now being pursued by Pfleiderer Wind Energy and
WinWinD.
25
The Finnish company WinWinD has developed the
WinWinD (Figures 1.19 and 1.20), a 1.1 MW, 56 m diam-
Figures 1.20: The WinWinD 1.1 MW Wind Turbine
Historically, gearbox problems are divided equally with the procedures still traced back to their boat building origins.
low speed (planetary) stage and high speed stages. Thus, Modern blade manufacturing and testing is illustrated in
the hybrid design is a trade-off between improved gearbox Figures 1.21 to 1.23. Rotor diameters in excess of 100 m
reliability and reduced gearbox cost and increased genera- are now being designed, manufactured and tested for off-
tor mass and cost (compared to a conventional high speed shore applications. LM Glasfiber in Denmark has domi-
generator). Taking account of component costs (capital and nated the independent blade market.
maintenance) and layout issues impacting on structure

costs, there is no fundamental reason why there should not Many manufacturers want to secure component supply by
be an optimum configuration mid-way between the power setting up their own blade manufacturing capability.
train with high speed generator and the power train with Vestas has long been in this position, as has Enercon.
direct drive. As yet, however, it is unclear whether the hybrid Bonus Energy A/S is now manufacturing blades using glass
design route is simply a viable alternative on a par with epoxy with resin infusion technology. NEG Micon owns NEG
other options or if it has a definite advantage. Micon Rotors, the UK plant where wood epoxy blades are
produced. In most cases, the aim is not necessarily to meet
1.3.3 ROTOR BLADE DEVELOPMENTS all blade demand from in-house supply, but rather to have
options for technical and commercial security. Thus, these
General Rotor Blade Development manufacturers and others have also purchased many
blades from independent suppliers.
The vast majority of WT blades are made from glass poly-
26
ester or glass epoxy. Although there is some automation Lightning can cause serious damage to blades and blade
involved in the process it is labour intensive with the tips and all leading manufacturers can offer lightning pro-
1
VOLUME
tection systems usually with metallic tip inserts and down Figure 1.22: NEG Micon Blade Manufacture
conductors embedded in the blades. Lightning is a com-
plex and unpredictable natural phenomenon, but high volt-
age testing helps to prove design solutions. Testing is of
prime importance for new blade designs and ultimate test-
ing and fatigue testing are now routine (Figure 1.21).
With the need for higher dimensional quality, higher specific

strength and mass reduction of large blades, the industry is
being weaned from the basic “boat building” technology of
the lower grades of glass fibre combined with polyester resin
that has served it well over several decades. For the larger
blades, all established manufacturers switched from poly-
ester to epoxy resin infusion some years ago and all new
manufacturers use epoxy resin based systems.
The spar and shell design, both manufactured using

prepregs, is particularly favoured by Vestas. It has advan-
tages in realising fast production with good quality control
and suits the manufacture of lightweight, flexible blades.
These advantages are somewhat offset by a premium in
the material components.
Figure 1.21: NEG Micon Blade Testing Figure 1.23: Bonus Blade Manufacture

© NEG Micon E/S
27
The design of a blade starts with the aerodynamic shape. 1.3.5 OFFSHORE TECHNOLOGY
The aerodynamic properties are principally determined by
the choice of aerofoil. Considerable efforts have been Offshore Wind Farm Installations
made to design aerofoils specifically for WT use. The
requirements of a WT aerofoil are significantly different to Currently, offshore installations only constitute a very small
those of more conventional aeronautical applications and part of the WT market, but offshore wind is set to develop
hence this has been a particularly demanding task. Even in a significant way and the potential offshore market is the
within the WT discipline, stall and pitch regulated rotors main driver for large turbine technology development.
have different design parameters with pitch regulated
aerofoils closer to conventional applications. Various Table 1.4: Operational Offshore Wind Farms
“families” of aerofoil have now been developed. Further
Location Capacity (MW) Turbines Year of
computational efforts may be expected to bear some fruit. installation
Vindeby, Denmark 5 11 Bonus 450kW 1991
Rotor Aerodynamic Devices Lely, The Netherlands 2 4 NedWind 500kW 1994
Tunø Knob, Denmark 5 10 Vestas V39 500 kW 1995
For stall regulated designs a more pragmatic approach Dronton, The Netherlands 17 28 Nordtank 600kW 1997
is needed and a variety of aerodynamic devices is used Bockstigen-Valor, Sweden 3 5 Wind World 500kW 1998
to fine-tune the performance of stall regulated rotors. Blyth, UK 4 2 Vestas 2MW 2000
Middelgruden, Denmark 40 20 Bonus 2MW 2000
These include vortex generators, stall strips, fences,
Utgrunden, Sweden 10 7 GE Wind 1.425MW 2000
dinotails and Gurney flaps. Vortex generators can inhib-
Yttre Strengund, Sweden 10 5 NEG Micon 2MW 2001
it flow separation and increase lift before stall.
Samsø, Denmark 23 10 Bonus 2.3MW 2003
Sometimes this will improve the power curve so increas-
North Hoyle, UK 60 30 Vestas 2.0MW 2003
ing output power in wind speeds just below rated when
Horns Rev, Denmark 160 80 Vestas V80 2MW 2003
a rapid development of stall regulation is then desirable. Nysted, Denmark 158.4 72 Bonus 2.2MW 2003
Stall strips may be used to induce an earlier stall in the Arklow Bank, Ireland 25 7 GEWE 3.6MW 2003
outboard blade sections of a blade that is producing too Total Offshore Installed
much power around rated wind speed. The “dinotail” is Capacity 522.4 2003
an interesting development of Bonus A/S in which a ser-

rated trailing edge – similar to the back tail plates of a
stegosaurus, hence dino(saur) tail, was tried out to mod- Table 1.4 indicates that 522.4 MW of offshore wind have
ify vortex shedding and reduce acoustic emissions. It been installed in European waters, and at the time of
was found to reduce drag generally and hence improve writing, there are declared plans for about 3.5 GW of off-
power performance. shore wind up to a horizon of 2007. About 10 years ago,
the technology started with a “toe in the water” approach
1.3.4 SINGLE BEARING ARRANGEMENT to test turbine operation in the offshore environment. The
turbines were “marinised” with some extra protection, in
Some WT designs have sought to achieve a lower weight some cases de-humidified nacelle space, but otherwise
and more integrated power train by using the gearbox were essentially the same as the land-based technology.
input bearing as the main rotor bearing ( e.g. designs of
Zond Energy Systems and Wind Energy Group, neither The largest WTs now being designed primarily for offshore use
company is now trading). Such a bearing has higher reveal design changes, mainly higher tip speeds (as discussed
friction than a smaller diameter bearing and may be in Section 1.2) and built-in handling equipment in the nacelle.
appreciably more expensive, but it can also realise very With turbines now available of 2 MW rating and above and two
28
significant economies in avoiding a low speed shaft and projects (Horns Rev + Nysted) of over 150 MW capacity each,
in reducing nacelle weight and space demands. the commercially viable offshore wind farm is at hand.
1
VOLUME
Logistics of Offshore Wind Farms always, is low cost and reliable electricity. The two new
considerations are access and construction. The former
Figure 1.24: Blade Handling for Transport by Sea will play a vital role in determining the energy produced and
the latter a large part in determining the capital cost.
The logistics involved in manufacture, transport, erection

and maintenance of offshore multi-MW WTs is a severe
challenge and on a commercial scale is likely to involve
integrated dockyard assembly facilities. In the case of
blades which may be more than 50 m in length, direct
access to the sea from the manufacturing plant is highly
desirable if not essential.
Technology for Offshore

Ironically, given that the offshore environment is generally
considered hostile, for WTs offshore conditions are often Typical stages in the establishment of offshore wind
more benign than many onshore sites. Design issues, farms are illustrated in Figures 1.25 and 1.26 which are
constraints and drivers are different. The ultimate goal, as composite images from various projects.
Figure 1.25: Erection of Offshore Wind Farms
29
Figure 1.26: Erection of Offshore Wind Farms
© Elsam A/S
It is unlikely that there will be any consensus on offshore Offshore Foundations
erection methods in the near future. Some issues are very
site-specific and a variety of craft and handling tools will It is well recognised that the balance of plant and main-
be tried. In some cases, wholly assembled rotors or tower tenance costs will be critical for the viability of offshore
top systems may be handled. In others, the assembly is wind. On land, machine costs may be about 75% of total
much more piecemeal, as with land-based sites. costs with the balance of plant hardware plus lifetime
maintenance costs accounting for the remainder.
Offshore, this split may well be reversed and much atten-
tion is being given to regulating such costs by design.
Table 1.5: Summary of Foundation Concepts
Foundation Type/Concept Application Advantages Disadvantages

Mono-piles Most conditions, preferably Simple, light, versatile. Expensive installation
shallow water and not deep soft Of lengths up to 35 m. due to large size. May
material. Up to 4 m diameter. require pre-drilling a
Diameters of 5-6 m are the next step. socket. Difficult to remove.
Multiple-piles (tripod) Most conditions, preferably Very rigid and versatile. Very expensive construction
not deep soft material. Suits water and installation. Difficult to
depth above 30 m. remove.
Concrete gravity base Virtually all soil conditions. Float-out installation. Expensive due to large weight.
Steel gravity base Virtually all soil conditions. Lighter than concrete. Easier Costly in areas with significant
Deeper water than concrete. transportation and installation. erosion. Requires a cathodic
Lower expense since the same protection system. Costly compared
crane can be used as for with concrete in shallow waters.
erection of turbine.
Mono-suction caisson Sands, soft clays. Inexpensive installation. Installation proven in limited
Easy removal. range of materials.
Multiple suction caisson (tripod) Sands, soft clays. Inexpensive installation. Installation proven in limited
Deeper water. Easy removal. range of materials.
More expensive construction.
Floating Deep waters – 100 m. Inexpensive foundation construction. High mooring and platform costs.
Less sensitive to water depth than Excludes fishing and navigation
30 other types. from most areas of farm.
Non-rigid, so lower wave loads.
1
VOLUME
Particular attention is being given to developing cost effec- Some of the foundation concepts discussed in Table 1.5
tive foundations and a variety of concepts are under con- are illustrated in Figures 1.27 and 1.28. In general, it
sideration. To date, the mono-pile is the most favoured would appear that fewer larger foundations will be more
solution, but much depends on wave loading, ice loading, economically constructed and installed than many small
water depth and seabed conditions. According to ones. This is a significant driver to develop offshore tur-
Milborrow (2003), at least a 20% reduction in foundation bine units of very large capacity.
costs is expected by 2012.
Future of Offshore Wind Technology
A summary comparison of foundation concepts is pre-
sented in Table 1.5. Up to 1990, the general economic view of offshore wind was
rather negative. Two to four times the unit generating cost
compared to land-based installations was typically projected.
Figure 1.27: Monopile, Tripod and Gravity Based Foundations
Figure 1.28: Floating Support These projections ignored various factors that are now
evident with the advent of commercial offshore wind:
• Perhaps most significantly, higher mean wind speeds
are often available at the offshore sites that would
compete with land-based sites to serve a given popu-
lation area.
• There is excellent offshore resource near centres of
consumption (south east England and Long Island
New York, for example) where little land-based devel-
opment is feasible.
• Although infrastructure costs are necessarily much
higher, there is some mitigation of turbine machinery
costs. With relaxation of acoustic noise constraints,
higher tip speeds are feasible, so reducing drive train
torque and cost. Also, there is generally reduced wind
31
turbulence offshore compared to land-based sites.
Figure 1.29: Wind Farm at Horns Rev wind industry, it may not have off-the-shelf equipment that
is optimum for wind farm establishment.
Optimal design for access in testing sea-states and

optimal strategies for maintenance will be some time in
evolving. There is undoubtedly much work ahead in cost
optimisation of offshore wind farm technology with regard
to all the issues of infrastructure (foundations, erection and
maintenance technology and logistics especially), but the
European wind industry has clearly accepted the challenge.
1.4 Technology Status
Figure 1.29 shows the development at Horns Rev, 1.4.1 OVERALL DESIGN TRENDS
Denmark.
How has WT technology evolved since the early 1980s?
It is expected that the wind industry will continue to
develop with an ever sharper focus on the specific needs Although there has always been a wide variety of designs
of offshore technology. Some of this development is seen on the margins of commercial technology, in the early days
in the turbines themselves, in the tendencies towards the Danish, three-bladed, single fixed speed, stall regulat-
increased tip speed and specific maintenance aids. ed turbine dominated the market at rated power levels of
Although the use of helicopters for installation and main- generally less than 200 kW. Blades were almost invariably
tenance operations may be prohibitively expensive, and manufactured from glass-polyester resin.
helicopters are very limited in lift capacity, some manu-
facturers provide helipads on the nacelle of their offshore In 2003, the focus of attention is on technology around
turbines to increase access opportunities for mainte- and above 1.5 MW rating and commercial turbines now
nance engineers. Some offshore turbines have in-built exist with rotor diameters in excess of 100 m. Designs
cranes whilst others have provision for winches to be with variable pitch and variable speed predominate while
brought to the turbine in order to exchange components. direct drive generators are becoming more prevalent.
Although a mature European offshore industry exists in Epoxy-based resin systems predominate in blade manu-
the context of oil and gas recovery, the demands of off- facture and carbon fibre reinforcement is increasingly used
shore wind farms are quite specific and ongoing develop- in big blades. Some manufacturers produce wholly carbon
ment is expected in the areas of foundations, access, blades and many use carbon in cap spars. One company
wind farm electrics, transportation and erection. In the oil has developed means of effectively combining carbon with
and gas industry, maintaining production is of overriding wood laminate. If the trend towards increasing use of car-
importance and justifies high capital cost solutions. In the bon continues, and the offshore market develops sub-
wind industry, production is also vital, but so also is min- stantially, the wind industry could lead world demand for
imisation of capital costs. Oil rigs are massive one-off quality carbon fibre, so driving further cost reductions for
constructions whereas quantity production issues will fig- carbon fibres and prepregs.
ure in the installation and maintenance of a large offshore
wind farm which may have hundreds of turbine units. This Figure 1.30 shows the evolution from the original main-
32
implies that while the existing offshore industry may have stream architecture, stall regulated, fixed speed and with
knowledge and experience of considerable value to the geared transmission to the present, pitch regulated, vari-
1
VOLUME
Figure 1.30: Technology Trends
Time
able speed and with direct drive transmissions appearing, expansion has, of course, promoted incremental techno-
along with continued use of gearbox transmissions. logy improvements in design, materials, processes and
logistics that have contributed very significantly to cost
These design changes are not in any significant degree a reductions. There are significant technological gains from
path to cost reduction. Variable speed may offer a little advances in WT engineering, but no significant cost
more energy capture but this is largely offset by added cost. reduction has come from the most visible changes in
The design changes have largely been driven by market mainstream technology direction - variable speed, direct
demands - better acoustic noise regulation, better output drive and pitch regulation.
power quality, reduction of gearbox problems, etc.
1.4.2 SIZE LIMITATIONS
Since the initial commercialisation of wind energy in the
early 1980s, there have, of course, been huge cost reduc- Frequently, the wind energy industry is asked: “Is there an
tions and this is a direct consequence of the dramatic optimum size for a wind turbine?” The answer is a complex
growth in the market. mixture of economics and technology, and is not well
defined. All predictions show very shallow minima for most
Thus, modern WTs are more sophisticated and adaptable parameters; hence the answer will be very sensitive to the
than their predecessors on account of technology devel- assumptions made and, ultimately, to the practicalities of
33
opment and are also much cheaper (discounting infla- turbine manufacture. The overall trend in price per kW is
tionary factors) on account of market expansion. Market presented in Figure 1.31. Note that this figure reflects pub-
lished price data and project prices are, in general, signifi- Figure 1.32: Price Trends of Large Wind Turbines
cantly lower. Very small turbines are extremely expensive
per kW. They require disproportionately tall towers to clear
boundary layer obstacles and their control systems repre-
sent a relatively high proportion of total cost.
Figure 1.31: Trends in List Price
The specific cost could readily be interpreted as constant

above 40 m diameter. However, excluding the largest off-
shore machines (75 to 80 m plus) which, having higher tip
speeds, form a separate new class on a lower cost curve,
the general trend is suggestive of a rising cost curve.
At larger sizes these factors diminish and costs reduce. In general, design loads increase cubically with rotor
However, the cost (net manufacturing cost of turbine ex- diameter and for constant design stress, cubic increase in
works) rises gradually for rotor diameters in excess of component size will accommodate this trend. This is exem-
about 50 m. Taking account of land utilisation, infrastruc- plified with clarity in Figure 1.33 where the mass of a series
ture costs and maintenance costs, there is, however, still of Enercon main shafts over a wide size range follow an
a net cost benefit in larger turbines for energy production almost perfect cubic curve. It should also be noted that,
rather than for capital cost and this is accentuated off- from a fracture mechanics stand point, design for constant
shore where infrastructure costs dominate. stress independent of scale may not be acceptable at very
large sizes as the probability of a critical flaw existing in any
Energy capture improves with increasing hub height, but given material sample increases with sample size.
the biggest machines are justified for offshore application

and the wind shear effect (which is typically an increase Figure 1.33: Main Shaft Mass – Enercon Designs
of average wind speed with height) is less offshore than
at landbased sites. It is often asserted that component
mass and costs increase less than cubically with scale.
However, the underlying physics is often confused with
the effects of technology development and the influence
of volume on production cost. Often different design con-
cepts are used in large-scale projects, e.g. for gearboxes
or in materials technology, so reducing the specific mass
of very large blades.
The latest and largest offshore designs benefit from

34
increased tip speed ratio and hence are in a different class
from earlier designs. There is a hint of this in Figure 1.32.
1
VOLUME
It is easy to find data suggesting less than cubic scale-up Table 1.6: Blade Mass versus Diameter Trend Line Exponents
of WT components. On a sound basis it can be argued
Manufacturer Exponent
that control system mass or costs will not scale-up cubi-
NOI 2.05
cally. Generators and electrics will scale only as power
ATV 3.07
(diameter squared) if, unlike the input speed provided to
Polymarin 2.76
the gearbox, the generator shaft speed is held constant.
Euros 2.92
However, this is just a trade-off between mass and cost in
LM Glasfider 2.30
gearbox or generator which can vary between the conven-
tional geared transmission, the system with a single
stage of gearing and direct drive. A different picture emerges when the data of Table 1.6
are considered.
A power law fit to the data of Figure 1.34 would give an
exponent of about 1.3, significantly less than cubic. There is further clarification of blade mass trends in
Figure 1.34 where blade mass data from manufacturers’
However, design development with time is being confused data sheets are presented. It is apparent from this figure
with inherent physical scaling. Old, relatively heavy, that the small carbon blades of ATV (A Tout Vent) are very
blades are mixed with new ones where great effort and light for their size, as would be expected with CFRP as the
the benefit of longer manufacturing experience have con- principal material. The wood epoxy blades of NEG Micon
tributed to significant specific mass reduction. Rotors (Aerolam in the legend of Figure 1.34) are also of
low specific mass.
Thus, blade mass can appear to scale with increasing
blade size by a power law that is much less than cubic. Table 1.6 corroborates the comments explaining why
This is principally because: blade mass apparently increases as less than cubic.
• the large blades of 40 m length or more are very sub- Trend line equations (not shown in Figure 1.34 to avoid
stantial structures and particularly focus development undue complication) in the form of best fit power law
effort to reduce mass and cost; curves were determined for the data of each manufactur-
• the largest blades are the most recent and at the er. In the case of ATV who has a consistent technology for
most advanced stage of manufacturing technology. small blades and Euros who are new entry manufacturers
with little development time behind them, the power law
exponents are close to the predicted cubic relationship.
Figure 1.34: Blade Mass Trends Based on Blade
Manufacturers’ Data The other manufacturers LM, NOI (formerly Aerpac) and
Polymarin have been trading for a long time, their tech-
nology has gone through major developments and the
power law exponents appear to be less than cubic. It is
logical that the lowest exponent of all applies to NOI since
their blade range includes small Aerpac blades originally
manufactured with polyester resin in a wet lay-up process,
whereas all recent designs of large blades are based on
resin infusion using epoxy resin.
Scaling of WTs is inevitably more complex than can ade-

quately be addressed here. Technology developments
will confront the up-scaling problems, as is now happen-
35
ing with increased focus on mass reduction of MW-scale
systems.
To a certain extent, whether turbine-specific cost rises been adjusted for inflation. It is valuable to have data
appreciably with up-scaling is controversial and the cost- from a single manufacturer where there is consistency of
optimum turbine size remains uncertain (and is properly a design and cost evaluation. The decrease in turbine price
secondary issue to the overall cost factors applicable to is very evident.
specific projects). However, the industry is faced with an
important issue, to look much more carefully at scaling Figure 1.35: Wind Turbine Price Reduction
trends in order to identify an economic size limit of con-
ventional turbine technology and consider new concept
development for very large-scale offshore units.
This is not a make or break issue as world market growth

can be well satisfied with turbines within the compass of
present technology, i.e. up to about 5 MW, but it is
nevertheless of some importance in striving for the best
economics in future offshore projects.
1.4.3 THE SUCCESS OF WIND TECHNOLOGY
Figure 1.35 shows the reduction in turbine cost with time.

It is based on sales prices provided by Bonus and has
Figure 1.36: Typical Availability for a Large Wind Farm Since Erection
36
1
VOLUME
Figure 1.36 shows a time history of availability from a • Design studies of systems rated above 5 MW for off-
large European wind farm. It demonstrates that the avail- shore including possibly multi-rotor systems.
ability in mature operation is in excess of 98%. It also • Variable speed DC or AC HV generation for offshore.
shows the way in which the availability climbs as the tur- • Offshore meteorology - hardware for measurements
bines “run in”. This project was built and commissioned and modelling issues.
in three phases. These phases can be seen coming on- • Integration of support structure design for offshore
line during the first two years. In conjunction with decreas- turbines.
ing cost, performance has risen. Aerodynamic design • Improved access methods for offshore turbines.
improvements and variable speed operation have realised • Condition monitoring of critical components.
small energy gains. But above all availability, which • SCADA for offshore - development for remote interven-
reflects the operational reliability of the product, has risen tion.
(Figure 1.36) to average levels generally above a target of • Development of alternative and deep water foundation
97.5%; this compares very favourably with all other elec- structure arrangements
tricity generating technologies. • Floating turbines.
1.5 CONCLUDING REMARKS AND FUTURE There are also many issues around standards, develop-
R&D ment of manufacturing processes and computer design
tools. This list is not comprehensive and does not sug-
The development of WTs is a remarkable success story gest any priorities but rather gives a flavour of the many
which is not yet complete. The wind industry is now poised areas where R&D support can benefit the wind industry.
at a stage where it is regarded by some as a mature tech-
nology and able to stand on its own commercially. While
that status is a great achievement, it is important to
realise the potential for yet greater growth that can best
be furthered by continuing vigorous R&D efforts. Much of
the current R&D focus will be on supporting the new drive
towards offshore technology. The design drivers are
always reduction in cost and increased reliability. A WT is
a complicated integrated structure - all its elements inter-
act and each will play its part in the optimisation. Whilst
Chapter Five will deal in greater detail with R&D require-
ments, a short indicative list is presented here to give a
broad view of future R&D demands.
• More intelligent control systems with additional sensors

measuring system vibrations.
• Advanced adaptable rotor concepts.
• Aerofoil design targeted at control of loads.
• Higher tip speed designs for offshore.
• Rotors making increasing use of carbon reinforcements.
• Low solidity, downwind, flexible rotor designs.
• Direct drive PMG technology including rare earth mag-
nets and alternative electrical machine topologies.
37
• Special designs of systems and components for erec-
tion, access and maintenance of offshore turbines.
2 WIND RESOURCE ESTIMATION
2.1 Introduction These questions, and more, are addressed in this vol-
ume. The first section looks at strategic “raw” resource
Wind is the fuel for wind power stations. Small changes issues, while the following sections provide a detailed
in wind speed produce large changes in the commercial step-by-step evaluation of the assessment process.
value of a wind farm. For example, a two-thirds increase A worked example of a real wind farm is provided in
in the wind speed might be expected to double energy Appendix C and, finally, some recommendations are
production over the lifetime of a wind farm. made on the key issues that need to be tackled in the
near future to help wind energy reach its full potential.
This chapter explains why knowledge of the wind resource
is important for each and every stage of the development
of a wind farm, from initial site selection right the way 2.2 Regional Wind Resources
through to operation.
Naturally, wind energy developers are very interested in
Europe has an enormous wind resource. It can be con- the energy that can be extracted from the wind, and how
sidered on various levels. At the top level, the potential this varies by location. Wind is ubiquitous; in order to
resource can be examined from a strategic standpoint: make the choice of development site both an affordable
• Where is it? and a manageable process, some indication of the rela-
• How does it compare to EU and national loads? tive size of the “wind resource” across an area is very
• Which regions and areas offer good potential? useful. The wind resource is usually expressed as a wind
speed or energy density and, typically, there will be a cut-
At the next level, it is necessary to understand the actual off value below which the energy which can be extracted
wind resource on a site in detail: is not sufficient to merit a wind farm development.
• How is it measured?
• How will it change with time? On-site Measurement
• How does it vary over the site?
• How is it harnessed? The best, most accurate, indication of the wind resource
at a site is through on-site measurement using an
It is at this stage that commercial evaluation of a wind anemometer. This is, however, a fairly costly and time-
farm is required and accurate estimates must be provid- consuming process.
ed which are bankable. Once the wind speed on the site
has been estimated, it is then vital to make an accurate Computer Modelling

and reliable estimate of the resulting energy production
from a wind farm which might be built there. This requires On a broader scale, wind speeds can be modelled using
wind farm modelling and detailed investigation of the envi- computer programs which describe the effects on the
ronmental and ownership constraints. wind of parameters such as elevation, topography and
ground surface cover. These models must be primed with
As the contribution of wind energy to electricity production some values at a known location; usually, this role is ful-
increases, in the context of liberalised energy markets, filled by local meteorological station measurements or by
new questions are beginning to emerge, which are criti- other weather-related recorded data.
cally linked to the nature of the wind:
• How can wind energy be consolidated, traded and Typically, these wind-mapping programs will derive gridded
generally integrated into conventional electricity systems? or contour values for a specified height to create a “wind
• Will an ability to forecast wind farm output help this atlas.” Wind atlases have been produced on a very wide
38
integration? range of scales, from the global down to a local govern-
ment region. They represent the best estimate of the wind
1
VOLUME
resource across a large area. They do not substitute for of the development. Because technology undergoes pro-
anemometry measurements; rather, they serve to focus gressive development, and the nature of constraints will
investigations and indicate where on-site measurements evolve as solutions are found and new constraints
would be merited. emerge, energy resource estimates tend to be valid only
for a limited time.
As a further stage in investigations, theoretical wind
turbines (WTs) can be placed in a chosen spacing, within
a geographical model containing wind speed values as a 2.3 Wind Atlases
gridded dataset. This is usually computed in a geographi-
cal information system (GIS). Using assumptions about the 2.3.1 ONSHORE
technology conversion efficiency to units of energy, it is
possible to derive an energy estimate which corresponds Figure 2.1 shows the onshore wind energy resource as
to a defined area. This is typically expressed as region X computed on a broad scale for the 1989 European Wind
having a wind energy potential of Y units of energy. Atlas. The map shows wind speeds at a height of 50 m
above ground level, which reflects the height of WTs at
Constraints that time. Because wind speeds increase with height, and
because higher wind speeds mean that much more ener-
Most wind energy resource studies start with a top-level gy can be extracted (this is discussed in more detail in
theoretical resource, which is progressively reduced by later sections), the average height of WTs has shown a
including so-called constraints. These are those consider- steady increase in the past decade. So, wind speeds
ations which will tend to reduce the actual area available experienced by today’s commercial technology are higher
to the wind energy developer. They can be geographically- than those shown in Figure 2.1.
delineated conservation areas, for instance, or areas
where the wind speed is not economically viable, or areas The wind speed above which commercial exploitation can
of unsuitable terrain. These areas are then sequentially take place varies according to the specific market condi-
removed from the area over which the energy resource is tions. While countries such as the UK and Ireland clearly
summed. have exceptional potential, every European country has an
exploitable wind resource.
Different estimates of the potential energy resource can
be calculated according to assumptions about the area The European Wind Atlas employs meteorological data
which will be available for development. The resource from a selection of monitoring stations, and shows the dis-
without constraints is often called the “theoretical” tribution of wind speeds on a broad scale. It has been
resource. Consideration of technical constraints results in used extensively by developers and governments to esti-
the estimation of a “technical” resource and, finally, con- mate the size of the wind resource and regional variations.
siderations of planning, environmental and social issues It is possible to map wind speeds at a higher resolution
results in the estimation of a so-called “practical” using, for instance, more detailed topographical data and
resource. a larger sample size of meteorological data, in order to
show more local variations in wind speed; these can be
There are, inevitably, limits to which these modelling exer- used by developers looking for sites in a particular country.
cises can reflect reality – data availability is the main lim-
itation, but also some constraints simply cannot be mod- There are far too many examples of national, regional and
elled accurately. Such studies are useful in estimating local wind atlases, for Europe and the rest of the world, to
upper bounds on deployment, the effects of known con- mention them all here. When investigating a particular region
39
straints, interactions between constraints and likely pat- or country for its wind development potential, one of the first
terns of development. A GIS also helps visualise the scale questions is – “Is there a wind atlas for this area?”
A review of national wind atlases for European countries the development of a wind-mapping tool called WAsP
has been undertaken for this edition of Wind Energy – The which is used widely for both broad-scale wind mapping
Facts, the results of which are shown in Table 2.1. Where and more site-specific applications. Table 2.1 distinguish-
permission has been granted, map reproductions are con- es between the use of WAsP and other wind mapping
tained in Appendix A. The European Wind Atlas resulted in methods.
Figure 2.1: European Wind Atlas, Onshore (EU-12)

40
Source: Risø National Laboratory, Denmark. See Appendix A for colour version.
1
VOLUME
Table 2.1: Wind Atlases in the EU-25 2.3.2 OFFSHORE
Country Coverage in the Other WAsP Other Model

There are two published offshore wind maps for Europe.
European Wind Atlas Application
One is an extension of the onshore European Wind Atlas
EU-15 (see Figure 2.2). Note that wind speeds are provided for a
Austria 1
range of heights. The 100 m height values are the most
Belgium appropriate for current offshore turbines. There is also a
Denmark 2
3 1995 European Commission-funded study (Garrad Hassan,
Finland 4 Germanischer Lloyd, Windtest, 1995), which produced off-
France shore wind maps for each of the (then) EU countries (repro-
Germany 5 duced for each country in Appendix B). Another European
offshore wind map is forthcoming from the POWER project,
Greece 6
also funded by the European Commission.

Ireland 7
8
Italy 9
Luxembourg
2.4 Energy Estimates
The Netherlands
Portugal
2.4.1 ONSHORE
Spain
Sweden 10
There are very few estimates of the onshore wind energy
UK 11
potential for the whole of Europe, and those that do exist
New Member States make assumptions which are, given today’s technology,
Cyprus conservative. Employing wind data (Grubb and Meyer,
Czech Republic 1 1993) made one of the first worldwide estimates of
Estonia 13
14 onshore wind energy potential. For Western Europe, they
Hungary 1 estimated a gross potential of 31,400 TWh/year and a
“second order” potential of 4,800 TWh/year, the latter
Latvia 13
employing constraints based on population density.

Lithuania 13
Corresponding estimates for Eastern Europe and the

Malta
Soviet countries were 106,000 and 10,600 TWh/year. At
Poland 15
the time the estimates were made, Grubb and Meyer pre-
Slovakia 1
dicted future turbines would reach a height of 50 m.

Slovenia 1
Others A 1993 assessment of the technical onshore wind resource

Armenia 16
for OECD countries (van Wijk and Coelingh, 1993) and ref-
Croatia 1
erenced in the first Wind Energy - The Facts publication
Norway 12
(1999), remains the only European-wide estimate of
Russia 17 onshore wind potential which gives a comparable estimate
for each country. The study presents figures for the “mete-
1
Dobesch and Kury (1997) 10
Krieg (1992, 1999)
2
Risø (1999) 11
Burch and Ravenscroft (1992) orological” potential, which is expressed as the land area
3
Petersen et al. (1981) 12
Vector (2001, 2003)
4
Tammelin (1991) 13
Rathman (2003) on which wind speeds of 5.1 m/s and over are experi-
5
Traup and Kruse (1996) 14
Steinrücke, et al. (1996)
6
CRES (2001) 15
Sander et al. (2003)
enced, as well as the “site” potential, which is the former
7
Watson and Landberg (forthcoming) 16
Elliott et al. (2003) land area minus land considered to be unsuitable for terrain
8
TrueWind Solutions (2003) 17
Starkov et al. (2000) 41
9
Podesta et al. (2002) or climatic reasons. It then estimates a “technical” wind
energy potential, which further reduces the site potential by
Figure 2.2: European Wind Atlas, Offshore
Source: Risø National Laboratory, Denmark. See Appendix B for colour version.
assuming that only 4% of the land could be used as a result total electricity production and wind energy production
of practical and social constraints. An energy estimate is (expressed in TWh and as a percentage of the van Wijk and
42
derived assuming 8 MW/km . The technical potential is
2
Coelingh technical potential).
reproduced in Table 2.2 below, together with figures for
1
VOLUME
While updated assumptions on both the technology (larger ity factor of 30% – the same as that employed by the
machines) and constraints would increase van Wijk and study to derive the MW estimates (see Table 2.3). The
Coelingh’s technical potential (Table 2.2) is nonetheless wind energy potential estimates were based on previous
instructive in two key respects. It shows that even with con- estimates for the USSR in a 1989 publication Master Plan
servative estimates of technical potential, only Germany of Wind Power Development in the USSR until 2010, and
has managed to exploit its wind resource to anything other supporting information gathered by the study team.
approaching the potential estimated in 1993. It also
demonstrates very clearly that those countries with the Table 2.3: Eastern European Countries, Onshore Potential, EBRD
biggest resources are not necessarily those which have
Country Yr 2001 EBRD Wind Energy
exploited them most effectively. Net TWh Assessment Production
Consumption1 (TWh/yr) (TWh); and
% of EBRD
Table 2.2: EU-15 and Norway, Technical Onshore Potential Assessment
Country Yr 2002 van Wijk & Coelingh Wind Energy

Consumption Technical Potential Production New Member States2
(TWh)1 (TWh/yr) (TWh, 2002)2; Czech Republic 55.6 5.8 0.06 1
and % of Tech-
nical Potential Estonia 6.2 1.3 0.02 2
Austria 60.15 3 0.24 8 Hungary 35.1 1.3 0.01 1
Belgium 81.73 5 0.088 2 Latvia 6.0 1.4 0.06 4
Denmark 34.01 29 5.28 18 Lithuania 8.7 1.3 0 0
Finland 79.64 7 0.08 1 Poland 118.8 10.5 0.19 2
France 431.83 85 0.20 0.2 Slovakia 24.4 0.7 0 0
Germany 531.78 24 18.49 77 Slovenia 13.8 0.3 0 0
Greece 48.60 44 0.65 1 Other EBRD
Ireland 22.14 44 0.27 1 Albania 5.9 0.1 0 0
Italy 295.08 69 1.18 2 Armenia 5.8 1.1 0.09 8
Luxembourg 5.65 0 0.048 - Azerbaijan 16.6 3.9 Negligible 0
The Netherlands 105.81 7 1.36 19 Belarus 26.7 0.5 Negligible 0
Portugal 42.55 15 0.39 3 Bosnia/Herzegovina 8.1 0.1 0 0
Spain 221.42 86 11.95 14 Bulgaria 32.5 8.9 0 0
Sweden 138.16 41 0.66 2 Croatia 14.3 2.6 0 0

UK 349.20 114 1.45 1 Georgia 7.6 6.0 0 0
Norway 114.94 76 0.26 0.3 Kazakhstan 48.4 21.0 0 0
TOTAL 2,562.69 649 42.60 6.6 Krgyzstan 10.5 3.9 0 0
FYR Macedonia 6.1 0.1 0 0

1
Extrapolated from 2001, using IEA data from “Electricity Information 2003”.
Moldova 3.2 1.3 0 0
2
Estimated from installed capacity, using capacity factors derived from year 2000 Eurostat
production data. Romania 46.1 7.9 Negligible 0
Russia 773.0 157.7 0.01 0.006
Tajikistan 14.5 2.6 0 0

The European Bank for Reconstruction and Development
Turkmenistan 8.5 26.3 0 0
(EBRD) has recently commissioned a series of renewable
Ukraine 152.4 13.1 0.11 0.8
energy assessments in its countries of operation (Black
Uzbekistan 47.1 2.6 0 0
and Veatch, 2003). This included an estimate, for each
FR Yugoslavia 32.4 0.3 0 0
country, of the realisable wind energy potential (in MW
capacity) by 2020, and a figure for currently installed/ 1
From US Department of Energy.
2
Those not covered by the EBRD assessment are not included. 43
under construction wind energy capacity. Both of these
estimates have been converted to energy, using a capac-
2.4.2 OFFSHORE Figure 2.3 explains the upward trend in turbine size. For line
configurations of wind parks, the installed power per unit of
The only publicly available, consistent, energy estimates length increases at a rate greater than linearly with diame-
for the offshore wind resource are from the previously ter which would be expected from simple considerations.
mentioned European Commission study (Garrad Hassan,
Germanischer Lloyd, Windtest, 1995). These are repro- Figure 2.3: Relationship between Power and Turbine
duced in Table 2.4. The Commission-funded CA-OWEE Dimensions
project (Concerted Action on Offshore Wind Energy in
Europe, Delft University et al., 2001), collated estimates
provided by each member state, and these are also
shown in Table 2.4. These estimates are based on a vari-
ety of source material.
Table 2.4: Offshore Wind Energy Estimates, Europe
Country Yr 2002 Consumption GH-GL Study CA-OWEE Survey

(TWh) (TWh/y) (TWh/y) P: installed power (kW)
D: rotor diameter (m)
H: tower height (m)
Belgium 78.0 24 4
Denmark 37.2 550 26
Finland 71.9 20 Source: ECN (2003).
France 533.3 477 44
Germany 543.5 238 45
Greece 48.9 92 5 As an example of how more up-to-date assumptions
Ireland 22.9 184 11 can change resource estimates, Table 2.5 compares
Italy 270.3 154 10
estimates for the UK’s onshore wind resource pub-
The Netherlands 92.4 137 33
lished in 1994 and later revised in 1997 (Brocklehurst,
Portugal 44.1 48 2-3
Spain 216.3 140 7 1997). The same underlying wind data were used for
Sweden 143.4 22.5 both studies. Principal differences were in the “typical”
UK 361.5 986 230 - 334 technology assumptions made in 1994 and 1997
Total 2,463.7 3,030 459.5 - 564.5 – 300 kW and 25 m hub height in 1992 compared with
600 kW and 45 m in 1997 – and in the availability and
use of constraints data. The latter has no effect on the

2.4.3 UPDATING RESOURCE POTENTIAL technical or “feasible” resource, but allows a more
sophisticated and thorough estimate of the practical
An estimate of the potential wind energy resource is not a resource.
fixed quantity. It changes over time as the technology
develops and as more is learned about its performance Table 2.5 shows data for the so-called feasible, accessi-
and about the technical, environmental and social con- ble and practical resource. The practical resource is sig-
siderations which influence the density and location of tur- nificantly greater in 1997 than in 1994 due largely to the
bine deployment. This changing quantity is not unique to increase in machine size. The accessible resource is
the wind industry - it is an established concept in the oil slightly smaller in 1997, due to assumptions on protect-
and gas industry, where estimates of recoverable oil and ed areas. The practical resource is greater in 1997 due to
gas reserves are continually being revised as the technol- a combination of factors relating to technology, deploy-
ogy improves and as new discoveries are made. ment and economic factors.
44
1
VOLUME
Table 2.5: UK Onshore Resource Estimates recording an appropriate set of site wind data and the
methodologies which can be used to predict the expected
Feasible (TWh) Accessible (TWh) Practical (TWh)
long-term energy production of a project.
1994 204.120 343.730 37.407
1997 660.787 317.854 57.627

Figure 2.4 provides an overview of the whole process. The
sections below describe this process step-by-step.
Appendix C provides a worked example of a real wind farm
2.4.4 CONCLUDING REMARKS which has been operational for a year and for which these
techniques were used to estimate its long-term energy
This brief section has collected together the available production.
wind energy estimates for the EU-15 and the new member
states. There is no complete integrated report which can
be drawn upon to provide a single estimate of the wind
energy resource in the EU-25. It would seem an appropri-
ate and useful task to develop a single common approach
so that the wind energy potential for an enlarged EU can
be assessed. Plans can then be made to promote wind
energy as widely as possible. The real future for wind is
through large-scale integration and hence this approach
would have both technical and strategic merit.
2.5 Local Wind Resource Assessment

and Energy Analysis
2.5.1 INTRODUCTION
The previous section has presented wind maps for Europe

and has considered the wind resource at a strategic level.
The purpose of this section is to consider the resource
assessment and modelling at a local, wind farm, level. To
the wind farm developer, regional wind maps are valuable
tools for site finding, but will not be of sufficient accuracy
to justify the financing of a development. The single most
important characteristic of a site is its wind speed, and
the performance of a wind farm is very sensitive to uncer-
tainties and errors in the basic wind speed estimate.
For the majority of prospective wind farms, the developer

must carry out a wind resource measurement and analy-
sis programme. This must provide a robust prediction of
the farm’s expected energy production over its lifetime.
This section discusses the issues which are pertinent to
45
Figure 2.4: Overview of the Energy Prediction Process
46
1
VOLUME
2.5.2 THE IMPORTANCE OF THE WIND sitive to wind speed. A change in wind speed of just a
RESOURCE few per cent makes an enormous difference financially.
Wind energy has the attractive attribute that the fuel is In summary, the single most important characteristic of
free and will continue to be free for the project’s lifetime a wind farm site is the wind speed. Every effort should
and beyond. The economics of a project depend, crucial- be made to maximise the length, quality and geographi-
ly, on the site wind resource. At the start of the project cal coverage of the data collected. However, measure-
development process, the long-term mean wind speed at ments are undertaken at the very beginning of a project
the site is not known. To illustrate the importance of this and some compromise is therefore inevitable.
measurement, Table 2.6 shows the energy production of
a 10 MW project for a range of long-term annual mean 2.5.3 BEST PRACTICE FOR ACCURATE WIND
wind speeds. SPEED MEASUREMENTS
It can be seen that when the long-term mean wind speed These results illustrate the importance of having an accu-
is increased by 67% from 6 m/s to 10 m/s, energy pro- rate knowledge of the wind resource. A high quality site
duction increases by 134%. This range of speeds would wind speed measurement campaign is therefore essential
be typical of Bavaria at the low end and Scotland or to reduce uncertainty in the predicted energy production of
Ireland at the high end. As the capital cost does not a proposed project. The goal for a wind measurement
strongly depend on wind speed, the sensitivity of the campaign is to provide information to allow the best pos-
project economics to wind speed is clear. Table 2.6 illus- sible estimate of the energy on the site to be provided.
trates the importance of having as accurate a definition The question then arises of how many masts to use and
of the site wind resource as possible. The sensitivity of how high they should be.
energy yield to wind speed variation varies with the wind
speed. For a low wind speed site the sensitivity is Number and Height of Meteorological Masts
greater than for a high wind speed site. For example, at
a low wind speed site a 1% change in wind speed might For a small or medium sized wind farm site it is likely that
result in a 2% change in energy, whereas at a high wind one meteorological mast will be sufficient to provide an
speed site the value might be only 1.5%. accurate assessment of the site’s wind resource. For
large wind farms, say in excess of 20 MW, located in
Table 2.6 illustrates long-term mean wind speeds at dif- complex terrain, it is likely that more than one mast will
ferent sites. The same comments apply at an individual be required to give an adequate definition of the wind
site. The commercial value of a wind farm development resource across the site. In simple terrain, and where
depends on its energy yield which, in turn, is highly sen- there is already a lot of experience at neighbouring wind
Table 2.6: Sensitivity of Wind Farm Energy Production to Wind Speed
Wind Speed (m/s) Wind Speed Energy Production Energy Production Capital Cost Normalised to
Normalised to 6 m/s of 10 MW Wind Farm Normalised to 6 m/s site 6 m/s site
(%) (MWh/annum)1 (%) (%)
5 83 11,150 63 100
6 100 17,714 100 100
7 117 24,534 138 102
8 133 30,972 175 105
9 150 36,656 207 110
10 167 41,386 234 120
47
1
Assumes typical turbine performance, air density of 1.225 kg/m3, total losses of 12 % and Raleigh wind speed distribution.
farms, the performance of these wind farms can be used hub height measurement is not made then it will be nec-
in lieu of a measurement campaign. A great many tur- essary to estimate the shear profile. This can be done,
bines have been sited in this way in North Germany and but it produces uncertainties. Commercial wind turbines
Denmark. However, great caution must be exercised in often have hub heights in excess of 60 m. Just five years
extending this approach to more complex areas. ago, a typical hub height was 30 to 40 m.
The locations and specifications of the mast or masts A 40 m meteorological tower can be erected by a small
need to be considered on a site-specific basis but, in gen- crew of experienced people and is relatively cheap at
eral terms, if there are significant numbers of turbines approximately €20,000. Higher masts are much more
more than one kilometre from a meteorological mast in expensive to erect, around €100,000, and also more
terrain which is either complex or in which there is signif- awkward to handle. Given the cost sensitive nature of this
icant forestry, it is likely that additional masts will be stage of the development there is a compromise to be
required. In such circumstances, discussion with the ana- made between expensive accurate measurements at hub
lyst who will have responsibility for assessing the wind height or cheaper measurements at a lower height which
resource at the site is recommended at an early stage. will be subject to more uncertainty. Often early prospect-
ing is undertaken with a short mast and further higher
The wind speed generally increases with height, as illus- masts are added if the site appears promising.
trated in Figure 2.5.
Specification of Monitoring Equipment and Required
Figure 2.5: The Atmospheric Boundary Layer Shear Profile Signals
A typical anemometry mast will have a number of

anemometers (devices which measure wind speed)
installed at different heights on the mast and one or two
wind vanes (devices which measure wind direction).
These will be connected to a data logger at the base of
the mast via screened cables. It is unusual for there to
be a power supply at a prospective wind farm site, so the
whole anemometry system is usually battery operated.
Some systems have battery charging via a solar panel or
small wind turbine (WT). For some systems, particularly

in cold climates, temperature measurement is important
to detect icing of the anemometers. In such circum-
stances, the use of heated or “ice free” anemometers is
beneficial; however, their use without an external power
source is usually impractical. Measurement of the atmos-
pheric pressure at the site is desirable, but often not
essential.
The figure shows, schematically, the way in which the
wind speed grows. This characteristic is called “shear”
and the shape of this profile is known as the “wind shear
profile”. Given the discussion above about the importance
of accurate wind speed measurements, it is clear that it
48
will be important to measure the wind speed as near to
the hub height of the proposed turbine as possible. If a
1
VOLUME
Signals which would typically be recorded for each sensor It is important to be aware of the potential influence of
with a 10-minute averaging period are: the support structure on the measured data. Detailed
• Mean wind speed guidance is provided in IEA Annex XI edition 1999, IEC 6-
• Maximum three second gust wind speed 1400 Part 12, (www.measnet.com) on specific separa-
• True standard deviation of wind speed tion distances which are required to reduce the influence
• Mean wind direction of the support structure on the measurement to accept-
• Mean temperature able levels. Illustrative examples which demonstrate
• Logger battery voltage good and poor mounting arrangements are presented in
Figure 2.6.
In recent years, it has become increasingly common to
download data via modem. This approach has made man- Figure 2.6: Summary of Good Practice (left) and Poor
aging large quantities of data from masts on a range of Practice (right) Mounting Arrangements
prospective sites significantly more efficient than manual
downloading. It also has the potential to improve data cov-
erage rates.
Recommendations provided by the International

Electrotechnical Committee (IEC), the International
Energy Agency (IEA) and MEASNET (IEA Annex XI edition
1999, IEC 6-1400 Part 12, www.measnet.com) provide
substantial detail on minimum technical requirements for
anemometers, wind vanes and data loggers. It is strong-
ly recommended that anyone intending to make “bank-
able” wind measurements should refer to these docu-
ments. Historically, a notable deviation from best prac- If the guidance presented above is followed, a high qual-
tice as defined in the IEC and IEA documents is the use ity set of wind data should become available, in time,
of anemometers which have not been individually cali- from a prospective site. The absolute minimum require-
brated for the assessment of the wind resource at the ment is one year to ensure that any seasonal variation is
site. Each sensor will have a slightly different operational properly captured. In addition to specifying and installing
characteristic as a result of variations in manufacturing appropriate equipment, vigilance is required in the regu-
tolerances. The use of individually calibrated anemome- lar downloading and checking of data to ensure high lev-
ters has a direct impact on reducing the uncertainty in els of data coverage are achieved. It will be necessary to
the predicted wind speed at a site and is therefore to be demonstrate, either internally or externally, the prove-
recommended. nance of the data on which important financial decisions
are being made. Therefore, it is important to keep accu-
Over the past decade, perhaps the most significant rate records regarding all aspects of the specification,
shortcoming of wind speed measurements at prospec- calibration, installation and maintenance of the equip-
tive wind farm sites has been poor sensor mounting ment used.
arrangements. There has been an increasing body of
measured data which has demonstrated that, if the sep- 2.5.4 THE ANNUAL VARIABILITY OF WIND
aration of anemometers from the meteorological mast, SPEED
booms and other sensors is not sufficient, then the wind
speed recorded by the sensor is not the true wind speed “Wind rose” is the term given to the way in which the joint
49
since it is influenced by the presence of the other wind speed and direction distribution is displayed. An
objects. example is given in Figure 2.7. The wind rose can be
thought of as a wheel with spokes spaced, in this example, different levels. This characteristic can be very important
at 30˚ intervals. For each sector the wind is considered and is often inadequately treated. It is the combination of
separately. The length of time that the wind comes from the wind speed distribution and the power curve of the
this sector is shown by the length of the spoke and the proposed turbine which combine together to determine
speed is shown by the thickness of the spoke. the energy production.
The design of a wind farm is sensitive to the shape of the Consider, as an example, two sites, A and B, both with a
wind rose for the site. In some areas, particularly where mean wind speed of 9 m/s. At one extreme, Site A, the
the wind is driven by thermal effects, the wind rose can be wind blows at 9 m/s all the time and the wind farm would
very unidirectional. For example in Palm Springs, USA the be very energetic. At the other extreme, Site B, the wind
wind comes from a sector 10˚ wide for 95% of the year. blows at 4 m/s (below cut-in wind speed for a typical WT)
At this type of site, the WTs will tend to be arranged in for one-third of the time, at 26 m/s (above cut-out wind
tightly packed rows perpendicular to the wind with large speed for most turbines) for one-third of the time and at
spaces downwind. In Northern Europe, the wind, although 9 m/s for one-third of the time. The mean wind speed at
predominantly from the south west, also comes from Site B would be (1/3) x (4 + 9 + 26) = 13 m/s, much high-
other directions for a significant part of the time and er than Site A, but the energy yield at Site B would be only
hence the WTs will tend to be more uniformly spaced in all one-third of that at Site A.
directions.
Although this example is unrealistic, it serves to illustrate
Figure 2.7: Wind Rose a point; that wind speed alone is not adequate to describe
a site’s potential energy yield. Some more realistic site
wind speed distributions are shown in Figure 2.8 where
the actual wind speed distribution is also shown as a
“Weibull fit” to the distribution. The Weibull distribution is
a mathematical expression which has been found to pro-
vide a good approximation to measured wind speed dis-
tributions. This empirical curve fit is therefore used to
characterise a site. Such a distribution is described by
two parameters: the Weibull “scale” parameter which is
closely related to the mean wind speed; and “shape”
which is a measurement of the width of the distribution

parameter. This approach is useful since it allows both
the wind speed and its distribution to be described in a
concise fashion. However, as can be seen from this
figure, care must be taken in using a Weibull fit. It is often
a very good likeness but it can be misleading.
The annual variability in wind rose and wind speed fre-

quency distribution are also important in assessing the
The description above has concentrated on the wind uncertainty in the annual energy production of a wind
speed and wind rose. The other important parameter farm, and are described in detail in a later section of this
which determines the output of a wind farm is the wind chapter. For illustrative purposes, only the variation in
speed distribution. This distribution describes the amount annual mean wind speed is considered, as the other fac-
50
of time on a particular site that the wind speed is between tors usually have a secondary effect.
1
VOLUME
Figure 2.8: Some Example Wind Speed Distributions Figure 2.9a: The Annual Mean Wind Speed Recorded at Malin
Head, Ireland
Table 2.7: Wind Speeds and Energy Production for the Average,
Lowest and Highest Wind Speed Year in the Period 1979 to
1998 Based on a Nominal 10 MW Wind Farm at Malin Head
Annual Mean Percentage of Energy Percentage

Wind Speed Average Year Production of Average
(m/s) (%) (MWh/annum) Year (%)
Lowest wind
speed year 7.77 93.3 29,491 89.8
(1987)
Average year 8.33 100.0 32,847 100.0

Variability of One Year Periods
Highest wind
speed year 9.16 110.0 37,413 113.9
As discussed above, annual wind speed variability has a
(1986)
strong influence on the analysis methodologies developed
for the assessment of the long-term wind resource at a
site and the uncertainty in such predictions. Before Figure 2.9a presents the annual mean wind speed record-
describing some typical methodologies, an example is ed at Malin Head meteorological station over a 20-year
used to illustrate typical levels of annual variability of wind period. It can be seen that there is significant variation in
speed. The example seeks to answer the following ques- the annual mean wind speed, with maximum and mini-
tions: mum values ranging from less than 7.8 m/s to nearly 9.2
• If there is one year of wind data available from a poten- m/s. The standard deviation of annual mean wind speed
tial wind farm site, what error is likely to be associated over the 20-year period is approximately 5% of the mean.
with assuming that such data are representative of the
long term? Table 2.7 presents the average and annual maximum and
• If, instead, there are three years of data available from minimum wind speeds. As an illustration, the equivalent
the site, how does the picture change? annual energy productions for the example 10 MW wind farm
case described above are also presented.
51
Figure 2.9b: Annual Mean Wind Speed at Malin Head over a Table 2.8: Wind Speeds and Energy Production for the
20-Year Period – Three Year Rolling Averages Average, Lowest and Highest Three Year Periods within the
Period 1979 to 1998
Annual Mean Percentage of Energy Percentage

Wind Speed Average Year Production of Average
(m/s) (%) (MWh/annum) Year (%)
Lowest wind
speed year 8.10 97.2 31,540 96.0
(1989)
Average year 8.33 100.0 32,847 100.0
Highest wind
speed year 8.51 102.2 33,871 103.1
(1990)
Table 2.8 illustrates that, if three years of data are avail-

Table 2.7 shows that, had wind speed measurements able from a site, the maximum deviations of the wind
been made on the site for just one year, and this one year speed and energy production over these periods from
had been assumed to be representative of the long term, long-term averages is substantially reduced. The devia-
then the predicted long-term wind speed at the site could tions of 10% and 14% in wind speed and energy for the
have been in error by 10%. It is often the case that little analysis based on one-year data sets reduces to devia-
on-site data is available and hence this situation can tions of 3% and 4% respectively when three-year periods
arise. In terms of energy production it is evident that the are considered.
predicted figure could be in error by some 14% if the
above assumption had been made. For a lower wind While the results presented here are site-specific, they are
speed site, a 10% error in wind speed could easily have a broadly representative of any wind farm in Europe. The reli-
20% effect on energy production owing to its higher sen- ability of long-term data and the consistency of the wind is
sitivity to changes in wind speed at lower wind speeds. central to the commercial appraisal of a wind farm.
Variability of Three-Year Periods Some substantial work has been undertaken (Raftery,
1997) to try and identify the key characteristics of long-
Figure 2.9b illustrates the same data as in Figure 2.9a, term behaviour of the wind. This effort consisted of iden-
but applying to a three-year rolling average. It is immedi- tifying reliable long-term data sets from around the world
ately apparent that the variability in the mean wind speed and attempting to tease out some common characteris-
over the three-year period is substantially reduced com- tics. One of the results of this approach is illustrated in
pared with the one year period. Figure 2.10. Data sets of at least 30 years in duration
have been assembled for each site. The mean of the 30
The results presented in Table 2.7 are reproduced in annual figures was calculated, together with their stan-
Table 2.8, this time based on the highest and lowest dard deviation. The ratio of the standard deviation to the
three-year averages. mean was then calculated and it was found that this var-
ied very little from location to location. This same trend
was observed all over the world, in Australia, Japan and
the US, as well as Europe. This finding is useful in order
to determine the expected variation in long-term wind
52
behaviour.
1
VOLUME
In summary, this work indicates that the annual variability 2.5.5 ANALYTICAL METHODS FOR THE
of long-term mean wind speeds at sites across Europe PREDICTION OF THE LONG-TERM WIND
tends to be similar and can reasonably be characterised REGIME AT A SITE
by a normal distribution with a standard deviation of 6%.
This result plays an important role in assessing uncer- From the above, it is clear that the key element in assess-
tainty in the prediction of wind farm energy production. ing the energy production of a proposed wind farm site is
the prediction of the long-term wind regime at the site
Figure 2.10: Wind Map of Europe – Inter Annual Variation mast or masts. The outcome of the analyses described in
Shown as Standard Deviation as a Percentage of Mean this section is a long-term wind speed distribution togeth-
er with the wind rose. Other meteorological inputs to the
energy production analyses are the long-term site air den-
sity and site turbulence intensity - a measurement of the
“roughness” of the wind - which, while important, are of
secondary influence to the energy production of the wind
farm; their derivation is not therefore considered in detail
here. It should be noted that the turbulence intensity is
very important in determining the loading on a WT and
hence its expected operational lifespan.
Overview
There are essentially two methods which can be used for

the prediction of the long-term wind resource at a site
where on-site measurements are available. These are:
1 Correlate on-site wind data with wind data recorded at
a long-term reference station.
2 Use only on-site wind data.
Unless a long-term data set is already available for a site,

it is desirable to use Method 1 for predicting the long-term
wind resource at a site. Typically, a reliable result can be
obtained with as little as one year of site data. As illus-
trated by the example presented for Malin Head above, if
Method 1 cannot be used and Method 2 is used with only
one year of data, the uncertainty caused by the assump-
tion that the year of data recording is representative of the
long term is substantial.
It is therefore normal practice to find a suitable source of

longer term data in the vicinity of the wind farm site. This
allows a correlation analysis to be undertaken and, if only
relatively short data sets are available from the site itself,
is likely to result in an analysis with a significantly lower
53
uncertainty than that which would result from use of the
on-site data alone. However, before a data set from a
long-term reference station can be used in an analysis, it station is less than half that of the site, prove to be
is vital that thorough checks on its validity are carried out. reliable long-term reference data sets.
• The data are well correlated with those recorded at the
Before discussing the details of this approach it may be site.
helpful to consider the broader picture. It would be ideal if
every site benefited from a long-term data set of, say, 10 Where there have been changes in the consistency at a
years. Now and again this happens, but it is very rare. It reference long-term data source, or where a reliable cor-
is therefore necessary either to use limited on-site data or relation cannot be demonstrated, it is important that the
to try and use other data to gain a longer term view. The use of a prospective source of long-term data is rejected.
correlation approach can be thought of in the following If no suitable reference meteorological station can be
way. Data are gathered on the site using good quality cal- found, then the long-term wind resource can only be
ibrated equipment. These data provide absolute meas- derived from the data recorded at the site itself. It is like-
urements of the wind speed on the site during the meas- ly that longer data sets of two or more years are required
urement period. If it can be established that there is a to achieve similar uncertainty levels to those which would
close relationship (a good correlation) between these site have been obtained had a high quality long-term reference
data and a reference mast, then it will be possible, by data set been available.
using the mast’s long-term reference data, to re-create
the wind speeds on the site. Thus, it is possible to “pre- Experience of wind energy project analysis across Europe
tend” that long-term wind speed records exist for the site. indicates that the density of public sources of high quali-
If a good correlation exists, this is a very powerful tech- ty wind data is greater in northern than in southern
nique but, if the correlation is weak, it can be misleading Europe. This observation, combined with the generally
and hence should be used with caution. more complex terrain in much of southern Europe, often
leads to analyses in southern Europe being based only on
Necessary conditions for an off-site wind data set to be the data recorded at the wind farm site or other nearby
considered as a long-term reference are set out below: wind farm sites. In contrast, for analyses in northern
• The reference data set includes data which overlaps Europe, correlation of site data to data recorded at nation-
with the data recorded on site. al meteorological stations is more common. Clearly, this
• It can be demonstrated that the data have been record- statement is a generalisation and there are numerous
ed using a consistent system over the period of both exceptions to it. However, the establishment of a good
the concurrent and longer term data. This should set of long-term reference masts specifically for wind
include consideration not just of the position and energy use in areas of Europe where wind energy projects
height of the mast and the consistency of equipment are likely to be developed would be an extremely valuable
used, but also potential changes in the exposure of the asset. An EU-wide network of this sort would be highly
mast. For example, the construction of a new building beneficial.
at an airport or the erection of a wind farm near an
existing mast will corrupt the data. The absolute values Correlation Methodologies
recorded at the reference station are not important, but
any changes in either process or the surrounding envi- Some detailed discussion about different correlation tech-
ronment, will render it useless as a reference site. This niques is provided in Appendix D.
investigation is therefore very important and is usually
done by a physical visit to the site, together with an The process of comparing the wind speeds on the site
interview with site staff. with the wind speeds at the reference station and using
• The exposure of the reference station should be good. the comparison to estimate the long-term wind speed on
54
It is rare that data recorded by systems in town cen- the site is called measure correlate predict (MCP). This
tres, or where the mean wind speed at the reference process is also described in some detail in Appendix D.
1
VOLUME
Once the MCP process has been completed, an estimate Energy Production Prediction Methodologies
exists of the long-term wind speed on the site.
Typically, the prediction of the variation in wind speed with
This stage – shown as Milestone 1 on Figure 2.4 - is a height, the variation in wind speed over the site area, and
very important one, since it marks the point at which reli- the wake interaction between WTs are calculated within a
able information on the site’s long-term wind speed bespoke suite of computer programs which are specifi-
behaviour at a single point (or points if there are multiple cally designed to facilitate accurate predictions of wind
masts) becomes available. This estimate will contain both farm energy production. The use of such tools allows the
the mean long-term expected value and the uncertainty energy production of different options of layout, turbine
associated with that value. So far, however, we know type and hub height to be established rapidly once mod-
nothing of the distribution of the wind speed across the els have been set up. Such programs are commonly
site and neither have we considered the way in which the termed wind farm design tools (WFDT).
energy values can be converted into energy.
Within WFDTs, site wind flow calculations are commonly
2.5.6 THE PREDICTION OF THE ENERGY undertaken using the WAsP model (Troen and Petersen,
PRODUCTION OF A WIND FARM 1989) which has been widely used within the industry over
the past decade. Other commercial models which are phys-
In order to predict the energy production of the wind farm ically similar to WAsP are also sometimes used. This area
it is necessary to undertake the following tasks: of the wind farm energy calculation is in need of the great-
• Predict the variation in the long-term wind speed over est level of fundamental research and development. Flow
the site at the hub height of the machines based on models for use in commercial wind farm development have
the long-term wind speeds at the mast locations. to be quick to execute and be both reliable and consistent.
• Predict the wake losses which arise as a result of one At present, the industry opts for simple but effective tools.
turbine operating in front of another. More rigorous and much more time consuming tools may
• Calculate or estimate other losses. be adopted in the future if they can be made cost effective
and not too sensitive to modelling assumptions.
Information Required for an Analysis
The challenge is to take a topographical map and, for any
In addition to the wind data described in the earlier sec- point on it, determine the long-term wind rose. This informa-
tions, inputs to this process are typically as follows: tion is then used to calculate the long-term wind speed at all
• Wind farm layout and hub height. the points on the map where it is intended to place WTs.
• Turbine characteristics, including power curve (the
curve which plots the power output of a turbine as a The WAsP model does have shortcomings under certain
function of the wind speed) and thrust curve (the equiv- topographical and flow conditions. It needs to be used
alent curve of the force applied by the wind at the top with care and experience and not as a “black box”. In par-
of the tower as a function of wind speed). ticular, it does not include “viscous effects” which cause
• Predicted long-term site air density and turbulence the wind to “separate” as it flows over a sharp change in
intensity (the turbulence intensity is the “roughness” of topography. The WAsP model will follow the terrain where-
the wind). as the real wind will behave as shown in Figure 2.11.
• Definition of the topography over the site and
surrounding area. There is an enormous amount of work in progress in all
• Definition of the surface ground cover over the site and aspects of engineering, quite distinct from the wind energy
surrounding area. industry in which developments are being made in the
55
Figure 2.11: An Example of Flow Separation over a Hill It is important to appreciate that, as the distance of the
turbine from the meteorological mast increases, the
uncertainty in the prediction also increases. This increase
in uncertainty is typically more rapid in complex than in
simple terrain. Experience of the decrease in accuracy
with distance from the mast when using models such as
WAsP are inherent in making recommendations regarding
the appropriate number of meteorological masts for a
wind farm site, as discussed above. WFDTs also allow
numerical prediction of complicated flows. Most notably, environmental constraints to be included – areas of the
these efforts centre on aerospace problems – accurately site which may not be used because of noise, visual intru-
predicting flow over aircraft wings and fuselage for example, sion, shadow flicker, land ownership, presence of pro-
or predicting internal flows in turbo-machinery. Efforts are tected flora and fauna, etc. These considerations are dis-
now being made to apply such models to the arbitrary ter- cussed in more detail below.
rain which defines a wind farm. There is still a long way to
go before these models can be considered either reliable or The ability of WFDTs to provide an integrated model of a
commercially feasible for wind farm design. The energy esti- wind farm also allows them to be used to optimise wind
mate is only as good as its weakest link and hence its accu- farm design. This task is performed automatically. The tool
racy is largely defined by this step – the topographical wind positions the turbines to achieve the best possible balance
model. Data sets now exist which can be used for the vali- between exploiting maximum wind speed and minimising
dation of new codes and further developments are expect- wake losses. This process can be successfully undertaken
ed. The task is, nevertheless, a demanding one and accu- at the same time as observing the environmental and own-
rate calculation of wind flow over steep terrain is still some ership constraints. The emergence of WFDTs has been a
way off. At present it is necessary to use a mixture of com- significant development in wind farm design. The success-
putation and human insight. ful completion of an energy calculation using a WFDT may
be considered as Milestone 2 in Figure 2.4.
Once the topographical effects on the flow have been
computed, it is necessary to determine how the individual Wind Farm Energy Loss Factors
turbines affect one another (the wake effects). If a turbine
is working downstream of another (i.e. in its wake) then When WFDT calculations have been undertaken, potential
the turbine will come into contact with less wind than it sources of energy loss must be considered on a site-spe-
would if it were in the free stream. For some types of wind cific basis. Specific sources of energy loss other than
rose and wind farm design this effect needs to be care- those associated with wake effects are described below.
fully calculated as it can be significant – of the order of
15% in lost energy. The models which estimate this loss Turbine Availability
are knows as “wake models”. Different complexities of
wake model are used in the various commercially avail- A figure is assumed for turbine availability based on data
able WFDTs. These tools are now well validated for a vari- from modern operational wind farms. However, availabili-
ety of different types of wind farm layout. However, it is ty may be a matter of warranty between the owner and the
well known that they do not work well for very tightly turbine supplier and the assumed figure should be
packed wind farms, such as those described above at reviewed when the terms of that warranty are clear. In
Palm Springs, and further fundamental work is required to some circumstances it is appropriate to consider a time
improve modelling in this area. varying value of availability.
56
1
VOLUME
Electrical Transmission Efficiency Utility Down Time
The electrical transmission efficiency from the terminals at The wind farm will be unable to export energy if the grid is
the base of the turbine tower to the wind farm metering point not available. This needs to be considered on a site-spe-
will depend on the site’s detailed electrical design. A formal cific basis with suitable historic information supplied by
calculation of the electrical loss should be undertaken when the grid operator.
the electrical system has been defined. Some of the WFDTs
described above may be used for such calculations. Power Curve Adjustment
Turbine Icing and Blade Degradation Power curve adjustment to the energy prediction accounts
for variations in the actual turbine performance in com-
This factor reflects losses associated with icing of the tur- parison to the supplied power curve. An example of such
bine, temporary “fouling” of the turbine blades by dirt or an adjustment is provided in Appendix C.
insects and long-term degradation of the blade. It needs
to be considered on a site-specific basis. Columnar Control Loss
High Wind Hysteresis If turbine spacing is close, the site conditions may exceed
the wind conditions within the WT certification criteria. In
This is caused by the turbine cut-in and cut-out control cri- these circumstances it may be necessary to shut down
teria for high wind speeds. The magnitude of the loss is turbines which are closely spaced when the wind direction
influenced by three factors: is parallel to the line of turbines. The turbine supplier
1 The turbine will cut-out when the mean wind speed should indicate whether such a strategy is required at the
exceeds a maximum level and it will not cut-in again time of a tender.
until the mean wind speed is lower than the cut-out
mean wind speed. These losses can combine to a significant total – often
2 The turbine will cut-out if the instantaneous gust wind between 9% and 12% of the gross energy yield. Applying
speed exceeds a maximum level and the turbine will these additional losses to the gross energy is represented
not cut-in until the wind speed drops. by Milestone 3 in Figure 2.4.
3 The accuracy of the calibration of the instruments that
determine the wind characteristics at the turbine. 2.5.7 DEFINITION OF UNCERTAINTY IN
PREDICTED ENERGY PRODUCTION
These three effects will cause the turbine to lose production
for some proportion of high mean wind speed occurrences. The uncertainty analysis is an important part of any
This is clearly a site-specific issue which is more significant assessment of the long-term energy production of a wind
for high wind speed sites than low wind speed sites. farm. Although an uncertainty analysis needs to be con-
sidered on a site-specific basis, the process can be
Substation Maintenance shown as follows:
The wind farm substation will require maintenance which • Identify the different inputs to and processes within the
is likely to be at a time which is outside the control of the analysis.
wind farm owner and, therefore, a small loss may be expe- • Assign an uncertainty to each of these elements both
rienced. in terms of the magnitude of the uncertainty and the
shape of the distribution.
57
• Convert each of the uncertainties into common units of
energy.
• Combine the various uncertainties to define a total 2.6 Offshore Wind Farm Design and
uncertainty for the entire prediction. Resource Estimation
• Present uncertainty statistics at requested levels.
This section describes the differences in wind flow moni-
Research work reported by Raftery et al. (1999) defined a toring and data analysis offshore compared with onshore.
comprehensive risk register for wind power projects and
included detailed Monte Carlo based analysis techniques 2.6.1 FUNDAMENTALS
to assess the uncertainty in the results obtained. Based
on the results of this work, use of an uncertainty analysis Onshore, topographic effects are one of the main driving
with a number of simplifying assumptions can be justified. forces of the wind regime. With no topographic effects off-
The main simplifying assumptions are that it is reason- shore, other factors dominate wind behaviour.
able to consider a relatively small number of key uncer-
tainties and that these individual uncertainties can be The surface roughness (a parameter used to describe the
assumed to be normally distributed. Making these roughness of the surface of the ground) is low, which
assumptions, it is possible to define energy production results in a steeper boundary layer profile. The different
levels with a defined probability of exceedance. values are illustrated in Table 2.9. Offshore, the surface
roughness length is typically assumed to be 0.001 m or
It is common to present uncertainty results for both a long less. This assumption is reasonable for relatively calm
future period of, say, 10 years and also for a shorter peri- weather, but it does not take into consideration the effect
od of one year. It is now normal practice for banks and of wind speed on wave size. However, calculating this
financial institutions to be presented with such figures, in parameter is complex as the sea surface does not pres-
parallel with central energy production estimates, when ent fixed roughness elements in the same way as trees,
wind farm financing is being arranged. hills and buildings. Low surface roughness also results in
low turbulence intensity.
The uncertainty analyses presented within energy assess-
ments typically assume that the turbines will perform Table 2.9: Typical Values for zo
exactly to the defined availability and power performance
Type of Terrain zo(m) α
levels. The power performance and availability levels are
Mud flats, ice 0.00001
usually covered by specific warranty arrangements and
Smooth sea 0.0001
hence any consideration of the uncertainty in these
Sand 0.0003 0.10

parameters needs machine-specific and contract-specific
Snow surface 0.001
review which is generally outside the scope of a “stan-
Bare soil 0.005 0.13
dard” energy analysis.
Low grass, steppe 0.01
Fallow field 0.03

Uncertainty in the energy estimates is a vital part of the
Open farmland 0.05 0.19
result.
Shelter belts 0.3
Forest and woodland 0.5
Suburb 0.8
City 1 0.32
The coastal zone, where the properties of the boundary

layer will be changing, extends away from the shore for
58
varying distances, and this can result in variations in wind
speed and boundary layer profiles across the wind farm.
1
VOLUME
2.6.2 MEASUREMENT OFFSHORE 2.6.4 ENERGY PREDICTION
Turbines for offshore wind farms are larger than those This step is essentially the same as for onshore predic-
onshore. Their size presents several issues including the tions. There is generally only minor predicted variation in
need to understand the characteristics of the boundary wind speed over a site. For large offshore sites, wake
layer up to and above heights of 150 m. Measurements losses are likely to be higher than for many onshore wind
offshore are expensive. A typical mast will cost some farms. Such losses are increased due to the lower ambi-
€750,000, some 50 times that required for equivalent ent turbulence levels since offshore wind is much
onshore work. Monitoring towers offshore are un-guyed smoother. There is, therefore, less mixing of the air
and therefore need to be wider, which can mean that behind the turbine, which results in a slower re-energising
measurements is more susceptible to wind flow effects of the slow moving air, meaning that the wake lasts
from the tower. longer. Recent research has, for the first time, validated
wake modelling techniques offshore (Risø National
If monitoring equipment is not available, there are other Laboratory, 2002).
sources of information which can be used to determine
the approximate long-term wind regime at the offshore Offshore machines are likely to experience more down
location. For example, there are some offshore databas- time than those onshore, due to difficulties associated
es for wind data including light vessels and observation with access. If a turbine has shut down and needs main-
platforms. None of these is suitable for a bankable report, tenance work, access to it may be delayed until there is a
however. suitable window in the weather. This aspect of offshore
wind energy is likely to be the most important element in
2.6.3 WIND ANALYSIS OFFSHORE determining real cost.
Depending on the amount of data available, different ana- 2.6.5 OTHER EFFECTS TO CONSIDER
lytical methods can be employed. A feasibility study can OFFSHORE
be carried out based on available wind data in that area.
WAsP can be used from coastal meteorological stations Tidal rise and fall effectively shifts the location of the tur-
to give a prediction offshore, aided by its latest tool, the bine in the boundary layer. Over a 12-hour period, this can
coastal discontinuity model (CDM) (Bartholemie, 2003). cause variation in mean wind speed and also impact on
Existing offshore measurements can also be used. There the shear across the turbine rotor itself. Taking the UK as
are problems associated with using long distance model- an example, tidal heights vary significantly, with
ling, especially around the coast, due to the differences in Avonmouth having mean spring tides of 12.2 m in height,
predominant driving forces between onshore and offshore the largest range in the UK.
breezes, and variation in the coastal zone in between.
Temperature driven flows due to thermal inertia of the sea
For a more detailed analysis, measurements offshore at initiate localised winds around the coastal area. The sea
the site are necessary. MCP from a mast offshore to an takes longer to heat up and cool down than the land. During
onshore reference station can be used. With several the day, as the land heats up, the warmer air rises and is
measurement heights more accurate modelling of the replaced by cooler air from over the sea. This creates an
boundary layer will help extrapolate to heights above the onshore wind. The reverse effect can happen during the
monitoring mast. Given the absence of topography, off- night, resulting in an offshore wind. The strength and direc-
shore measurements from such a mast can be consid- tion of the resulting wind is influenced by the existing gra-
ered representative of a much larger area than would be dient wind which in some situations may be cancelled out
59
possible onshore. by the sea breeze, leaving an area of no wind.
2.7 Forecasting for all the different energy forms to be considered on an
equitable basis. Proper, formal statistical analysis of both
So far, this chapter has only considered the wind indus- renewable and conventional plant is therefore necessary.
try’s ability to estimate long-term energy production from This task should be considered as an essential element of
a wind farm. Usually this is the most important task since, a wind energy development strategy.
to date, most of the power purchase agreements are
“take or pay”, meaning that the utility or other customer As a result of its strategic importance, forecasting has
is obliged to buy all the energy produced by the wind farm. been the focus of considerable technical attention in
As the penetration of wind power generation in the overall recent years. A good source of general review materials,
energy mix increases, it will cause the fluctuations in as well as detailed papers, can be found in Landberg et
energy output caused by variations in wind speed to be al. (2003). Although there is a variety of different tech-
more visible. The independent system operators (ISOs) niques being used, they all share similar characteristics.
working to balance supply and demand on regional or It is therefore possible to provide a generic description of
national grid systems will need to predict and manage this existing techniques, whereby data is provided by a weath-
variability to avoid balancing problems. The point at which er forecast and production data is provided by the wind
this is required changes from system to system. farms. The two sets of data are combined together to
provide a forecast for future energy production.
As the level of penetration of wind energy into individual
grids increases it will be necessary to forecast over short To integrate wind energy successfully into an electricity
to medium time scales (one hour to two days) how much system at large penetration levels, it will be necessary to
energy will be produced. In some countries, forecasting is predict wind energy production as accurately as possible.
already required. New wind farms in California are
required to “use best possible means available” to fore- The numerical weather predictions (NWP) models run by
cast the output and send such estimates to the California national institutes are typically of continental, if not glob-
ISO. In European countries where there is already a high al, scale. Consequently, their resolutions tend to be too
level of penetration - Spain, Germany and Denmark - oper- coarse for wind energy needs. For example, the model
ators and managers are routinely forecasting output from run by the UK Meteorological Office for north-west
their wind farms. These forecasts are used to schedule Europe has a minimum horizontal resolution of approxi-
the operations of other plant, but are also used for trad- mately 12 km.
ing purposes.
The methods of achieving the transformation between

Forecasting wind energy production will increase in impor- coarse NWP forecasts and site-specific ones are varied.
tance as the level of installed capacity grows. The wind Despite this variation, they can largely be grouped into
industry must allow ISOs to use wind energy to its best two main types - physical and statistical models.
effect, by forecasting output from wind farms as accu-
rately as possible. In the UK, where the market is already Physical models primarily aim to improve the resolution of
deregulated, energy traders are using crude forecasts to the “original” NWP model. The models used to achieve
trade wind energy on the futures market. this can include:
• Simple linear-flow models, such as WAsP.
At the same time as improving the predictability of the • Fine resolution NWP models. These are essentially local
output of wind energy plant through improvements in fore- (nested) meso-scale versions of the original NWP model
casting techniques, awareness of the true behaviour of and are often termed “storm-scale” or “convective-
conventional plant should be considered. In order to provide scale”. They aim to model local thermal and terrain
60
the best mix of plant and technologies it will be important effects that are not apparent at the coarse-scale.
1
VOLUME
Figure 2.12: A Schematic Representation of a Forecasting Figure 2.13: Method Overview
Approach
NWP
Forecasts
Site-specific
Models
Site-specific
Forecasts
Power Models
Power
There are aspects of the physical model approach which
Forecasts
need to be considered:
• Skill: The implementation of local meso-scale models
requires the skill and competency of a meteorologist.
There is always the possibility of a poorly formed model forecasts (dotted line) to represent what is actually hap-
introducing further errors. pening at the site. The example shown is for a T+12 hour
• Computational requirements: The formulation and exe- (12 hours ahead) forecast horizon, for a meteorological
cution of the models is computationally expensive. mast on a wind farm situated in complex terrain.
2.7.1 OVERVIEW OF THE METHOD Figure 2.14: Example Time Series of Wind Speed Forecast, T+12h
There are several groups working in this area and they all
have slightly different approaches (see, for example,
Landberg et al., 2003). All, however, create power output
forecasts through a two-stage process. First, there is the
creation of site-specific meteorological forecasts (for some
pre-defined reference point, such as a site meteorological
mast). These meteorological forecasts are then trans-
formed, via site-specific power models, to power output fore-
casts. This process is shown schematically in Figure 2.13.
To enable the meteorological model to be both auto-regres-

sive and adaptive, feedback data from the site is also
required. In other words, the method needs to know what is
happening at the site where the predictions are being made The time series starts on 19 November with the site
so it can “learn” and adjust the forecasts accordingly. An model having been initialised on 1 November. Therefore,
61
example time series plot is shown in Figure 2.14. This the model has adapted to this accurate transformation in
shows how well the model transforms the initialising NWP less than three weeks.
2.7.2 IMPROVEMENT OVER PERSISTENCE 2.7.3 POWER OUTPUT
Persistence is the rather grand name given to the crudest The next stage of the process is to convert the meteoro-
of forecasting techniques: the wind speed will stay the logical forecasts to forecasts of power output. This trans-
same as it is now! For short time scales, as common formation is typically achieved via a wind farm power
sense would suggest, it works well. It is often used as a matrix, using multiple direction and wind speed bins to
standard yardstick against which to evaluate other more represent the power output of the wind farm. It should be
sophisticated forecasting techniques. The improvement in stressed that the method of producing this power matrix
uncertainty (based on reduction of standard deviation of is crucial if this stage is not to introduce further uncer-
errors), over the basic persistence method, is presented tainty into the forecasts.
in Figure 2.15 for the same site. The evaluation was
undertaken over a total period of two months (November The meteorological forecasts shown previously were con-
and December). Three lines are shown: verted to power using a less than optimal power matrix.
• Overall (line only). The resulting time series is shown in Figure 2.16.
• Low wind speeds (line with triangle).
• High wind speeds (line with star). Figure 2.16: Time Series of Power Forecast, T+12h
Typically, the evaluation of forecasting methods is pre-

sented as the “overall” case. However, it is clear that the
spread of improvement is not uniform across all wind
speed ranges. In this particular example, the crucial low
(cut-in) and high (cut-out) wind speed forecasts show sig-
nificantly greater improvements than the overall case. The
cut-out prediction is particularly important since for a
small increment in wind speed, say from 24.5 m/s to 25
m/s, the whole plant will shut down. These performance
differences can have significant effects on the value of
the forecasting tool, depending on its specific application.
Figure 2.15: Improvement over Persistence
To an “engineering eye” this prediction looks good; the

forecast has captured the shape of the profile rather
well. The maximum power level could certainly be
improved, but that is a matter of fine tuning. To the “com-
mercial eye”, however, the situation is not as good. The
fact that the phase of the forecast and the actual power
are different is crucial. For example, on 20 November,
the steep rise in power is lagged by the prediction so
although the shape is good the absolute error on an
hourly basis will be large. For hourly trading purposes
such a prediction would then be very poor. Whether or not
the forecast is “good” or “bad” therefore depends very
62
strongly on its precise purpose. For scheduling plant
maintenance it is acceptable, whereas for hourly trading
1
VOLUME
it is poor. The purpose of the forecast needs to be very Europe. It will be necessary to determine the scale at
carefully defined. This is a strategic as well as a techni- which this work is required - local, national or European.
cal issue. Different techniques will be needed for different scales.
The work is now well established but certainly has
2.7.4 PORTFOLIO EFFECTS greater potential than is presently realised.
Improvements in forecasting will be derived from within
As the geographical spread of an electricity system the wind energy industry and also from the great strides
increases, the wind speeds across it become less corre- being made in numerical weather forecasting generally.
lated. Some areas will be windy, some will not. Some
areas will have rising power output, others will have falling
power output, etc. The effect of aggregation on wind farm 2.8 Future Developments
power fluctuations, and on capacity credit issues, has
been looked at by several analysts, and forecasting of the Wind speed and energy prediction is, and will remain, the
output of large numbers of wind farms for the ISO is now most critical part of the development of a wind farm.
commonplace in both Denmark and Germany. Enormous investments are made based on the estimates
provided. The confidence of lenders and investors must
For certain sizes of weather system, the effect of aggre- be maintained or boosted. Improvement in these tech-
gation will certainly be to smooth the output of the wind niques is therefore an important part of European wind
farms as a whole. This effect is powerful and can be used, energy development. Below is a list of important topics for
subject to further analysis and validation, to dispel the future development:
argument that 100 MW of wind requires 100 MW of con- • Continued improvement in the quality and quantity of
ventional generation to back it up. The level of backup wind data recorded at wind farm sites.
needed will depend on the geographical disparity of the • Consideration should be given to the establishment of
wind farms, the size of the weather systems and the size a network of reference masts in the new member
of the interconnected system. This matter is an important states to help “kick start” their wind energy activity.
topic for further work which will allow the robustness of • Some more sophisticated flow modelling tools are
the European wind resource to be properly quantified. starting to be used for the prediction of wind flow over
wind farm sites. While further validation work and
2.7.5 CONCLUSIONS development is required before such models can be
extensively used for the assessment of wind farm proj-
The present status of forecasting techniques may be sum- ects, it is considered that such models have the poten-
marised as: tial to significantly improve the modelling of the flow at
• The statistical meteorological model has proven adapt- wind farm sites, particularly in complex terrain. Initially,
able, transferable and accurate. Improvements in these models may be useful for investigating sites in
uncertainty (over the basic persistence method) of up complex terrain as well as wind farms located close to
to 70% can be forecast 12 hours ahead and beyond. mature forestry. Application of existing CFD tools to the
• Knowledge, interpretation and mitigation of forecast complex topography of wind farm sites would be a
uncertainties are all of primary importance when it rewarding activity if it could be proved that such tools
comes to the application of a forecast. This is closely can be used both efficiently and reliably. This is an
linked with the nature of the market in which the fore- area where future endeavour would be well worthwhile.
casts are being used. One very effective way of reduc- It offers the possibility of applying sophisticated tools
ing the overall forecast uncertainty is to aggregate the developed elsewhere (e.g. the aircraft industry) to the
forecasts from several wind farms. wind business.
63
• Improving forecasting techniques will greatly enhance • Remote sensing techniques look promising for making
the commercial and strategic value of wind energy in measurements at hub height of large machines and
have possible applications offshore. If they could be
made both robust and reliable, and their validity
demonstrated by working in parallel with conventional
means, then their application would be highly benefi-
cial. The development of accurate Sodar or Lidar tech-
niques have the potential to improve the quality of
measured wind data at a wind farm site and may also
be useful in the future for undertaking power perform-
ance tests. To date, however, it is not considered that
these methods are sufficiently accurate to replace con-
ventional measurements. Again this topic would prove
a fruitful R&D exercise.
• The wind speed on a potential wind farm site will
remain by far the most important parameter to deter-
mine the viability of the development. It is the key
parameter for lenders and they are becoming increas-
ingly sophisticated in their demands in the analysis of
the uncertainty often estimates as well as the long
term consistency. The effect of climate change on wind
speed has not been covered in this work. It is, howev-
er, becoming a common question for lenders and
should be addressed.
• The improvement of forecasting techniques is vital to
allow wind to compare with conventional plant.
Significant investment in fine-tuning these tools would
return a good reward and help wind realise its full
potential in Europe.
• A unified wind resource map for the new member
states combined with the EU-15 does not exist and
would be a very useful strategic tool.
64
3 WIND FARM DESIGN
1
VOLUME
3.1 Introduction for a project it is necessary to install anemometry equip-
ment at the site. The preliminary layout allows the wind
Previous chapters have discussed the turbines and the measurements to be made in appropriate locations. The
wind resource. This chapter presents a brief summary of preliminary layout also allows more detailed discussions
the design of a wind farm as a whole. to be held with relevant parties to better define the con-
straints.
3.2 Preliminary Layout Design

3.3 Detailed Layout Design
Once a site has been identified and the decision has been
taken to invest in its development, the wind farm design A key element of the layout design is the minimum turbine
procedure commences. This is inevitably an iterative spacing used. In order to ensure that the turbines are not
process. The first task is to define the constraints on the being used outside their design conditions, the minimum
development: acceptable turbine spacing should be obtained from the
• Maximum installed capacity (due to grid connection or turbine supplier and adhered to. The appropriate spacing
power purchase agreement terms). for turbines is strongly dependent on the nature of the ter-
• Site boundary. rain and the wind rose at a site. If turbines are spaced
• Set backs from roads, dwellings, overhead lines, closer than five rotor diameters (5D) in a frequent wind
ownership boundaries, etc. direction, it is likely that unacceptably high wake losses
• Environmental constraints. will result. For areas with predominantly uni-directional
• Location of noise sensitive dwellings if any and wind roses, such as the San Gorgonio Pass in California,
assessment criteria. or bi-directional wind roses such as Galicia in Spain,
• Location of visually sensitive viewpoints if any and greater distances between turbines in the prevailing wind
assessment criteria. direction and tighter spacing perpendicular to the prevail-
• Turbine minimum spacings as defined by the turbine ing wind direction will prove to be more productive. Tight
supplier. spacing requires approval by the turbine supplier if war-
• Constraints associated with communications signals ranty arrangements are not to be affected.
such as microwave link corridors, if any.
With the wind farm constraints defined, the layout of the
These constraints may change as discussions and nego- wind farm can be optimised. This process is also called
tiations with various parties progress. wind farm “micrositing”. The aim of such a process is to
maximise the energy production of the wind farm whilst
When an idea of the likely constraints is known, a prelimi- minimising the infrastructure and operating costs. For
nary design of the wind farm can be produced. This will most projects the economics are substantially more sen-
allow the size of the development to be established. For sitive to changes in energy production than infrastructure
the purpose of defining the preliminary layout it is neces- costs. It is therefore appropriate to use the energy pro-
sary to define approximately what sizes of turbine are duction as the dominant layout design parameter.
under consideration for the development, as the installed
capacity achievable with different sizes of turbine may vary The detailed design of the wind farm is facilitated by the
significantly. The selection of a specific turbine model is use of commercially available wind farm design tools
often best left to the more detailed design phase when the (WFDTs). Once an appropriate analysis of the wind regime
commercial terms of the various suppliers are known. at the site has been undertaken, a model is set up which
can be used to design the layout, predict the energy pro-
65
The wind resource at the site is the key parameter in duction of the wind farm, and address economic and
determining its economic viability. To assess the energy planning related issues.
For large wind farms, it is often difficult to manually derive For a typical onshore wind farm, the cost of the turbines
the most productive layout. For such sites a computa- is approximately 75% of the total cost of the farm. This
tional optimisation using a WFDT may identify a layout for infrastructure is often called the “balance of plant”.
which substantial gains in predicted energy production are There are three essential elements of the balance of
achieved. Even a 1% gain in energy production from plant: the foundations; the electrical grid; and the super-
improved micrositing is worthwhile. The computational visory, control and data acquisition (SCADA) system which
optimisation process will usually involve many thousands links all the turbines to a central computer and acts as
of iterations and can include noise and visual constraints. the wind farm’s “nerve centre”. A typical cost breakdown
WFDTs conveniently allow many permutations of wind is shown in Figure 3.1.
farm size, turbine type, hub height and layout to be con-
sidered quickly and efficiently, so increasing the likelihood Figure 3.1: Typical Cost Breakdown for an Onshore Wind Farm
that an optimal project results. Financial models may be
linked to the tool so that returns from different options
can be directly calculated, further streamlining the devel-
opment decision-making process.
In many countries the visual influence of a wind farm on the

landscape is an important issue. The use of computation-
al design tools allows the zone of visual influence (ZVI), or
visibility footprint, to be calculated to identify from where
the wind farm will be visible. The tools may also be used
to provide visualisations, to facilitate the production of pho-
tomontages and to predict the noise and shadow flicker
from a proposed development. These are often key
aspects of the project’s environmental impact assessment.
In conclusion, the design of a wind farm is a compromise 3.4.1 CIVIL WORKS

between high energy, easy access, easy permitting and
commercial viability. The foundations must be adequate to support the turbine
under extreme loads. Normally, the design load condition
for the foundations is the extreme, once-in-50-year wind

3.4 The Infrastructure speed. In Europe, this wind speed is characterised by a
three-second gust which would probably lie between 45
The energy output of a wind farm is the key determinant and 70 m/s. At the lower end of this range it is likely that
for success. Without adequate wind resource a wind farm the maximum operational loads will be higher than the
cannot be economic. The previous chapters in this loads generated by the extreme gust and would therefore
volume have described the process of wind assessment govern the foundation design. The first step towards the
in great detail. It is clear that a reliable, economic wind proper design of the foundations is the specification of a
turbine (WT) must be used – a topic which has also been load. The turbine supplier would normally provide a com-
discussed in detail. In addition to these basic principles, plete specification of the foundation loads as part of a
it is important to devote adequate attention to the infra- tender package.
structure needed to support the turbines and to extract
the energy and transmit it to the grid. Once the specification has been prepared in detail, design
66
of the foundation structure can be undertaken. Although
extremely important, this process is a relatively simple
1
VOLUME
civil engineering task. A typical foundation would be, per- The design requirements for the internal grid will be in two
haps, 13m across a hexagonal form and one to two parts: the losses must be kept to a minimum (usually less
metres deep. It would be made from reinforced concrete than 2.5% of annual energy); and the design must allow
cast into an excavated hole. The construction time for the turbines to connect safely to the utility grid and satis-
such a foundation, from beginning to end, is normally less fy both the local grid requirements usually in the form of
than a week. a “grid code” and also the turbine specifications.
3.4.2 ELECTRICAL WORKS The most cost-effective way in which to develop a wind farm
is to find a location which is close to the grid. For some large
The turbine generator voltage is normally classed as developments, however, it may be necessary to build an
“low” and is often 690 V although some more large interconnecting line. Such connections are very expensive.
modern turbines generate at 10-12 kV. For the vast
majority of onshore wind farms, the low voltage output of 3.4.3 SCADA AND INSTRUMENTS
the turbine generator is connected to a pad mount trans-
former which steps the voltage up to a level used by the In addition to the essential equipment needed for a func-
internal grid – usually between 10 and 20 kV. The trans- tioning wind farm - the turbines and associated balance of
former is either mounted on a plinth beside the turbine plant - it is also advisable, if the project size can warrant the
foundation or, for bigger turbines, is contained within the investment, to erect permanent anemometry. This equip-
base of the tower. The individual transformers are then ment allows the performance of the wind farm to be care-
connected to underground cables in an internal grid fully monitored and understood. If the wind farm is not per-
which takes the power to a substation or interconnector. forming to budget, it is important to find out whether this is
A typical layout is shown in Figure 3.2. The substation due to poor mechanical performance or less than expected
usually contains another transformer which steps the wind resource. Without good quality wind data, it is not pos-
voltage up from the internal grid level to the distribution sible to make this determination. Large wind farms there-
or transmission level. This final level will depend on the fore usually contain a permanent meteorological mast
local utility grid. It can be anywhere in the range from 10 which is installed at the same time as the turbines.
kV upwards; a typical level would be 20 to 50 kV. The
metering for the wind farm will usually be located at the A vital element of the wind farm is the SCADA system which
substation. It can be at the medium or at the high volt- connects the individual turbines, the substation and mete-
age level. orological stations to a central computer. This computer
and the associated communication system allow the oper-
Figure 3.2: A Typical Electrical Layout ator to supervise the behaviour of all the WTs and also the
wind farm as a whole. It records activity every 10 minutes
and allows the operator to determine what corrective
action, if any, to take. It also records energy output, avail-
ability and error signals which will act as the basis for any
warranty calculations and claims.
3.5 Construction
A wind farm may be a single turbine or it may be a large

number – possibly many hundreds. The design
67
approach and the construction method will, however, be
almost identical whatever the size. The record of the
wind industry in the construction of wind farms is very 3.7 Commissioning and Operation
good. Few wind farms are delivered either late or over
budget. Once construction is complete, commissioning will begin.
Commissioning of an individual turbine takes little more
Newcomers to the wind industry tend to think of a wind than two days. The long-term availability of a commercial
farm as a power station. There are, however, some wind turbine is usually in excess of 97%. This means
important differences between these two types of power that, for 97% of the time, the turbine will be available to
generation. A conventional power station is one large work if there is adequate wind. This value is superior to
machine which will not generate power until it is com- values quoted for conventional stations. It will usually
plete. It will often need a substantial and complicated take a period of some six months for the wind farm to
civil structure, and construction risk will be an important reach full commercial operation; during that period avail-
part of the project assessment. New lenders will, there- ability will increase from around 90% immediately after
fore, always wish to make a careful assessment of the commissioning to the long-term level of 97% or more.
construction risk. However, the construction of a wind
farm is more akin to the purchase of a fleet of trucks Commissioning tests will usually involve standard tests of
than it is to the construction of a power station. The tur- the electrical infrastructure and the turbine, and inspec-
bines will be purchased at a fixed cost agreed in tion of routine civil engineering quality records. Careful
advance and a delivery schedule will be established. In testing at this stage is vital if a good quality wind farm is
a similar way, the electrical infrastructure can be speci- to be delivered and maintained.
fied well in advance - again probably at a fixed price.
There may be some variable costs associated with the It is normal practice for the supplier of the wind farm to
civil works, but this cost variation will be very small com- provide a warranty for between two and five years. This
pared to the cost of the project as a whole. The con- will often cover lost revenue, including downtime to cor-
struction time is very short compared to conventional rect faults and a test of the power curve of the turbine. If
power; a 10 MW wind farm can easily be built within a the power curve is found to be defective then reimburse-
couple of months. ment will be made through the payment of liquidated dam-
ages. For modern wind farms there is rarely any problem
in meeting the warranted power curves, but availability,
3.6 Costs particularly for new models, can be lower than expected in
the early years of operation. During the first year of tur-
Wind farm costs are largely determined by two factors: the bine operation some “teething” problems are usually
complexity of the site and the likely extreme loads. The experienced. For a new model this effect is more marked.
site may be considered complex if the ground conditions As use of the model increases, these problems are
are difficult – hard rocks or boggy, for example - or if resolved and the availability rises.
access is a problem. A very windy site with high extreme
loads will result in a more expensive civil infrastructure as After commissioning, the wind farm will be turned over to
well as higher turbine specification. Typical installation the operations and maintenance (O&M) crew. A typical
costs for 2001 are reported in Volume 2 as between 900- crew will be two people for every 20 to 30 WTs installed.
1,150 €/kW. In 2003, the installed cost for a large wind For smaller wind farms there may not be a dedicated O&M
farm was between €850 and €1,100 per kilowatt crew but arrangements will be made for regular visits from
installed. a regional team. Typical routine maintenance time for a
modern WT is 40 hours per year. Non-routine mainte-
nance may be of a similar order.
68
1
VOLUME
There is now much commercial experience with modern
WTs and high levels of availability are regularly achieved.
Third party operations companies are well-established in
all of the major markets and it is likely that this industry
subsector will develop very much along the lines associat-
ed with other rotating plant and mechanical/electrical
equipment.
The building permits obtained in order to allow the con-

struction of the wind farm may have some ongoing envi-
ronmental reporting requirements, for example the moni-
toring of noise, avian activity or other flora or fauna inter-
est. Similarly, there may be, depending on local bye-laws,
regulatory duties to perform in connection with the local
utility. Therefore, in addition to the obvious O&M activity,
there is often a management role to perform in parallel.
Many wind farms are the subject of project finance and
hence will also be reporting to lenders.
3.8 Concluding Remarks
This chapter has briefly outlined the physical activities

required to design, build and operate a wind farm and has
shown how the principles described in previous chapters
are put into practice. It has been demonstrated that a
wind farm has different challenges and demands from a
conventional power station. It is quick to build and oper-
ates using free fuel. Its operational demands are simple
but must be thoroughly exercised in order to obtain the
long life required for economic success.
69
4 TRANSMISSION AND DISTRIBUTION NETWORKS
4.1 Introduction • Generators, especially large ones, are often distant

from centres of demand. This necessitates transmis-
In other chapters of this volume, the amount of raw ener- sion of electricity across large distances.
gy available in the wind has been discussed and the ways • Generation and demand must be balanced at any point in
in which it can be measured and assessed have been time across the network. There is little energy
described in detail. The WTs, which convert the raw ener- storage in large power systems compared to the energy
gy into electricity, have also been described. In order to be being consumed. There is, of course, substantial hydro
useful, this electricity has to be transported from the capacity in many European countries (approximately 16%
source – the wind farm – to the load, which may be a long of EU generating capacity). Some of this can act as stor-
distance away. It also has to be integrated with electricity age in principle but there is very little which is actually
from other generators. These two aims are achieved intended for that purpose, i.e. pump storage (5% of EU
through the grid – local, national and international. generating capacity). This is for economic reasons: ener-
gy storage technologies are relatively expensive, both in
The raw wind resource is vast. Technological improve- capital cost and in energy losses.
ments can be made to the turbines, but they already work • There is a culture of operating the system to serve
very efficiently on a large scale. Therefore, the key strate- demand, i.e. electricity consumers are uncontrolled, or
gic element which will determine the degree to which wind are controlled only by season, time of day or other
energy may realise its potential is its interaction with, and price signals. The generators are expected to meet
integration into, the European grids. For Irish wind to their requirements (a “predict and provide” strategy).
power the German Ruhr is feasible. Whether or not it hap-
pens depends on the grid – technical, economic, regula- So electricity systems, as we traditionally know them in
tory and, ultimately, political considerations about the use Europe, consist of large thermal and nuclear generators
and purpose of the grid will govern the outcome. (with hydro where available), and extensive networks to
transmit the power long distances to consumers. The
In this section, the constraints and opportunities offered demand is variable, and generation responds to this by
by the grid for large-scale wind energy integration are dis- increasing and decreasing its output. The larger the elec-
cussed. The detailed electrical engineering of grid con- tricity system, the smaller (in relative terms) the variations
nection is also described in detail in Appendices E and F. in consumer demand are, due to geographical diversity.
The tasks which face the engineering and regulatory arms This is one of the factors that favour large, highly-inter-
of the system operators and the industry as a whole are connected systems over smaller isolated ones. The other
outlined. such factor is the ability of large systems to cope better

with the failure of any one element.
4.2 Setting the Scene Because of this reduction in variability in large systems,
some generators can operate as so-called “baseload”
4.2.1 LARGE INTERCONNECTED NETWORKS generation, i.e. they operate continually at a constant out-
put. Typically, these are nuclear generators, which cannot
An essential element of establishing wind energy is to readily alter output, or the largest thermal generators. To
ensure that the electricity generated can feed into the cope with the daily variability in demand, other generators
grid system, and so reach electricity consumers. The (“mid-merit” plant) will operate throughout the day but
electricity grid, or network, exists to accept output from shut down or reduce output during the night. Then there is
all generators and to transfer power to consumers. “peaking” plant, to cope with the short-term peaks in
There are several key features of this arrangement which demand, typically in the morning and evening. Hydro gen-
70
have driven the design and operation of large electricity eration and pumped-storage plants may be suitable for
systems: this task, due to their short start time.
1
VOLUME
Within these general principles there are wide variations When aiming to increase wind energy’s contribution to
across Europe, depending on availability and cost of fuels, electricity supply the key questions are:
population density, location of centres of population and • Is it feasible to overcome the technical issues and at
fuel sources, climate and other factors. The liberalisation the same time maintain the quality of supply that we
of the electricity supply industry in each country also has presently experience and expect to continue?
a major effect. • Is it realistic to expect these issues to be overcome? If
so, what are the costs, including the costs of develop-
Liberalisation has tended to move large electricity sys- ment and operation of the electricity system?
tems from central planning to a market system where the • How can the costs of the various potential solutions to
rules of the market determine the behaviour of all parties. reducing environmental emissions be objectively com-
pared?
There are also technical factors which are changing the
situation. The major ones are: The following sections provide a more detailed description
• Electronic equipment, consumer and industrial, is less of the situation outlined above, provide some answers to
forgiving of poor “power quality” and supply interrup- these key questions, and indicate what needs to be done
tions than old fashioned heavy industrial plant. to achieve complete answers.
• Communications technologies mean that, in principle,
customers can become more responsive to price signals. 4.2.2 SMALL ISOLATED NETWORKS
• Gas-fired generation plant is available with cost benefits.
• Renewable energy technologies are emerging, particu- This chapter is mainly concerned with the connection of
larly wind generation. wind generation to large electricity networks, because this
• Smaller generators are connected to the lower levels of task is required in order for wind energy to make a major
the electricity network, perhaps even in domestic cus- contribution to achieving European environmental and
tomers’ premises. energy policy goals.
• New energy storage technologies may become
available. The vast majority of electricity consumers in Europe are con-
nected to large networks. Both policy goals and economic
The expansion of wind energy poses a number of issues for factors will tend to encourage further interconnection, as
electricity systems and their development which can be ref- small isolated networks generally have higher costs, lower
erenced back to the defining features of wind generation: reliability and/or higher atmospheric emissions.
• Wind energy output fluctuates. To an extent, this can
be controlled and/or predicted, but sometimes it can- However, small isolated networks will continue to play
not, or only with short notice. This is an added compli- their part, and there may well be other policy reasons for
cation for grid operators. It will be demonstrated that it encouraging the use of wind generation in these circum-
is important to distinguish clearly between predictabili- stances, particularly because such systems are often
ty and variability – terms which are often confused. located in rural or peripheral areas. Therefore, it should
• Wind energy can be in locations remote from demand be noted that most of the issues discussed in this chap-
and/or remote from existing conventional generators. ter for large systems are also applicable in some form for
This means that there need to be changes in the grid smaller isolated systems.
infrastructure. This may be required earlier than would
have been the case for simply replacing assets. The integration of wind energy into smaller networks is
• The technical characteristics of wind generation do not technically more complex, due to:
match the technical characteristics of conventional • Less geographical averaging of the variability of the
71
forms of generation, around which the existing elec- output of the wind generation.
tricity systems have evolved. • Less averaging of the variability of consumer demand.
For these reasons the management of variability becomes Distribution networks are less robust than transmission
particularly important on such systems. networks and their reliability decreases as voltage level
decreases, e.g. a connection at 33 kV could expect to
lose only a few minutes of connection per year on aver-
4.3 Electricity Networks age, whereas a low voltage connection at 230 V for an
individual domestic consumer in a rural area would, on
Electricity networks can be split into two major subsections: average, expect to lose at least an hour.
the transmission network and the distribution network.
There is very little so-called “active” management of dis-
The transmission network consists of high voltage power tribution networks. Rather, they are designed and config-
lines designed to transfer bulk power from major genera- ured on the basis of extreme combinations of circum-
tors to areas of demand. In general, the higher the volt- stances (for example, maximum demand in conjunction
age, the larger the transfer capacity. Only the very largest with high ambient temperatures, which reduce the capac-
customers are connected to the transmission network. ity of overhead lines), to ensure that even in these
Transmission network voltages are typically above 100 extreme circumstances the network conditions experi-
kV. They are designed to be extremely robust, i.e. they enced by customers are still within agreed limits.
can continue to fulfil their function even in the event of
several simultaneous failures of the network. Failure of a The addition of embedded generation to these networks
single element, such as a transformer or transmission creates challenges, for the following principal reasons:
line, is referred to as an “N-1” event, and transmission • The embedded generation adds a further set of cir-
systems should be capable of withstanding all such cumstances (full generation/no generation) with which
events. More complex cases of simultaneous failures of the network must cope, without negatively affecting the
multiple elements (for example, the failure of a transmis- quality of supply seen by other customers.
sion line when a parallel line has been disconnected for • The direction and quantity of real and reactive power
maintenance) can be termed “N-2” or similar. flows change, which may affect operation of network
Transmission systems should also be capable of with- control and protection equipment.
standing all such credible combinations. • Design and operational practices are no longer
suitable and may need modification.
Transmission systems are operated by transmission sys-
tem operators (TSOs) or independent system operators To set against these challenges, embedded generation
(ISOs). Responsibility for constructing or owning the net- also brings benefits to distribution networks, including:
work may belong to other organisations. • Reduction in network losses, in many situations.
• Deferring or avoiding network reinforcement otherwise
Transmission systems are actively managed through grid required to achieve standards for quality of supply.
control centres. Balancing the power entering and leaving To address these issues, distribution networks may
the high voltage network, and reconfiguring the network to become more “actively managed”. This implies cost, and
cope with planned and forced outages is a 24-hour activity. requires the development of suitable equipment and
Figure 4.1 shows the transmission network across Europe. design principles.
Distribution networks are usually below 100 kV and their

purpose is to distribute power from the transmission net-
work to the customers. At present, little generation is con-
nected to distribution networks, but it is growing rapidly.
72
Generation connected to distribution networks is often
termed “embedded generation”.
Figure 4.1: The European Transmission Grid
73
VOLUME
1
4.4 Considerations for Wind Energy This historical characteristic also has a strong influence
on the grid “culture”. Literally, how do the operators of the
There are three broad categories of issue which are rele- grid “feel” about the connection of load in this way? For
vant when considering connection of wind generation to some grid engineers this change in approach is difficult to
the grid. accept.
First, there is securing the immediate connection, which The capital cost of high-voltage equipment means that a
normally falls to the developer to negotiate with the rele- transmission system connection is only considered for the
vant network operator. This involves establishing whether largest wind farms, or where the transmission system is
there is sufficient capacity, and what effect the connec- much closer to the site than the distribution system.
tion will have on the network and other customers in the While the issues are similar to those for distribution net-
area. Second, there are considerations for operating the works, problems are less likely to occur, due to the
network. These relate principally to the intermittent nature greater strength of transmission systems.
of wind energy and ensuring that this does not impact
adversely on the operation of the network. Third, there are In addition, wind projects can be built much faster than
wider “strategic” considerations in planning for large conventional plants and, in particular, faster than the
amounts of new generation. transmission system can be modified or reinforced.
4.4.1 CONNECTIONS Appendix E provides a full commentary on the technical

and other issues which arise when seeking to connect a
Wind farm developers, like any other sponsor, wish to find wind farm to a grid.
the best point of connection for their development. “Best”
often means “cheapest”, but could, in some circum- All are considerations for any generation project. For wind
stances, mean “least risky” or “fastest to construct”. In energy, problems arise perhaps more commonly for the
some cases, the capacity of the available network con- following reasons:
nection will decide the maximum capacity of the generator. • It is a new technology, and practices developed for ear-
lier technologies (principally synchronous generators)
The majority of wind farms in Europe are connected to dis- are generally not applicable.
tribution systems, but proposals for larger onshore and • Its output is variable and less predictable.
offshore farms suggest that connection to transmission • Capacity factors are lower than other forms of genera-
systems will increasingly be sought. tion so it is harder to absorb the costs of network con-
nection, which are primarily capacity-related.
Similar issues arise when connecting to the transmission • The project location is often far from stronger sections
or distribution networks, although their relative impor- of the network.
tance changes. Most of the experience is at the distribu-
tion level. Here, the key influences are the preponderance In some cases, the network operators are unused to deal-
of connected consumers, and hence the need to maintain ing with embedded generation and, in particular, intermit-
power quality within acceptable levels, and the fact that tent generators like wind farms. The requirements and
these networks are not actively managed. design rules of network operators are not written with
wind generation in mind, and the application process for
Distribution networks are currently designed for power to a network connection is often not transparent. It will be
flow downwards from the transmission network to the cus- shown later that regulatory issues will be important in
tomers. Embedded generation, including wind farms, allowing wind to realise its full potential. On the other
74
changes the magnitude and sometimes the direction of hand, wind farm developers are sometimes reluctant to
power flow, and this can cause technical issues to arise. commit money to undertaking studies, or even to decide
1
VOLUME
upon the turbine type, the size of the wind farm or its lay- in the US, the UK, Scandinavia and Italy have also served
out. This can complicate the development process and to make this point.
lead to friction between the wind farm developer and net-
work operator. Wind generation is sometimes regarded as negative
load, since it can, to a degree, cancel out the demand
4.4.2 OPERATION for power at the point at which the transmission system
supplies the distribution system. For the purpose of
“Operation” means the day-to-day management of the net- matching demand and supply of a large electrical sys-
work. Distribution networks are not usually actively man- tem it is not possible to consider wind generation pure-
aged, so this section is more relevant to transmission net- ly as a negative load, however. Therefore, if there is a
works. However, there are pressures (due to embedded substantial proportion of wind generation on the system,
generation and other factors) to increase the manage- this adds another variable to the calculations that sys-
ment of distribution networks, and so this is expected to tem operators must perform in order to balance demand
become more relevant in the future. and supply. For this reason, wind generation may be
seen as increasing risk, and this perceived increased
Transmission systems are conventionally operated and risk is often resisted by the system operators. This
planned on the basis of a relatively small number of large resistance needs to be addressed in order to increase
generating plants, whose output can be varied at will. Wind the rate at which wind energy can be deployed on a large
generation, on the other hand, takes the form of larger scale.
numbers of smaller plant, whose output is variable and
less predictable. The science of prediction and forecasting It is important when discussing the intermittent nature of
has been covered in Chapter 2 and it is clear that this area wind generation to distinguish between two different con-
is likely to develop significantly in the near future, and will cepts: variability and predictability.
play an important role in the integration of wind power, but
will not provide a complete solution. Variability
Even if the output of wind generation were completely pre-
Balancing of Large Systems dictable, the variability of that output would increase the
difficulties of matching supply and demand. Setting aside
Large electricity systems operate with hardly any energy for the moment the issue of predictability, the variability
storage, as storage is expensive and the process results of wind generation is often seen as a problem. However,
in significant energy losses. Therefore, at any second, the analysis of data from operating wind farms and meteoro-
supply (output from all generators) has to be controlled to logical measurements from locations typical of wind farms
meet customer demand. produces the following conclusions:
• For a large individual WT, the variation in the output
In addition, the electricity system must be extremely reli- power is small for time-scales of less than a few sec-
able and robust, able to continue in the event of concur- onds, due to the averaging of the wind field across the
rent failures. For these reasons, system operators and rotor and the filtering effect of the turbine inertia (this
planners have to estimate demand and supply on time- is particularly true for variable-speed WTs operating at
scales of minutes to years. Given the size and complexity rated power).
of the plant this is a remarkable fact. • For an individual wind farm, the variation in the total
output power is small for time-scales of tens of sec-
The national economic consequences of complete failure onds, due to the averaging of the output of individual
of the electricity system, even for only a few hours, are so turbines across the wind farm.
75
great that it is worth network operators spending a great • For a number of wind farms spread across a large area,
deal of money and effort to reduce the risk. Recent events such as a national electricity system, the variation in
the total output power of all wind farms is small for Figure 4.2: Example of Wind and Demand over Two Days
time-scales from minutes or less, up to tens of (Winter 2003, Eltra System)
minutes. This is termed “geographic diversity”.
System operators only need to deal with the net output of

large groups of wind farms, and so the issue is what vari-
ability needs to be planned for and on what time-scales.
Analysis of available data allows estimates to be made of
the worst-case variation in net power output that can be
expected for a given time period, i.e. over 10 minutes or
over one hour. This form of analysis should continue to be
developed with high resolution data (i.e. sub-hourly peri-
ods) from larger numbers of wind farms, as suitable data
becomes available. As a starting point, see EWEA (2000)
and Commission for Energy Regulation/Office for
Electricity Regulation NI (2003).
This type of information can be used by system operators

to determine the level of reserve to maintain. tions may be controlled at critical times by measures such
as setting a temporary cap on the output of all wind
An extreme example of the variability of wind generation is farms, or by limiting the maximum rate of change of out-
illustrated in Figure 4.2 below, which shows two days in put (see below). For example, it is known from operational
early 2003 on the Eltra system (Denmark), which shows experience in northern Germany and Denmark that the
that wind power can maintain almost constant output over passage of a storm front may produce very severe
prolonged periods. The wind then falls in the evening just changes in wind generation output, as WTs shut down
when the evening demand peak is approaching. This fall from full power due to excessively high winds. This effect
is then followed by a rise in wind output at a time when can be limited by reducing the output power gradually over
demand is falling steeply. The coincidence of the steeply several hours in advance of the storm front.
falling wind supply and the steeply rising load requirement
makes the event extreme in nature. Clearly, limiting the output of wind generation wastes
“free” energy and should only be done when other means

Note that the wind generation is approximately equivalent have been exhausted.
to the minimum demand (overnight).
Predictability
The example given in Figure 4.2 is an extreme case, shown Important developments have been made in recent years
here to highlight the important issues. The majority of days with tools to aid in the forecasting of energy output from
do not exhibit this extreme situation. The counter example wind farms. Details have been provided in Chapter 2 of
is the growing use of wind energy as a source of spinning this volume. It has generally been found that over short
reserve, as a benefit from the very rapid response times of time frames, and with good data about the historic wind
wind turbines. regime at a site, it is possible to predict wind farm output
using a correlation with forecasted meteorological data
Conventional techniques available to system operators for from nearby weather stations. Although these tools are
such situations include keeping other generation operat- still in a relatively early stage of development, it is now
76
ing at low output, and making use of interconnections to possible to provide vital forecasting information, which in
neighbouring systems. Alternatively, these extreme varia- some cases is a requirement of the project’s power pur-
1
VOLUME
chase agreement. This information can then be used by Some difficulties in the process of modifying these codes
the system operator in balancing the generation with are listed below:
demand in their system, and significantly reduces the • The dominant WT technology at present, the doubly-fed
level of uncertainty to which wind generation has histori- induction generator (DFIG) is not well understood by
cally been attributed. network operators.
• The important characteristics of wind farms relevant to
The system operator can also undertake its own fore- network operators are not well defined, i.e. there is a
casts, perhaps using as an input the current and forecast lack of a common vocabulary for discussing these
output of each wind farm, as well as large-scale meteoro- issues and comparing characteristics of alternative WT
logical forecasts. types.
• There are some functions that are desirable, but the
This unpredictability can be dealt with by the system cost-to-benefit trade-off to provide them is not clear.
operator assuming a forecast level of wind generation into • The existing grid codes grew up around synchronous
the future (typically for up to 24 or 48 hours ahead). The generators so it is difficult to rewrite them from a
operator then applies a forecast error to calculate the generic or functional viewpoint.
amount of wind generation in future hours which can be • There may be some valuable functions currently pro-
treated as “firm”. It has been found in several studies that vided by conventional generation which have not been
the forecast error increases with look-ahead time. For formally identified in grid codes.
example, if the output of all wind generators is forecast to • There may be some functions that would now be better
remain constant for the next 24 hours, the firm contribu- provided by a market rather than by an obligation on all
tion of that generation will be assumed to decrease over generators.
that period. • It is desirable that all grid codes are similar but there may
be some real technical differences between systems.
Improved forecasting will allow the forecast error to be
both specified and reduced, thus allowing more wind ener- Because of the expected expansion of wind energy, time
gy to be treated as firm. is pressing and the need for rapid solutions to the prob-
lems conflicts with the need for widespread consultation
Grid Codes among all interested parties. This task requires immedi-
ate action.
Grid code documents set the requirements for users of
the transmission or distribution system, including genera- 4.4.3 STRATEGIC PLANNING
tors. These codes cover issues that are important for CONSIDERATIONS
operation of the system, although some aspects are also
relevant to system planning. As total wind generation capacity increases, its effects on
large electricity systems (national scale) will eventually
These codes have evolved to suit conventional genera- become significant. Both the transmission system and
tion: substantial modifications are required in order to the existing generators could be affected. This section
apply them to new forms of generation, particularly wind. discusses these strategic issues.
Such modifications have been produced, or are in the
process of being developed, in many European countries. The effect of wind generation on an electricity system
depends on the “penetration”. Two different definitions
Typical requirements of grid codes for wind generation, are often used:
published or in draft, are summarised in Appendix F. • Wind energy penetration: annual production (GWh) of
77
wind generation as a fraction of total consumption.
• Wind capacity penetration: wind generating capacity who should pay or be paid for this change in losses. This
(GW) as a fraction of total generating capacity. is a market issue. Several solutions are possible and,
indeed, different approaches may be suitable for different
Both issues are important when considering the effects of electricity systems. All that it is necessary to say at this
large amounts of wind generation. point is that any system of allocating costs of losses
should be able to allow for the contribution, positive or
Displacement of Conventional Generation negative, of intermittent generation, particularly embed-
ded generation.
Wind energy to date has tended to displace the output of
conventional plant, reducing the conventional plant’s con- Benefits of Interconnected Systems
sumption of fuel. This is effectively the rationale for its
promotion on environmental grounds. In some EU coun- Experience has shown the benefits of combining a diverse
tries, it displaces imported conventional fuels and uses a mix of demand and supply types, via interconnected trans-
national sustainable energy source in their place. It does, mission systems, enabling greater wind generation to be
however, raise some policy issues in the form of eco- connected.
nomic questions for the owners and financiers of conven-
tional plant. This is demonstrated by way of example. Table 4.1 shows
the characteristics of two contrasting electricity systems.
The costs of conventional generation can increase due to
the following causes: Table 4.1: Comparison of the Eltra and Crete Systems
• If forced to run at lower output, conventional genera-
Item Eltra System, DK Crete System
tion may operate at a lower thermal efficiency, thus (2001) (2001)
consuming more fuel per MW hour of electricity pro- Total conventional generation
duction and producing more pollution per MW hour of capacity [MW] 4,724 570
electricity production. Total wind capacity [MW] 1,932 70
• Due to the variability of wind generation, conventional Installed wind capacity penetration 29 % 11 %
generation will be called upon to start, stop and change Wind energy penetration 16 % 10 %
output more frequently. This also incurs costs. Transmission capacity to
other networks [MW] 2,640 0
In addition, if a generating plant has been financed on the

basis of some assumed annual output, and that annual

output can no longer be achieved, the plant may become The island of Crete is completely isolated from all other
uneconomic and may be closed. Those who financed the electricity systems. Wind energy penetration had already
plant may want compensation. reached the level of 10% by 2001, and continues to
increase. For technical reasons it has been found neces-
Such effects are not addressed in detail as part of this sary to keep large amounts of conventional generation
discussion, but they are central considerations for policy- operating during periods of low demand, even when there
makers, developers or financiers and can be expected to is high wind generation. This means that a substantial
precipitate important political debate. amount of the wind energy has to be curtailed.
Losses In comparison, the Eltra system has an even higher level

of wind energy penetration but has not (yet) needed to cur-
New wind generation can, depending on its location rela- tail wind output. This is because it is highly interconnect-
78
tive to the main loads, increase or decrease the electrical ed to the Nordpool system of the Scandinavian countries
losses within the network. This raises questions about to the north and to northern Germany to the south. The
1
VOLUME
Eltra system has, however, been subject to situations costs of rapid changes in output, and costs of operat-
where the generating output exceeds demand, particular- ing away from peak thermal efficiency.
ly overnight when there is high wind generation. In these • Providing other generation such as open-cycle gas tur-
circumstances, the system operator has had to sell energy bines which have low capital cost and high fuel cost,
cheaply to the other networks to which it is connected. This and which can start, stop and change output very
highlights the important point that the technical feasibility rapidly.
of high wind penetration is a separate issue to the
economic effects. It is not at all clear how these alternatives compare to
increased interconnector capacity. It is instructive to note
There are two particular features of the Eltra system that that, even over significantly large geographical areas, out-
are worth noting: put from all of the wind farms in the region could be very
• At present, the system operator must accept all wind low as a result of common weather systems (anticy-
generation that wishes to generate at any time, in pref- clones). This is certainly the case for areas the size of the
erence to generation from conventional sources. UK, for example. This is of particular concern in winter
• The same also applies to the output of district heating when anticyclones are accompanied for several days by
plants and combined heat and power plants. A large clear cold conditions and electricity demand can be
proportion of these plants are so-called “heat-led”, i.e. expected to be high.
their electricity production is determined by their heat
production; therefore, when temperatures are low and Therefore, in order to replace wind generation in an area
there is a high heat demand, their electricity production covered by an anticyclone by wind generation outside that
is also high. area, transmission system reinforcement for distances of
several hundred km will be necessary. The wind genera-
The Nordpool system to which the Eltra system is connect- tion capacity factor of this new transmission reinforce-
ed is highly suitable for connection to networks with large ment would be in the order of, say, 25% to 35%. As trans-
amounts of wind generation as it contains large amounts of mission reinforcement costs are dominated by capital
hydro generation which provides very fast, low loss storage costs, transmission system reinforcement for the purpos-
capacity. On the other hand, the north German networks to es of transfer of wind energy appears to be less attractive
the south of Denmark have themselves a large proportion than transmission system reinforcement for conventional
of wind generation, so when there is surplus wind in purposes.
Denmark it may be of little value as there is also likely to
be surplus wind in the north of Germany. This assumes, of course, that sufficient transmission sys-
tem capacity does not already exist. However, as trans-
Interconnection of electricity systems with wind energy mission systems are built in response to existing patterns
therefore brings benefits, and these can be characterised of generation and demand, this assumption seems gen-
as being due to the ability to reduce the effect of the vari- erally to be justified.
ability of wind.
A key parameter for the successful large-scale penetration
However, other means are available to provide similar of wind into the European grid will therefore be the size of
benefits. Some of these are: the interconnected system relative to the size of a large
• Curtailment of wind generation output. anticyclone. This point is demonstrated graphically in
• Energy storage (hydro generation, pumped-storage Figure 4.3 which might be subtitled Wind from Spain and
plant, or new forms of storage). Ireland powers the Ruhr. The interconnected grid will have
• Demand management. to be large enough to make sure that wind is available to
79
• Reducing the costs of operating conventional genera- meet the required load at all times.
tion in a more variable regime, such as start up costs,
Figure 4.3: A Large Weather System over Europe 4.5 Issues for High Wind Penetration in
Europe
The key questions defined at the start of this chapter are

repeated here for reference:
1 Is it technically feasible to overcome the technical
issues and, at the same time, maintain the quality of
supply that we presently experience?
2 Is it realistic to expect these issues to be overcome? If
so, what are the costs, including the costs of develop-
ment and operation of the electricity system?
3 How can the costs of the various potential solutions to
reducing environmental emissions be objectively com-
pared?
For the first question, it is clear that it is technically fea-

Generation Adequacy and “Capacity Credit” sible to have very high wind penetration on the European
electricity systems, without affecting the quality of supply.
“Generation adequacy” means the likelihood that available
generation capacity will be insufficient to meet demand. For example, a combination of large-scale transmission
Detailed probabilistic calculations are carried out to deter- reinforcement, new interconnector capacity, new energy
mine the probability of this occurring. If the result is below storage capacity (probably hydro generation and pumped
the accepted requirement then something has to be done. storage), and wind curtailment would, in principle, allow
The traditional solution was to build more generating 100% of Europe’s electricity consumption to be provided
capacity, but this is harder to arrange in a deregulated or by wind energy if sufficient land and sea area were made
liberalised system. The system is designed to cope with available for wind farms and energy storage. This is a
failures of the transmission system and individual genera- hypothetical case, as the cost of reaching this 100% fig-
tors, but wind generation has the additional risk that all ure would be very high, and would almost certainly not be
wind generation may shut down at the same time due to the optimum means to achieve such a large reduction in
low winds over a large area, as discussed above. emissions from conventional generation.
Formal probabilistic assessment studies show wind does However, a major part of this cost would be for the ener-
provide some contribution to reducing this risk of inability gy storage capacity required to supply the electricity sys-
to meet demand with supply; for example, see tem through periods of several days or weeks of low wind.
Transmission System Operator Ireland, Generation This would not be economically justifiable; it would be
Adequacy Report 2003-2009 (2002), which calculates a cheaper to provide that “storage” by storage of fuel for
“capacity credit” of approximately 20% depending upon suitable conventional generators, possibly gas turbines,
circumstances. This capacity credit means that, in princi- or by using other renewable technologies.
ple, less conventional generation need be built. If capaci-
ty or availability payments are made to generators for their A strengthened European grid would reduce the need for
contribution to generation adequacy (which is not com- storage. Development of sustainably produced hydrogen
mon) then, in principle, payments should also be made to and superconducter technology may also play a role.
wind generation.
80
1
VOLUME
As more conventional generation is added to the above faster than energy penetration. It is more important to
scenario, the cost of such a system will reduce, and establish what the costs are, in order to compare them
atmospheric emissions will rise. against the costs of competing options. These costs will
depend on the generation mix, the level of interconnection
The second question cannot be answered fully at present, and forecasting ability.
as the costs of operating an electricity system with high
wind penetration (say above 30%) are not yet clearly Variability
understood. The sources of these increased costs are,
however, becoming more clearly understood, as are Because of the major benefits (for system operators) of
means to mitigate them. geographical diversity, the variability of wind generation is
not as severe as is often perceived. Variability can also be
There have been several studies (EWEA, 2000; damped by relatively simple measures such as capping or
Commission for Energy Regulation/Office for Electricity ramp rate control of wind farms during critical periods.
Regulation NI, 2003; Dany et al., 2003; UK Department of However, it remains a major disadvantage that becomes
Trade and Industry, 2002), on this issue which provide more important at high penetrations, principally because
some illumination. It is necessary to continue such stud- it increases the amount of reserve that an operator must
ies at higher wind penetrations than have so far been carry on the system, and/or requires some other solution
envisaged, updated with operating data from large num- such as curtailment of wind generation or increased inter-
bers of wind farms. connector capacity.
The answer to the third question is also not clear at It is important to gain experience of system operation with
present, principally because the second question cannot be wind generation, and system operators should be encour-
fully answered. This is, of course, also partly due to the aged to record data from their systems which will allow
uncertainty in the costs of other renewable technologies: in these issues to be studied.
particular, whether they are able to achieve the same major
benefits from volume production that the wind industry has Predictability
achieved. However, it is clear that a major part of the addi-
tional cost borne by electricity systems with high wind pen- As for the issue of variability, the unpredictability of wind
etration is due to the variability and unpredictability of wind. generation requires system operators to carry additional
The latter cost is set to reduce as prediction techniques reserve or possibly force curtailment of wind generation.
improve. Therefore, other renewable technologies which do
not have these characteristics will have an advantage at Significant improvements in wind output forecasting are
high wind penetrations. anticipated through work currently in progress.
The above questions have helped to identify the important Technical Requirements for Wind Generation
issues for achieving high wind penetration in the electrici-
ty networks of Europe. These issues can be summarised Grid codes for distributed generation and wind generation
as follows. are being developed. It is expected that the wind industry,
given sufficient evidence of a continuing market, will devel-
Technical Limits to Wind Energy Penetration op technical solutions to meet these requirements, at costs
which will not be significant in the context of overall project
In principle, there are no technical limits. Very high wind costs. If suitable solutions do not appear (which seems high-
energy penetrations appear achievable. The additional ly unlikely), then it is possible that this issue could slow the
81
costs appear to be low at the local high penetrations cur- rate of expansion of wind generation.
rently achieved, and are expected to rise proportionately
It is important that these requirements are formalised as In this context, it can be noted that no form of generation
soon as possible, consistent with adequate consultation is completely reliable; all require the electricity system to
with interested parties. carry some reserve, in the form of additional generation
that can start up or increase its output rapidly. Therefore,
It appears that it will not be possible to develop one set the point at issue is how much this level of reserve needs
of requirements for all electricity systems in Europe. For to be increased to cope with wind generation.
example, it appears that requirements in the UK and
Ireland for transient stability may be more onerous than It is also generally true that the energy production of wind
on the European mainland, due to their island nature. generation is expected to grow more rapidly than electric-
However, this does not appear to cause any particular ity demand. Therefore, the existing conventional genera-
problems for WT manufacturers. tion can expect to be in use less often: it is expected that
any increase in reserve requirements due to increased
System Operation Costs wind generation will not lead to an overall increase in the
total capacity of conventional generation.
The additional costs imposed on operators of electricity
systems are not clear, as noted above. The costs at high
penetrations (above 30%) are particularly unclear. 4.6 Concluding Remarks
Further work in this area is required, making use of oper- When considering the connection of wind generated elec-
ational data from systems with distributed wind farms as tricity to the grid, the starting point often seems to be why
it becomes available. it cannot be done rather than how it can be done. This
starting point will prove crucial in the successful integra-
Curtailment of Wind Generation tion of large amounts of wind. If a political goal is set, as
it has been at both European and national levels, then the
It appears that the variability and unpredictability of wind challenge is to allow that goal to be met in a systematic
adds to system operational costs principally by introduc- and rigorous fashion. It is possible to meet environmental
ing “extreme” events such as storm fronts. These policy objectives and keep the lights on, but to do so
extreme events happen rarely (a few times per year) and requires some imaginative thinking and new research. It
therefore coping with them by capital-intensive means is this aspect which will either allow wind energy to fulfil
such as building additional fast-response generating its promise or leave it as a marginal player.
plant, or similar means, appears to be less attractive than

curtailing wind generation by capping it or controlling ramp The grids have been designed for large-scale central gen-
rate. Any further studies of system costs should take eration, whose power is transported outwards through the
these options into account. transmission and distribution systems. They were not
designed for, and is operated with, substantial distributed
Capacity Credit and “Back Up” Generation Sources generation. That requirement is coming, not just as part
of the introduction of renewable energy, but also as a
This issue is now becoming important, with a belief result of a much increased interest in smaller scale com-
among opponents of wind generation that 100 MW of mercial generation as the electricity industry is liberalised.
wind generation requires an additional 100 MW of new A change in attitude and operation will be required to
conventional generation to back it up during calm spells. accommodate this change.
Further work in this area is required, preferably based on
analysis of the probability of failing to meet generation The grid codes were written with conventional generation
82
adequacy targets, and using several years of operating in mind and also, in particular, with synchronous genera-
data from distributed wind farms. tors in mind. Thus, it is historical precedent rather than
1
VOLUME
pure technical necessity for the preclusion of large-scale Grid operators must be educated to recognise that
wind generation from the grid codes. These codes should although the wind is variable it is also predictable and
be revised, not in a bid to compromise the security of the hence it can, when considered in significant aggregated
system through relaxation of the terms, but rather to capacity, be scheduled at a time-scale which is commen-
recognise that new types of generator and new sources of surate with conventional plant.
energy are available. The grid codes should be optimised
to allow the best possible mix of generation rather than to The level of penetration which can be achieved by wind is
prefer one particular category. essentially limited by cost rather than by some funda-
mental technical considerations. Investigation of the
Geographical averaging is a powerful tool to smooth the cost-penetration relationship merits serious investiga-
variations in wind energy output on all time-scales. It also tion. This relationship will be system-specific, as it
increases the extent to which power system planners can depends on factors such as the other forms of genera-
rely on wind energy to meet future demand – the so-called tion available. Interconnection is not the only solution;
capacity credit. It can only do so if the size of the inter- careful curtailment of energy production during
connected system is large enough to compensate for the predictable extreme events may well lead to a more
effect of the weather systems likely to occur over Europe. cost-effective approach.
Systematic investigation of the relative size of the weath-
er and electrical systems will provide a valuable insight Rigorous work on the establishment of the real capacity
into the strategic value of wind energy and the cost of credit which should accrue to wind generation is well over-
security of supply. due and is amenable to systematic investigation.
83
5 RESEARCH AND DEVELOPMENT
5.1 Wind Industry Research and European R&D programmes over the last 15 years have
Development Overview been at the forefront of today’s industry. Results of such
programmes include the development of large MW turbines,
Early R&D was technology driven but, as the industry the first European wind atlas, and funding demonstration
expanded, other issues came to light, such as public atti- and pilot projects, such as the first offshore wind farm. It is
tudes, noise levels, environmental impacts and wind ener- essential that this R&D continues with the support of EU
gy financing. Furthermore, the technology emphasis has research programmes, such as FP6 and its successors. As
sometimes been superseded by a strong market push - to the IEA puts it:
bring larger machines with greater output into circulation “In order to achieve a 10 to 20% part of the world-
as soon as possible. Consequently, some more wide energy consumption provided by wind, major
fundamental, promising options in design have not been steps have to be taken…it is for this objective that
implemented. This is not always the case however, direct there is a need for long-term R&D.”1
drive turbines being a good example, but it is certain that
the need for technology driven research is still important, It is interesting to note that wind technology has been
and support for it is needed. influenced by, and exerts influence on, other industry
sectors, such as, for example, aerofoil design in the aero-
This chapter provides an overview of the different fields nautics industry, and in the shipping industry where the
related to R&D in the wind power industry. Some key requirements of the offshore wind power industry has lead
needs are highlighted, but it must be stressed that this to nautical R&D to provide suitable craft for erecting wind
chapter does not prioritise specific needs. The Wind turbines (WTs).
Energy R&D Network provides further information on R&D
issues (see below). Wind Energy R&D Network
Objectives The Wind Energy R&D Network, coordinated by EWEA, is

formulating a strategy for R&D in the European wind indus-
A principal objective of wind industry R&D is to meet the try. It brings together actors from across the wind energy
levels of wind penetration described in the EWEA feasibil- sector and enables dialogue between industries and other
ity study Wind Force 12 – A Blueprint to Achieve 12% of the stakeholders with the common goal of substantially increas-
World’s Electricity by 2020 (EWEA, 2003c). ing the share of wind energy technology in global electricity
markets. Finalisation of the strategy is set for 20052.
To achieve this goal, the industry needs to:

• continue making cost reductions. Priority R&D Areas
• enable increased penetration of wind power.
• minimise environmental and social impacts. The Wind Energy R&D Network has established priority
R&D areas as below:
During the last two decades, R&D programmes have been • Economic, Policy and Market Issues:
a pre-condition for the successful development of the for example, assessment affecting wind farm invest-
wind power industry to date. In its 2001 report Long Term ments and market barriers.
Research and Development Needs for Wind Energy for the • Environmental and Social Impacts.
Time Frame 2000 to 2020 the IEA states: • Wind Turbine and Component Design Issues:
“Thanks in large part to successful R&D, the wind for example, basic research in aerodynamics, structural
energy market is in a state of rapid development. dynamics, structural design and control.
R&D has been an essential activity in achieving the • Testing, Standardisation and Certification:
84
cost and performance improvements in wind power for example, common accepted certification procedures
generation to date.” for WTs and wind farms.
1
VOLUME
• Grid Integration, Energy Systems and Resource Prediction: As turbines become larger and more powerful, requiring more
for example, forecast of wind resource. advanced technology, expertise and refinement, greater
• Operation and Maintenance (O&M): development costs, such as increased engineering hours,
for example, advanced condition monitoring. become apparent. Increasingly complex control requirements
• Location of Wind Farms: due to the complex design of WT systems and farms also
for example, in complex terrain and remote areas where contribut to increased development costs.
satellite technology can be used, among others, in the
formulation of wind atlases – showing the wind resource. Reductions in project costs can be expected through
• Offshore Wind Technology: rigorous technical and economic standardisation, facili-
for example, research into the control and efficiency of tating the specification of components, and their certifi-
very large wind farms and more cost effective founda- cation. This means greater transparency and increasing
tions, transport and installation techniques. competitive pressure leading to greater efficiency in the
• Megawatt and multi-megawatt Wind Turbines: market place.
for example, application of new materials with improved
strength-mass ratio and development of lighter 5.2.1 TRANSPARENCY
components.
Despite EU efforts to remove legal barriers between mem-
Underlying these research areas is the drive to increase ber states, many country-specific regulations remain,
economic efficiency through reduction of uncertainties in hampering transparency and the free market. Greater
fields ranging from resource prediction to improvements European-wide harmonisation is vital if a future internal
in component reliability. Reducing uncertainties in assess- market in wind power is to be developed.
ing technical risks will reduce the cost of services provid-
ed by finance and insurance companies and allay fears R&D Objectives
relating to security of supply. These uncertainties are:
• Resource assessment and wind speed measurement. • Categorisation of turbine types on the basis of
• WT reliability. manufacturer data; further use of ISO/IEC and
• Performance predictions. CEN/Cenelec standards; and development of new
• Prediction of O&M costs. standards.
• Increased maintainability of machines. • Development of a risk assessment standard for wind
• Lifetime design methods. turbine projects.
• Grid assessment. • Certification of standards for wind energy projects
addressed to financiers and insurance companies.
• Increased transparency in respect of weather related
5.2 Socio-Economic, issues on the one hand, and with regard to
Policy and Market Issues operational damage on the other, taking into account
the perspective of the insurance industry.
There is an overall need for research into methods of cost • A continuously updated web based database of
reduction and risk management that can increase the member state rules and regulations: policy instru-
value of wind energy. Milestones towards cost reduction ments, planning and construction regulations, permit
need to be defined. and environmental issues, tax law, corporate law, etc.
85
5.2.2 INCREASING THE VALUE OF WIND POWER 5.3 Environmental and Social Impacts
At present, value is not only determined by the avoided It is essential to express continually to the general public
fuel cost of fossil-fuelled plants through installed wind the predominately positive social and environmental
capacity. Increasing predictability in power output, and an aspects of wind energy, in order to maintain and improve
increase in capacity factor has further raised the value of its support for wind power. R&D efforts should aim to
wind electricity. Equally relevant are cost components in increase public involvement, and to further minimise
electricity production, such as environmental benefits and social impacts.
consumer preference for green electricity, which are not at
present internalised in electricity prices. As the Social R&D Objectives
cost/value ratio is optimised, the economic drive to
realise wind energy plants will grow. • Development and verification of public participation
models.
R&D Objectives • Assessment of the positive social effects of wind
energy, such as local employment, investment,
• Development of output forecasting models. taxes, etc.; and the creation of local networks to
• Improved controllability of large wind farms. express these local benefits.
• Effective quantification of external benefits of wind • External costs: an accepted methodology for the
energy; and development of methods to quantify the assessment of environmental savings through the
cost/value ratio. use of wind energy is needed, to establish a
quantifiable and cogent benefit.
• Clear understanding of the external social and
5.2.3 EDUCATION AND HUMAN RESOURCES environmental costs of conventional power
DEVELOPMENT generation, as well as of possible external costs of
large wind penetration, is needed so accurate
1.8 million job-years will be required to meet the comparisons may be drawn.
Windforce 12 target of 12% wind electricity worldwide in
2020. Education and training are required in both techni-
cal and non-technical capabilities in order to provide a
skilled workforce to satisfy future demand. In 2003, the
European Academy of Wind Energy was established with

the aim, inter alia, of providing for this demand. Visit
www.eawe.org for further details.
R&D Objectives
• Establish where skill shortages will occur along the

growth curve.
• Joint international R&D programmes in universities.
• Develop training schemes to supplement work-
based training.
• Establish specialised professorships at universities.
• Develop educational material for primary and sec-
86
ondary schools.
1
VOLUME
Environmental R&D Objectives issues with regards to a specific application such as off-
shore or multi-MW scale turbines, or O&M.
• Recommendations for limits on stroboscopic
effect/shadow flicker (taking into account seasonal vari- All categories of turbines are, to a varying degree, on the
ations) in residential areas and other sensitive areas. frontier of technology research. Enormous progress has
• Methods should be developed to integrate turbines been made to date in increasing efficiency, and reducing
visually into the landscape on an individual turbine and the cost of turbine production. R&D into the manufactur-
wind farm basis. This may be approached in two ways: ing process is key to further cost reductions. For example,
through design of individual turbines; and their effective an assembly line approach to WT construction would to
siting in the landscape, with the use of camouflage and lead to benefits through economy of scale. Continued
stealth techniques. R&D efforts are needed in order to improve the technical
• The reduction of noise impacts to decrease the mini- and economic efficiency of WTs and their components.
mum required distance between turbines and residen-
tial areas. This would increase the potential for wind New turbine models are increasingly complex, yet once a
energy utilisation in populated areas. Methods to pre- turbine is deemed ready for market, little time is dedicat-
dict the noise level generated by turbine blades, geared to prolonged testing. This is due to market pressure for
boxes, generators and transformers need to be new, larger turbines. As a result, new preventive features
improved through more fundamental R&D, leading to that might help reduce O&M requirements are not always
the production of low noise blades, gearboxes, and gen- integrated in new designs. Turbine cost reductions of 15-
erators. 20% may be realised through the implementation of such
• Turbine interference on telecommunications/radar research results. R&D must be combined with market driv-
needs to be quantified. At present, large areas with high en development, and a doubling of accumulated capacity
wind potential are restricted by the military. Real dia- to achieve this cost reduction.
logue is needed among industry and the military, to
establish an acceptable level of understanding in rela- Horizontal integration between the manufacturing sector
tion to national defence. and research institutes is at present limited. The two are
• The identification of areas where potential impacts exist arranged on different time-scales and there is a need for
on bird populations, habitats & flight paths, as well as further incentives for cooperation. Within the European
ways of mitigating such impacts. Research Area, a wider network for manufacturers, sub-
• A standard for turbine design involving the use of life suppliers and research institutions should be encour-
cycle analysis to identify recyclable materials, and to aged, yielding integrated, active research and aggressive
specify how to dispose of non-recyclable elements, e.g. implementation of research findings.
stand alone systems involving the use of batteries.
• Identification of potential cumulative environmental R&D Objectives
impacts of increased numbers of wind farms, across
the EU. • External design conditions (wind climate assessment
• Decommissioning alternatives. etc.), aerodynamic and aeroelastic design, structural
design, loads and safety.
• New materials with higher strength as well as higher
internal damping.
5.4 Turbine and Component Design Issues • Advanced manufacturing technologies.
• Feasibility studies of new wind turbine concepts and
This section discusses design issues related to turbines innovations, e.g. flexible blades & hubs, and variable
87
and their components, but it should be noted that other speed generator systems, to display potential for
sections in this chapter are also concerned with such reducing cost per kWh.
• Methods of reducing O&M costs in the design phase. reputation; as the market is sensitive to negative report-
• Reliability models leading to higher wind farm availabil- ing. Standards designed for one market segment can be
ity – particularly relevant regarding offshore turbines. inappropriate in another, and standards across the seg-
• Development of more efficient testing and verification ments should normally be limited to essential operating
methods to both shorten turbine development periods, and safety standards.
and improve the quality of verification process.
• Integration of demand side requirements in the design R&D Objectives
of turbines, e.g. electrical control system interaction
with grid requirement. • Identification of standards lacking, and initiation of
• Fundamental R&D into site-specific control of turbines appropriate actions for new standards and back-
to cope with variations in external conditions, e.g. high ground research.
turbulence levels. One solution is to build in methods • Background R&D into a standard for service and main-
for fine tuning the aerodynamic and structural perform- tenance concepts, including labour safety.
ance of rotor blades and associated fast field diagno- • Guidelines and standards describing the steps in pro-
sis instruments. E.g. if prohibitive vibrations appear dur- ject development, according to sector (deep/near off-
ing operation, control systems can be used to actively shore; mountainous/isolated/coastal onshore, etc).
damp vibrations after the turbine is put in operation. • Development of turbine type categories on the basis
• Component design, such as longer blades, and electri- of ISO/IEC and CEN/Cenelec standards.
cal components. • Co-ordination of system development & testing pro-
grammes in place in major European R&D centres,
with the full involvement of the manufacturing industry.
5.5 Testing, Standardisation, • Background R&D into standards for project perform-
Certification & Safety ance testing (production verification).
Also discussed in the section on Socio-economic, policy &

Market issues, standardisation, testing and certification
are themselves the potential results, or indeed overall 5.6 Grid Integration, Energy Systems
objectives, of the R&D process; they are essentially R&D and Resource Prediction
results applied to specific areas. Standardised certifica-
tion and testing techniques lead to increased transparen- European utilities are increasingly seeing wind power as a
cy in the wind power market, and thereby reduce costs. viable and reliable source of energy in the grid's supply
For example, a certification system covering not only portfolio. However, there is a reluctance to implement
power-curves, but the entire turbine and project develop- wind capacity, and further understanding of the total
ment process, along the lines of ISO 9000, would capacity of wind generation that can be absorbed locally
increase the insurability of the entire sector. Existing and regionally by grids is required.
design certification might be supplemented by requiring
testing of components under specific test protocols, as is Wind farms must not only ensure efficient operation in vary-
the case with blades under the Danish approval scheme. ing meteorological conditions, but have also to answer the
Project certification can be done to a degree under the requirements of the transmission and distribution (T&D)
IEC-CAP standard, however, the CAP standard is not suffi- networks to which they are connected - requirements such
cient as it stands. as high power quality and steady output. These require-
ments often do not take into account the distributed nature
High quality and efficient standardisation and certification of wind power generation as T&D systems are developed
88
are vital given the number of turbine types. Low quality tur- for large, centralised fossil fuel and nuclear power plants.
bines on the market would hamper the wind industry’s
1
VOLUME
R&D Objectives 5.7 Operation and Maintenance
• Development of scenarios for redesign of the EU O&M requirements are driven by site conditions, quality of
T&D grid, with high wind penetration. components and turbine design. Costs and electricity output
• Development of tools to enable grids to cope with depend on availability and, therefore, on the reliability and
large-scale wind power. accessibility of the WT installation. Figure 5.1 describes
• Increased predictability of system output. To devel- availability as a function of reliability and accessibility. In
op electrical output prediction tools (meteorological particular, WTs located in offshore locations and in moun-
forecast) to predict output 24-48 hours in advance tainous terrain are subject to potentially very high costs due
for wind farms. to availability losses through potentially avoidable O&M.
• Longer-term forecasting.
• Increased accuracy in pre-installation prediction of Figure 5.1: Availability as a Function of Reliability and
electricity output via tools such as anemometry, ter- Accessibility
rain calibration, and the translation of power wind
speed curves from test sites to installation sites.
• Energy management and storage systems for stand-
alone applications.
• Develop demand side management tools to encour-
age consumers to focus their electricity use during
high electricity output periods.
An increase in predictability of wind farm output will reduce

requirements for spinning reserve. Improved levels of grid
interconnection among European countries will also reduce
the need for spinning reserve, as intermittency becomes
less of an issue with increased geographical spread.
Source: Delft University, et al., (2001)
Analysis of spinning reserve requirements and intermit-
tency should be conducted on the basis of the require-
ments of the entire electricity system rather than on a In this respect, fast development of early failure detection
technology-by-technology basis, to achieve optimal ecosystems is essential. Data describing the properties of WTs
nomic and technological solutions. Distribution network should be gathered and used as inputs for improving the
operation should be remodelled taking wind farms into reliability of turbine design. Statistical data on the machine
account as large-scale as well as decentralised contribu- performance and component outage events are necessary
tors to total electricity production, rather than considering as a basis for maintenance optimisation. Much research is
wind as an element to be considered in isolation. needed regarding early failure detection and condition mon-
itoring to establish trends to assist in predicting part failure.
The fact that WTs are clustered into bigger units requires
specific R&D into the performance of large wind farms. On R&D Objectives
the one hand, national grids must be able to absorb large
amounts of varying electricity output while on the other, • Accelerated development of early failure detection
wind energy plants must meet targets in terms of the and condition systems.
amount of reactive power produced, harmonic distortion, • Cross industry analysis of turbine and component
89
predictability and controllability of power output. performance and maintenance statistics.
• Resolution of confidentiality issues. Complex Terrain
• Increased understanding of the drivers behind O&M
costs in offshore sites. There is an urgent need for more reliable methods for the
• Developments in preventative maintenance. prediction of wind conditions in complex terrain. At the
• Standardisation of components and turbine docu- moment, there is very limited support for R&D in this area.
mentation for ease of replacement. The wind energy community is collaborating with meteoro-
• Harmonised methods and certification of O&M com- logical institutes and carrying out its own research. Also,
panies to a high standard. the short-term prediction of the output of farms in complex
• Certification of service and maintenance concepts. terrain is limited because of a lack of knowledge and tools.
• Better techniques for assessing wind turbine and
wind park performance in situ. Icing Environments
The potential for WTs in icing environments needs further

investigation. Particular issues include blade heating ver-
5.8 Location of Wind Farms sus power performance and safety aspects of ice-throw
from the blades.
Already, in some densely populated countries with a high
level of installed wind capacity, the best sites, in terms of Developing World Markets
available wind resource, are being exploited. Space is lim-
ited also by the requirements of other activities (e.g. In the developing world, priority should be given to market
nature conservation, agriculture, military). development, before dedicated R&D efforts can be fully
utilised. Following identification of possible markets, mar-
In densely populated coastal countries, offshore sites are ket demand should be built and financiers reassured
the ”new” option, while in mountainous land-locked coun- through large-scale demonstration programmes. In paral-
tries, sites are found in complex terrain, such as funnels, lel, technology oriented R&D programmes should be initi-
hills and mountains. Other areas include those with cold ated, and carried out by industry in cooperation with R&D
and icing climates and, possibly, built up areas. institutes.
R&D Objectives
5.9 Offshore Wind
• Offshore resource assessment, also showing exclu-

sion zones: shipping lanes, grids, pipelines, military, Offshore and onshore R&D should be integrated to a
flora and fauna, water depth. degree. Although the parameters involved in onshore and
• Complex terrain and sensitive area resource assessment. offshore R&D differ, the issues are essentially the same.
• Cold and icing climates resource assessment. For example, loading must be analysed for both offshore
• Test site design conditions and methods to be and onshore turbines and, although different data sets
adapted to mirror potential and new sites. exist in each case, the software necessary to analyse
• Market surveys of developing country markets. them can be of similar type. Such integration would also
• Dedicated WT types for low wind speed inland help avoid potential conflicts of interest between indus-
locations, for high wind speed/high turbulence tries in the coastal countries of northern Europe and
locations, and for cold climates and offshore locations. southern Europe, the former typically having a larger off-
shore resource. In addition, it would reduce duplication in
90
1
VOLUME
R&D efforts. Furthermore, it should be borne in mind that R&D Objectives
although the offshore sector is growing fast, by 2020 it is
estimated that three-quarters of installed capacity will still • Fundamental WT design research (aerodynamics,
be onshore. aeroelasticity, structural design, loads and safety,
control, etc.).
R&D Objectives • Development of test facilities to follow turbine devel-
opments.
• Monitoring of environmental impacts of near and far • Adequate testing and certification of new turbine
offshore projects. technologies, for insurance and finance purposes,
• Potential conflicts of interest: defence, fisheries, such as “0-series” turbines (turbines for areas with
shipping, oil and gas exploration and pipelines, and low wind resource).
sand mining, etc. • Modelling of O&M requirements for large turbines,
• Legal research into offshore ownership in coastal before installation.
waters, Exclusive Economic Zones, etc. • Effective output forecasting methods for large tur-
• Higher tip speed designs, as noise issues are less bines.
significant offshore. • Transport requirements for blades, e.g. built in seg-
• Minimisation of 0&M-related downtime. The distance ments to reduce transportation size.
offshore and the water depth at the site have signifi- • Partnership between fundamental and market driven
cant impacts on O&M. research is essential in ensuring reliability.
• Special designs of systems and components for erec-
tion, access and maintenance of offshore turbines.
• Design studies of systems rated above 5 MW for off-
shore, possibly including multi-rotor systems. 5.11 Summary of R&D Objectives
• Offshore meteorology – short and long-term forecast-
ing; hardware for measurements. The Wind Energy Network, comprising discussions partici-
• Development of alternative, and deep water, founda- pated in by a large cross section of the wind energy sec-
tion structures. tor, puts forward the following initial R&D recommenda-
• Combined wind and wave loading. tions and conclusions. These will be refined and built on
in the final report of this study, to be released in the sum-
mer of 2005.
5.10 Multi-Megawatt Turbines Broad Requirements:

• Long term wind energy R&D programmes
The most important arguments for the development of to increase economic and technical efficiency of the wind energy sec-
larger machines are: for the exploitation of offshore sites, tor, and the European electricity sector as a whole.
where a higher wind resource exists (typically 40% more • Fundamental long term R&D
energy content in the wind compared to onshore); rela- in such fields as aerodynamics & aeroelastics, structural dynamics
tively lower foundation and grid costs; and reduced visual & design, loads & safety, integration into the European electricity
impact on the landscape per unit of installed power. transmission and distribution system, and resource assessment,
Demand drives the trend towards larger machines while and forecasting techniques.
R&D is increasingly expensive and complex as turbines • European standards for use by developers,
increase in size and use more advanced technology, while investors & insurance companies
yet new models are released with increasing frequency. on risk, economic viability, performance, reliability, and O&M of wind farms.
91
• European certification and accreditation systems
for components, turbines and projects. Includes testing of compo-
nents and turbines.
• European codes of practice for access of wind

power into transmission grids.
• Standard European planning procedures for site
assessment
taking into account wind regime, environmental and social impact,
accessibility, etc.
• Full public participation in wind energy exploitation

through exploration of local beneficial impacts.
• R&D into the efficiency and control of very large wind

farms.
• R&D into the dynamics of very large wind turbines.
• Concept development of integral optimised con-
cepts of large wind farms both on land and offshore.
Specific Tasks include:

• Development of tools to identify new sites
such as offshore, remote and complex terrain.
• Remodelling of European-wide grid systems

taking into account large-scale electricity production from wind, and
the benefits of distributed generation.
• Databases of member states’ rules and regulations

including national policies on planning, permitting and environmen-
tal issues.
• Databases of environmental impact issues

to include public opinion surveys, ecological impacts, etc.
• Evaluation of market stimulation programmes &

policy instruments
• Evaluation of European harmonization requirements
in general
to prepare the sector for eventual inclusion in the Internal
Electricity Market.
• Reduction in installation, generation and O&M costs

to optimise the cost/value ratio, and facilitate development of
megawatt and multi-megawatt machines.
Endnotes
1
35th IEA Topical Expert Meeting “Long Term R&D Needs 2000 – 2020” The
Netherlands, March 2001.
92 2
For more information, visit www.wind-energy-network.org
VOLUME 2
COSTS & PRICES
HOOFDTITEL
Acknowledgments
This volume was compiled by Poul Erik Morthorst,

Senior Research Specialist at Risø National Laboratory,
Denmark. Our thanks also to the national wind associa-
tions around Europe for their contributions of data, and to
the other project partners for their inputs.
WIND ENERGY - THE FACTS - COSTS & PRICES
94
VOLUME 2 - COSTS & PRICES: TABLE OF CONTENTS
2
CHAPTER 1 INTRODUCTION 96
CHAPTER 2 COST AND INVESTMENT STRUCTURES 97
VOLUME
CHAPTER 3 OPERATION AND MAINTENANCE COSTS OF WIND POWER 100
CHAPTER 4 THE COST OF ENERGY GENERATED BY WIND POWER 103
CHAPTER 5 DEVELOPMENT OF THE COST OF WIND POWER 105
CHAPTER 6 FUTURE DEVELOPMENT OF THE COST OF WIND POWER 106
CHAPTER 7 COSTS OF CONVENTIONAL POWER PRODUCTION 108
CHAPTER 8 EXTERNAL COSTS OF POWER PRODUCTION 110
95
1 INTRODUCTION
From a European, as well as a global perspective, wind

power is undergoing rapid development. Within the past
10 years the global installed capacity of wind power has
increased from approximately 2.5 GW in 1992 to a little
below 40 GW at the end of 2003, with an annual growth
rate of around 30%. However, only at few sites with high
wind speeds can wind power compete economically with
conventional power production at present.
This section focuses on the cost structures of a wind

power plant, including the lifetime of the turbine and oper-
ation and maintenance costs. Finally, it analyses how the
costs of wind power have developed in previous years and
how they are expected to develop in the near future.
Wind power is used in a number of different applications,

including both grid-connected and stand-alone electricity
production, as well as water pumping. This section analy-
ses the economics of wind energy primarily in relation to
grid-connected turbines which account for the vast bulk of
the market value of installed turbines.
96
2 COST AND INVESTMENT STRUCTURES
The main parameters governing wind power economics

include the following:
• Investment costs, including auxiliary costs for
Figure 2.1: Development of the Average Wind Turbine Size
Sold in Different Countries
2
VOLUME
foundation, grid-connection, and so on.
• Operation and Maintenance (O&M) costs.
• Electricity production/average wind speed.
• Turbine lifetime.
• Discount rate.
Of these, the most important parameters are the wind tur-

bines’ (WT) electricity production and their investment
costs. As electricity production is highly dependent on
wind conditions, choosing the right site is critical to
achieving economic viability. The following sections out-
line the structure and development of capital costs and
efficiency trends of land based WTs.
In general, three major trends have dominated the devel- Source: BTM Consult
opment of grid-connected WTs in recent years:

• The WTs have grown larger and taller – thus, the aver- The wind regime at the chosen site, the hub height of the
age size sold has increased substantially. WTs and the efficiency of production mainly determine
• The efficiency of WT production has increased steadily. power production from the WTs. Thus, increasing the
• In general, investment costs per kW have decreased. height of the WTs has, by itself, yielded a higher power
production. Similarly, the methods for measuring and eval-
Figure 2.1 shows the growth in average size of WTs sold uating the wind speed at a given site have improved sub-
each year in a number of the most important wind power stantially in recent years, thus improving the siting of new
countries. The annual average size has increased signifi- WTs. In spite of this, the fast development of wind power
cantly within the last 10-15 years, from approximately capacity in countries such as Germany and Denmark
200 kW in 1990 to almost 1.5 MW in Germany and implies that most of the good wind sites are, by now,

Denmark in 2002. But, as shown, there is quite a differ- taken. Therefore, any new on-land turbine capacity has to
ence between the individual countries. In Spain, the US be erected at sites with a marginally lower average wind
and the UK, the average size installed in 2002 was speed. It should be added, however, that the replace-
approximately 850-900 kW, significantly below the levels ment of older and smaller WTs with new ones is getting
of Denmark and Germany of 1,450 kW and 1,400 kW increasingly important, especially in countries that have
respectively. The large increase in Denmark from 2001 to taken part in wind power development for a long time, as
2002 was mainly caused by the Horns Reef offshore wind is the case for Germany and Denmark. In 2002, a suc-
farm which came onstream in 2002 equipped with 80 cessful re-powering scheme in Denmark had a substantial
WTs of 2 MW each. impact on market development.
In 2002, the best-selling WTs had a rated capacity of 750- The development of electricity production efficiency owing
1,500 kW and a market share of more than 50%. But WTs to better equipment design, measured as annual energy
with capacities of 1,500 kW and above had a share of 30% production per swept rotor area (kWh/m2) at a specific
and have been increasing their market shares. By the end reference site, has correspondingly improved significantly
97
of 2002, WTs with a capacity of 2 MW and above were over the last few years. Taking into account all the three
becoming increasingly important, even for on-land sitings. mentioned issues of improved equipment efficiency,
improved turbine siting and higher hub height, overall effi- cally, by lower shares for foundation costs and the cost of
ciency has increased by 2% to 3% annually over the last the electrical installation. Thus, these three issues might
15 years. add significant amounts to the total cost of the WT. Cost
components such as consultancy and land rental normal-
Capital costs of wind energy projects are dominated by ly account for only a minor share of additional costs.
the cost of the WT itself (ex works ). Table 2.1 shows the
1
cost structure for a medium sized turbine (850 kW to In Germany, the development of these additional costs has
1,500 kW) sited on land and based on a limited data- been further investigated in a questionnaire carried out by
selection from the UK, Spain, Germany and Denmark. The Dewi (2002), looking at the actual costs for wind turbines
WTs share of total cost is typically a little less than 80%, installed in 1999 and 2001 (Figure 2.2). As shown, all addi-
but, as shown in Table 2.1, considerable variations do tional cost components tend to decrease over time as a
exist, ranging from 74% to 82%. share of total WT costs, with only one exception. The
increase in the share of miscellaneous costs is mostly on
Table 2.1: Cost Structure for a Typical Medium Sized Wind account of increasing prefeasibility costs. The level of auxil-
Turbine (850 kW – 1500 kW) iary costs in Germany has, on average, decreased from
Share of Typical Share approximately 31% of total investment costs in 1999 to
Total Cost of Other Costs approximately 28% in 2001.
% %
Turbine (ex works) 74-82 -
Figure 2.2: Development of Additional Costs (Grid-Connection,
Foundation 1-6 20-25
Foundation, etc.) as a Percentage of Total Investment Costs
Electric installation 1-9 10-15
for German Turbines
Grid-connection 2-9 35-45
Consultancy 1-3 5-10
Land 1-3 5-10
Financial costs 1-5 5-10
Road construction 1-5 5-10
Based on data from Germany, Denmark, Spain and UK for 2001/02.

Of other cost components, dominant ones are, typically,

grid-connection, electrical installation and foundation, but
other auxiliary costs such as road construction could rep-
resent a substantial proportion of total costs. There is
considerable variation in the total level of these auxiliary Source: Dewi (2002).
costs, ranging from approximately 24% of total turbine

costs in Germany and the UK to less than 20% in Spain The total cost per installed kW of wind power capacity dif-
and Denmark. The costs depend not only on the country fers significantly between countries, as exemplified in
of installation, but also on the size of the turbine. Figure 2.3. The cost per kW typically varies from approxi-
mately 900 €/kW to 1,150 €/kW. As shown in Figure 2.3,
Typical ranges of these other cost components as a share the investment costs per kW were found to be almost at
of total additional costs are also shown in Figure 2.2. As the same level in Spain and Denmark, while the costs in
seen, the single most important additional component is the data-selection were approximately 10% to 30% higher
the cost of grid-connection which in some cases can in the UK and Germany. However, it should be noted that
98
account for almost half the auxiliary costs, followed, typi- Figure 2.3 is based on limited data.
Figure 2.3: Total Investment Cost, Including Turbine,
Foundation, Grid-Connection, etc., Shown for Different Turbine
Sizes and Countries of Installation (€/kW)
when analysing WTs constructed to be used in low and
medium wind areas, where the rotor diameter is dimen-
sioned to be considerably larger compared to the rated 2
VOLUME
capacity.
Another reason for the increase in capacity costs is that,

in 2001, the 1,000 kW machine was fairly new. It is usu-
ally the case that, due to economies of scale, a reduction
in price is seen over time.
Figure 2.4: The Development of Investment Costs, Exemplified

by the Case of Denmark for the Period 1989 to 2001
Based on reported data from Germany 2, UK, Spain and Denmark.
Figure 2.4 shows how investment costs have developed,

exemplified by the case of Denmark for the period 1989
to 2001. The data reflect turbines installed in the partic-
ular year shown3. All costs at the right axis are calculated
per swept rotor area, while those at the left axis are
calculated per kW of rated capacity.
Swept rotor area is a good proxy for the turbines’ power

production and this measure is therefore a relevant index
for cost development per kWh. As shown in the figure,

there has been a substantial decline in costs per unit
Right axis: Investment costs divided by swept rotor area (€/m2 in constant 2001 €).
swept rotor area in the period under consideration and for
Left axis: Wind turbine capital costs (ex works) and other costs per kW rated power
all turbines. Thus, overall investment costs by swept rotor (€/kW in constant 2001 €).
area have declined by almost 3% per annum during the

period analysed, corresponding to a total reduction in cost Also, the share of other costs as a percentage of total
of approximately 30% over the past 12 years. costs has decreased. In 1989, almost 29% of total
investment costs were related to costs other than the
Looking at the cost per rated capacity (per kW), the same turbine itself. By 1997, this share had declined to
decline is found in the period 1989 to 1997. approximately 20%. The trend towards lower auxiliary
Surprisingly, however, investment costs per kW have costs continues for the last vintage of turbines shown
increased from the 600 kW machine to the considerably (1,000 kW), where other costs amount to approximately
larger 1,000 kW turbine. The reason is to do with the 18% of the total.
dimensioning of the turbine. With higher hub heights and
larger rotor diameters, the WT is equipped with a rela-
99
tively smaller generator although it produces more elec-
tricity. It is particularly important to be aware of this
3 OPERATION AND MAINTENANCE COSTS OF WIND POWER
O&M costs constitute a sizeable share of the total annu- strictly to O&M and installation, with this proportion split
al costs of a WT. For a new machine, O&M costs might into approximately half for spare parts and the rest equal-
easily have an average share over the lifetime of the tur- ly distributed between labour costs and fungibles. The
bine of approximately 20%-25% of total levellised cost per remaining 40% is almost equally split between insurance,
kWh produced – as long the WT is fairly new, the share rental of land4 and overheads.
might constitute 10%-15% increasing to at least 20%-35%
by the end of its life. Thus, O&M costs are increasingly In Germany, a questionnaire by Dewi (2002) also looked
attracting the attention of manufacturers seeking to devel- into the development and distribution of O&M costs for
op new designs requiring fewer regular service visits and German installations. For the first two years of its life, a
less out-time. WT is normally covered by the manufacturer’s warranty.
Thus, in the German study, O&M costs for the first two
O&M costs are related to a limited number of cost com- years were fairly low at 2%-3% of total investment costs,
ponents: corresponding to approximately 0.3-0.4 c€/kWh. After six
• Insurance years, total O&M costs had increased to constitute a lit-
• Regular maintenance tle less than 5% of total investment costs, which is equiv-
• Repair alent to approximately 0.6-0.7 c€/kWh. These
• Spare parts figures are in line with calculated O&M costs for newer
• Administration Danish turbines (see below).
Some of these cost components can be estimated with Figure 3.1 shows an average over the period 1997 - 2001
relative ease. For insurance and regular maintenance, it is of how total O&M costs were split into six different cate-
possible to obtain standard contracts covering a consid- gories based on the German data from Dewi. The cost of
erable portion of the WT’s total lifespan. On the other buying power from the grid and land rental (as in Spain)
hand, costs for repair and related spare parts are much are included in the O&M cost calculation for Germany.
more difficult to predict. Although all cost components
tend to increase, costs for repair and spare parts are par- Figure 3.1: O&M Costs for German Turbines as an Average
ticularly influenced by turbine age, starting low and over the Period 1997-2001
increasing over time.
Due to the newness of the wind energy industry, only a

limited number of WTs have existed for their expected lifes-
pan of 20 years. Compared to the average size WTs com-
mercially available nowadays, these older WTs are nearly
all small and have, to a certain extent, been constructed
using more conservative, less stringent design criteria
than that used today. Some cost data can be gleaned from
existing older WTs, but estimates of O&M costs should
nevertheless be considered highly uncertain, especially
around the end of a turbine’s lifetime.
Based on experiences from Germany, Spain, the UK and

Denmark, O&M costs are, in general, estimated to be at
a level of approximately 1.2 to 1.5 c€/kWh of produced
100
wind power seen over the total lifetime. Data from Spain
Source: Dewi (2002).
indicate that a little less than 60% of this amount goes
Figure 3.2: O&M Costs Reported for Selected Sizes and Types of Wind Turbines
VOLUME
Source: Jensen et al. (2002).
A recent study in Denmark has analysed the development their life. The picture for the 55 kW machine is patchy,
of O&M costs, insurance costs, etc., including the showing rapidly increasing O&M costs right from the start,
economic and technical lifetime of WTs. Based on a sur- and reaching a fairly high but stable level of approximate-
vey of national wind organisations and an existing data- ly 3-4 c€/kWh after five years.
base, time series for O&M cost components were estab-
lished going back to the early 1980s. Relevant O&M costs Furthermore, the figure shows that O&M costs decrease
were defined to include reinvestments – for example, for newer and larger WTs. The observed strong increase
replacement of blades or gears – if any. Due to the indus- for the 150 kW WTs after 10 years represents only a very
try’s evolution towards larger WTs, O&M cost data for old few machines; therefore, it is not known at present if this

WTs exist only for relatively small units, while data for increase is representative of the 150 kW type or not. For
younger WTs relate primarily to larger units. In principle, turbines with a rated power of 500 kW and more, O&M
the same sample should have been followed throughout costs seem to be under or close to 1 c€/kWh. What is
successive years. However, due to the appearance of new also interesting is that the 225 kW machine over its first
WTs, the scrapping of older ones, and general uncertain- 11 years has O&M costs at around 1-1.3 c€/kWh, close-
ty about the statistics, the sample is not constant over ly in line with estimated O&M costs in Germany, Spain,
time, particularly for the larger WTs. Some of the key the UK and Denmark.
results are shown in Figure 3.2.
Thus, the development of O&M costs appears to be
The figure shows the development of O&M costs for strongly correlated with turbine age. In the first few
selected sizes and types of turbines since the beginning years, the manufacturer’s warranty5 implies a low level
of the 1980s. The horizontal axis represents the age of of O&M expenses for the owner. After the 10th year, how-
the WT while the vertical axis is the total O&M costs stat- ever, larger repairs and reinvestments should be expect-
ed in constant 1999 €. As seen, the 55 kW WTs now have ed: from experience with the 55 kW machine, these are
101
a track record close to 20 years, implying that the first the dominant O&M costs during the last 10 years of the
serial-produced machines are now reaching the end of turbine’s life.
Figure 3.3 shows the total O&M costs as found in the With regard to the future development of O&M costs,
Danish study and details how these are distributed among care must be taken in interpreting the results of Figure
the different O&M categories, according to the type, size 3.3. Firstly, as WTs exhibit economies of scale in terms
and age of the turbine. Thus, for a three-year-old 600 kW of declining investment costs per kW with increasing tur-
machine, which was fairly well represented in the study6, bine capacity, similar economies of scale may exist for
approximately 35% of total O&M costs are for insurance, O&M costs. This means that a decrease in O&M costs
28% for regular service, 11% for administration, 12% for will, to a certain extent, be related to up-scaling of the
repair and spare parts, and 14% for other purposes. In gen- WTs. Secondly, the newer, larger WTs are more opti-
eral, the study found that expenses for insurance, regular mised with regard to dimensioning criteria than the old
service and administration were fairly stable over time, ones, implying an expectation of lower lifetime O&M
while, as mentioned above, costs for repair and spare parts requirements than the older, smaller machines. This
fluctuated heavily. Finally, in most cases, other costs were might, however, imply that newer WTs are not as robust
of minor importance. as older ones and are less capable of dealing with unex-
pected events.
Figure 3.3: O&M Costs as Reported for Selected Types and
Vintages of WTs Taking this reasoning into account, the O&M cost per-
centage for a 10-15 year old 1,000 kW WT could be
expected not to rise to the same level as seen today for
a 55 kW WT of the same age. Most likely, the O&M
costs for newer turbines will be significantly lower than
those experienced to date for the 55 kW WTs. How much
lower future O&M costs go will also depend
on whether the existing trend of up-scaling continues.
Source: Jensen et al. (2002).
Figure 3.3 clearly shows the trend towards lower O&M

costs for new and larger machines. Thus, for a three-year-
old turbine, O&M costs have decreased from approxi-
mately 3.5 c€/kWh for the old 55 kW machine to less
than 1 c€/kWh for the newer 600 kW. The figures for the
150 kW WTs are almost at the same level as the O&M
costs identified in the three countries mentioned above.
That O&M costs increase with turbine age is, again,

102
fairly clear, although not to the same extent as shown in
Figure 3.2.
4 THE COST OF ENERGY GENERATED BY WIND POWER
The total cost per produced kWh (unit cost) is traditional-

ly calculated by discounting and levelising investment and
O&M costs over the lifetime of the WT, divided by the
found in the UK, Ireland, France, Denmark and Norway.
Medium wind areas are generally found at inland terrain in
mid- and southern Europe in Germany, France, Spain, 2
VOLUME
annual electricity production . The unit cost of generation
7
Holland, Italy, but also at inland sites in Sweden, Finland
is thus calculated as an average cost over the lifetime. In and Denmark. In many cases, local conditions significant-
reality, actual costs will be lower than the calculated aver- ly influence the average wind speed at the site. Therefore,
age at the beginning of the life, due to low O&M costs, strong fluctuations in the wind regime are to be expected,
and will increase over the period of WT use. even for neighbouring areas.
The production of power is the single most important Figure 4.1: Calculated Costs per kWh Wind Power as a
factor for calculating the cost per generated unit of power. Function of the Wind Regime at the Chosen Site (Number of
Turbines sited at good wind locations are likely to be prof- full Load Hours)
itable, while those at poor locations may run at a loss. In
this section, the cost of wind-produced energy will be cal-
culated based on a number of assumptions. Due to the
importance of the power production, this parameter will
be treated on a sensitivity basis. Other assumptions
include the following:
• The calculations relate to a new land-based medium-
sized WT of 850-1,500 kW, which could be erected
today.
• Investment costs reflect the range given in section two,
i.e. a cost per kW of 900 to 1,100 €/kW. These costs
are based on data from Spain, UK, Germany and
Denmark.
• O&M costs are assumed to be 1.2 c€/kWh as an aver-
age over the lifetime of the WT.
• The lifetime of the WT is 20 years, in accordance with
most technical design criteria. For assumptions: see above.

• The discount rate is assumed to range within an interval
of 5% to 10% a year. In the basic calculations, an annual Approximately 75% of total power production costs for a
discount rate of 7.5% is used, and a sensitivity analysis WT are related to capital costs, i.e. costs for the WT itself,
of the importance of the interest range is performed. foundation, electrical equipment and grid-connection.
• Economic analyses are carried out as simple national Thus, WTs are a so-called capital-intensive technology
economic ones. No taxes, depreciation, risk premia, compared with conventional fossil fuel-fired technologies
etc. are taken into account. Everything is calculated at such as a natural gas power plant, where as much as 40%-
fixed 2001 prices. 60% of total costs are related to fuel and O&M costs. For
this reason, the cost of capital (discount or interest rate)
The calculated costs per kWh wind power as a function of is an important factor for calculating the cost of wind
the wind regime at the chosen sites are shown in Figure power; cost of capital varies substantially between individ-
4.1 below8. As shown, the cost ranges from approximate- ual EU member states. In Figure 4.2, the costs per kWh
ly 6-8 c€/kWh at sites with low average wind speeds to wind power are shown as a function of the wind regime and
approximately 4-5 c€/kWh at good coastal positions9. In the discount rate, where the latter varies between 5% and
103
Europe, coastal positions such as these are mostly to be 10% a year.
As shown in Figure 4.2, costs range between approxi-
mately 5 and 6.5 c€/kWh at medium wind positions, indi-
cating that a doubling of the interest rate induces an
increase in production costs of 1.5 c€/kWh. In low wind
areas, the costs are significantly higher, 6.5-9 c€/kWh,
while production costs range between 4 and 5.5 c€/kWh
in coastal areas.
Figure 4.2: The Costs of Wind Power as a Function of Wind

Speed (Number of Full Load Hours) and Discount Rate
104
5 DEVELOPMENT OF THE COST OF WIND POWER
The rapid European and global development of wind power

capacity has had a strong influence on the cost develop-
ment of wind power within the past 20 years. To illustrate
Figure 5.1: Total Costs of Wind Power (c€/kWh, Constant
2001 Prices) by Turbine Size
2
VOLUME
the trend towards lower production costs of wind power, a
historical case showing the production costs for different
sizes and vintages of WTs has been constructed. Due to
limited data, it has only been possible to construct this
case for Denmark, though a similar trend was observed in
Germany at a slightly slower pace.
Figure 5.1 shows the calculated unit cost for different

sizes of turbines based on the same assumptions as used
in the previous section. Thus, a 20-year lifetime is
assumed for all turbines in the analysis and an annual dis-
count rate of 7.5% is used. All costs are converted into
constant 2001 prices. Electricity production is estimated
for two wind regimes, a coastal and an inland medium wind
position, respectively. The starting point for the analysis is For assumptions on wind speed, see endnote 10.
the 95 kW machine that was mainly installed in Denmark

during the mid 1980s, followed by successively newer WTs newer WTs might have a track record of only a few years.
(150 kW, 225 kW, etc.), ending with the most recent - the Thus, the newer the WT, the more uncertain the
1,000 kW turbine typically installed around year 2000. It calculations.
should be noted that WT manufacturers, as a rule of
thumb, expect the production cost of wind power to decline In spite of this, Figure 5.1 clearly illustrates the economic
by 3%-5% for each new generation of WTs that they add to consequences of the trend towards larger WTs and
their product portfolio. Further cost reductions are there- improved cost-effectiveness. For a coastal position, for
fore likely to have occurred with the longer production example, the average cost has decreased from approxi-
series of WTs over 1,000 kW. Note that the calculations mately 8.8 c€/kWh for the 95 kW WT (mainly installed in
are performed for the total lifetime (20 years) of the WTs, the mid-1980s) to approximately 4.1 c€/kWh for a fairly

which means that calculations for the old WTs are based new 1,000 kW machine – an improvement of more than
on track records of up to 15 years (average figures), while 50% over a 15 year period at constant 2001 prices.
105
6 FUTURE DEVELOPMENT OF THE COST OF WIND POWER
In this section, the future development of the economics target of 40,000 MW of wind power by year 2010, com-
of wind power is illustrated by the use of experience pared to approximately 23,500 MW installed in the EU
curve methodology. The experience curve approach was at the end of 2002. The European Wind Energy
developed back in the 1970s by the Boston Consulting Association (EWEA) has recently published a target of
Group. Its main feature is that it relates the cumulative 75,000 MW for Europe by 2010. The EU target implies
quantitative development of a product with the develop- an annual growth rate of approximately 7% (a doubling
ment of its specific costs (Johnson, 1984). Thus, if the time of a little more than 10 years), while the EWEA tar-
cumulative sale of a product is doubled, the estimated get requires an annual growth rate of almost 16% (a dou-
learning rate gives you the achieved reduction in specific bling time of 4.8 years). In Figure 6.1 below are shown
product costs. the consequences for wind power production costs
according to the following assumptions:
The experience curve is not a forecasting tool based on • A learning rate between 9% and 17% is assumed,
estimated relationships. It merely points out that if implying that each time the total installed capacity
existing trends are to continue, then we might see the is doubled, then the costs per kWh wind power is
proposed development. It converts the effect of mass reduced by 9%-17%.
production into an effect on production costs, but other • The growth rate of installed capacity is assumed to
casual relationships are not taken into account. Thus, double cumulative installations every 5th, respective-
changes in market development and/or technological ly every 10th year.
break-through within the field might considerably change • The starting point for the development is the cost of
the picture. wind power as observed today, i.e. in the range of 5
to 6 c€/kWh produced for an average medium sized
For a number of projects, different experience curves turbine (850-1,500 kW) sited at a medium wind
have been estimated10, but, unfortunately, most used regime (average wind speed of 6.3 m/s at a hub
different specifications, which means that they cannot height of 50 m).
be directly compared. To get the full value of the experi-
ences gained, not only should the price-reduction of the Figure 6.1: Using Experience Curves to Illustrate the Future
WT (€/kW-specification) be taken into account, but the Development of Wind Turbine Economics until 2010
improvements in efficiency of the WTs production should
be included too. The latter requires the use of an ener-
gy specification (€/kWh) which excludes many of the

mentioned estimations (Neij, 1997 and Neij et al.,
2003). Thus, using the specific costs of energy as a
basis (costs per kWh produced), the estimated progress
ratios in these publications range from 0.83 to 0.91,
corresponding to learning rates of 0.17 to 0.09. That is,
when total installed capacity of wind power is doubled,
the costs per produced kWh for new turbines are
reduced by between 9% and 17%. In this way, both the
efficiency improvements and embodied and disembodied
cost reductions are taken into account in the analysis. Costs illustrated for a turbine installed in a medium wind regime with a
present day production cost of 5 to 6 c€/kWh.
Wind power capacity has developed very rapidly in recent

years, on average approximately by 30% per year during The consequences of applying the above-mentioned
106
the last 10 years. Thus, at present, total wind power results for wind power are illustrated in Figure 6.1. At
capacity is doubled every three years. The EU has set a present, the production costs for a medium sized WT
(850-1,500 kW) installed in an area with a medium wind
speed is approximately 5-6 c€/kWh produced power. If a
doubling time of total installed capacity of 10 years is 2
VOLUME
assumed, the cost interval in 2010 would be approxi-
mately 4.4 to 5.6 c€/kWh. A doubling time of five years
only would imply a cost interval in 2010 of 3.9 to 5.2
c€/kWh. If the WT is located in a coastal area with a high-
er wind speed (average wind speed of 6.9 m/s at a height
of 50 m), the costs per kWh produced in 2010 could be
as low as 3.1 to 4.4 c€/kWh in the case of a five-year
doubling time of total installed capacity.
107
7 COSTS OF CONVENTIONAL POWER PRODUCTION
The cost of conventional electricity production is deter- To get a comparable picture, “Projected Costs of
mined by three components: Generating Electricity - Update 1998” (OECD/IEA, 1998)12
• Fuel cost has projected the costs of electricity generation with
• O&M costs state-of-the-art coal-fired and gas-fired base load power
• Capital cost plants, given the following common assumptions:
• Plants are commercially available for commissioning
When conventional power is substituted by wind power, by the year 2005
the avoided cost depends on the degree to which wind • Costs are levellised using a 5% real discount rate and
power substitutes each of the three components. It is a 40-year lifetime13
generally accepted that implementing wind power avoids • 75% load factor
the full fuel cost and a considerable portion of O&M costs • Calculations are carried out in constant 1996 US$,
of the displaced conventional power plant. The level of converted to € 2001 prices
avoided capital costs depends on the extent to which wind
power capacity can displace investments in new conven- The OECD/IEA calculations were based on data made
tional power plants and is thus directly tied to the capacity available by OECD member countries. Costs related to
credit of wind plant. electricity production, pollution control and other environ-
mental protection measures were included in the calcu-
The capacity credit will depend on a number of different lated generation costs, while general costs, such as cen-
factors: among these is the level of penetration of wind tral overheads, transmission, and distribution costs were
power and how the wind capacity is integrated into the excluded. Losses in transmission and distribution grids
overall energy system and market. In general, for mar- were also not taken into account. Fuel price developments
ginal levels of wind penetration, the capacity credit for were projected in accordance with national assumptions.
WTs is close to the annual average capacity factor. Thus, Figure 7.1 shows the costs of conventional power as pro-
25% is considered to be a reasonable capacity credit for jected by OECD/IEA, updated to 2001 € prices.
wind power when the volume of wind electricity is less
than 10% of total electricity production11. This capacity Figure 7.1: Projected Costs of Conventional Power
credit declines as the proportion of wind power in the sys- (2001 c€/kWh)
tem increases; but even at high penetrations a sizeable
capacity credit is still achievable if the management and
future development of grid infrastructure are conducted

with a view to the expected increase in distributed gen-
eration from wind power and other renewable energy
sources.
The capacity credit of wind power depends heavily upon

the structure of power markets. Studies of the Nordic
power market, NordPool, show that the cost of integrat-
ing intermittant wind power is, on average, approximate-
ly 0.3-0.4 c€/kWh wind power at the present level of
wind power capacity (20% in Denmark). Under existing
transmission and market conditions, and as in the case
of capacity credit, these costs are supposed to increase
Source: OECD/IEA (1998), updated to 2001 € prices.
with higher levels of wind power penetration.
108
The figures are based on the above cost data from
OECD/IEA (1998) for a selected number of countries and
power technologies. The costs for the conventional tech-
Figure 7.2: Projected Avoided Costs of Conventional Power
Compared to Costs for Wind Electricity (2001 c€/kWh),
Assuming 25% Capacity Credit for Wind Power 2
VOLUME
nologies were originally stated in 1996 US$, but at the
aggregate level converted to 2001 € prices. Thus, con-
siderable uncertainty exists for the costs shown owing to
changes in exchange rates, national differences in infla-
tion rates and different national assumptions on fuel price
development. Finally, although no major changes are
expected, investment costs for conventional power plants
may have changed quite substantially since 1998.
Figure 7.2 shows those costs of conventional power which

are avoidable through wind electricity, assuming that all
conventional fuel and O&M costs are avoided and that
wind power is assigned a capacity credit of 25%. For
example, in Spain, for each kWh of electricity generated
by wind power which displaces a kWh of gas power,
Source: OECD/IEA (1998), updated to 2001 € prices.
approximately 5.2 c€/kWh are saved in gas fuel, O&M
costs and displaced capital costs. Therefore, if a wind tur-
bine could be installed in Spain at an average cost below electricity markets increasingly move away from cen-
5.2 c€/kWh, this would make wind power economically tralised generation planning and towards increased
competitive in comparison with new gas-fired power plant competition. Much of wind energy’s future competitive-
in Spain. For comparative purposes, the estimated total ness will depend on short-term wind predictability and
costs (including capital costs and calculated using an on the specific conditions which develop for bidding
annual discount rate of 5%) for a medium sized on-land into short-term forward and spot markets at the power
turbine at average coastal and inland sites are also exchange.
shown (4.2 and 4.8 c€ per kWh, respectively ). As shown
14
in Figure 7.2, under the assumption of a 25% capacity Finally, it should be stressed that the above-mentioned

credit for wind energy, a medium sized turbine is actually costs of conventional generation are based upon national
approaching competitiveness in terms of direct costs in a assumptions on the development of fossil fuel prices
number of countries, compared to technologies based on which, of course, are subject to significant uncertainties.
coal and gas. As discussed by Awerbuch (2003), these uncertainties
relating to future fossil fuel prices imply a considerable
Of course, if a higher capacity credit for wind than 25% is risk for future generation costs of conventional plants,
assumed, this would raise the avoided costs of conven- while the costs per kWh generated by wind power are
tional technologies and thus improve wind’s competitive- almost constant over the lifetime of the turbine when first
ness. Similarly, if a lower capacity credit were assumed, installed. Thus, although wind power today might be more
this would make wind power less economically competitive. expensive than conventional power technology per kWh, it
may nevertheless take up a significant share in investors’
Capital costs are more important for coal based power power plant portfolios, taking on the role of hedging
than for natural gas fired plants, and therefore assump- against unexpected rises in future prices of fossil fuels.
tions about wind’s capacity credit are particularly Thus, the constancy of wind power costs justifies a rela-
109
important regarding coal plants, as shown above. tively higher cost per kWh compared to the more risky
However, this importance may change in the future as future costs of conventional power.
8 EXTERNAL COSTS OF POWER PRODUCTION
The competitiveness of wind power is dependent on the Figure 8.1: An Illustrative Example of the Social Cost of Energy
particular market conditions where wind developments are
placed. Figure 7.2 shows that wind costs are marginally
higher than conventional power technologies such as coal
and natural gas. However, it is generally appreciated that
wind energy and other renewable energy sources have
environmental benefits when compared to conventional
electricity generation. But are these benefits fully reflected
in the market prices of electricity? And, on the other hand,
is conventional power generation charged for the environ-
mental damage caused by polluting emissions?
This section deals with these questions in order to esti-

mate the hidden benefits/costs of the different electricity
production activities not taken into consideration by the
existing pricing system. To establish a fair comparison of
the different electricity production activities, all internal
and external costs to society need to be taken into
account.
Hence, it is important to identify external effects of differ-

ent energy systems and to monetise their costs, especial-
ly if these are of a similar order of magnitude as the inter-
nal costs of energy and if the external costs vary substan-
Endnotes
tially between competing energy systems such as conven-
1
“Ex works” means that no site work, foundation, or grid connection costs are includ-
tional electricity generation and wind energy. The question
ed. Ex works costs include the turbine as provided by the manufacturer, including the
arises whether the inclusion of external costs – the exter- turbine itself, blades, tower, and transport to the site.
2
For Germany, an average figure for the installed capacity in 2001 is used.
nalities – in the pricing system (internalisation) could have 3
All costs are converted to 2001 euros.
4
an impact on the competitive situation of different elec- In Spain the rental of land is seen as an O&M cost.
5
In the Danish study, only the costs to be borne by the wind turbine owner are
tricity generating technologies. Results from different included, i.e. costs borne by the manufacturer in the warranty period and subse-
studies are shown in Figure 8.1. The external costs of 6
quently by the insurance company are not taken into account.
The number of observations was, in general, between 25 and 60.
conventional power systems make these technologies 7
The cost of wind energy should not be confused with the price of wind power. The
less competitive in comparison with wind energy as the latter relates to the amount per kWh the wind turbine owner receives for the power
he/she sells.
externalities are included to take account of the social 8
In the figure, the number of full load hours is used to represent the wind regime.
Full load hours are calculated as the turbine’s average annual production divided
cost of energy production. The internal cost of wind ener-
by its rated power. The higher the number of full load hours, the higher the wind
gy is practically unchanged by including the externalities. turbine’s production at the chosen site.
9
In this context, a coastal position is defined as a site with an average wind speed
of 6.9 m/s at a height of 50 m above the ground. Correspondingly, the medium
Volume 4 ‘Environment’, presents a more detailed analy- and low wind sites have average wind speeds of 6.3 and 5.4 m/s at a height of
50 m.
sis of the external cost of energy as well as the latest 10
See, for instance, Neij (1997), Neij (1999), Milborrow (2003) or Neij et al. (2003).
11
results obtained for different generation technologies. In EPRI (1997) suggests that wind turbines located in highly windy areas could
achieve capacity factors of 40%-45% by 2005.
addition, an analysis focusing on the avoidable external 12
This seems to be the most recent update of the projected costs of generating elec-
costs of wind energy for European member states, along tricity available.
13
National assumptions on plant lifetime might be shorter, but calculations were
with an estimation of the total avoided external costs, adjusted to 40 years.
110 14
Average wind power production costs calculated using an annual 5% discount rate
are also introduced.
as shown in chapter 4.
VOLUME 3
INDUSTRY & EMPLOYMENT
HOOFDTITEL
Acknowledgments
The employment chapters of this volume were compiled

by Henrik Jacobsen, Senior Researcher at Risø National
Laboratory, Denmark. Our thanks also to the national wind
associations around Europe for their contributions of
data, and to the other project partners for their inputs.
WIND ENERGY - THE FACTS - INDUSTRY & EMPLOYMENT
112
VOLUME 3 - INDUSTRY & EMPLOYMENT: TABLE OF CONTENTS
CHAPTER 1 INDUSTRY STATUS 114

1.1.1 New Developments 114
1.1.2 Market Changes 114
1.2 Demand Drivers 115
3
1.2.1 Market Types 115
1.2.2 Political Risk 116
1.3 The Onshore Market 116
1.3.1 Market Status (Cumulative Market) 116
VOLUME
1.3.2 Market Growth (Annual Market) 119
1.4 Offshore Market 121
1.5 Domestic Market 123
1.6 Export Market 124
1.7 Industry Segments 125

1.7.1 Manufacturers - Megawatt Scale 125
1.7.2 Manufacturers - Small Turbines 125
1.8 Wind Farm Developers 126
CHAPTER 2 EMPLOYMENT IN THE WIND TURBINE SECTOR 127
2.1 Direct and Indirect Employment Effect 127

2.2 Direct Employment in Wind Turbine Manufacturing 127
2.3 Direct Employment in Wind Turbine Installation 128
2.4 Direct Employment in Maintenance Activities 129
2.5 Indirect Employment in Manufacturing 130

2.6 Indirect Employment in Wind Turbine Installation 130
2.7 Total Direct and Indirect Employment (Manufacturing, Installation and Maintenance in Europe) 131
CHAPTER 3 EMPLOYMENT PREDICTION AND METHODOLOGY 133
3.1 Direct and Indirect Employment 133

3.2 Input-Output Methodology 133
3.3 Projection/Prediction Parameters 135
3.4 Sensitivity – Main Parameters 135
CHAPTER 4 SCENARIOS FOR EMPLOYMENT IN THE WIND TURBINE SECTOR 137
4.1 Projection of Employment Based on Wind Energy Installation in Europe and Globally 137
CHAPTER 5 DEVELOPMENT AND INNOVATION 139
113
1 INDUSTRY STATUS
1.1 Introduction (1 gigawatt, GW). With a more limited definition - that of a

single location - then the King Mountain project in Texas
Since the last Wind Energy - The Facts report published in is the largest single installation, at 278 MW. Continued
1999, the European wind energy industry has made sig- activity in the US suggests that such large projects,
nificant progress. There are several ways of monitoring although unlikely to become commonplace, will certainly
this progress, such as measuring electricity output in MW occur more frequently.
or kW hours. However, the usual method is to use a
measurement of installed capacity, so this chapter Growth in wind farm size has, to a degree, followed growth
demonstrates national markets and their growth in terms in wind turbine (WT) size. These same large wind farms can-
of MW capacity installed. not, however, be installed in all markets, and space require-
ments in some hitherto very active European markets will
Wind experienced a surge of growth in California in the ultimately limit growth onshore. This constraint has now
1980s thanks to a combination of state and federal ener- been recognised by several northern European govern-
gy and investment tax credits . From 1980 to 1995,
1
ments, and active plans for the development of offshore
around 1,700 MW of wind capacity was installed and, projects have begun. Indeed, the construction of Denmark’s
although there were some turbines of poorer quality, the second large-scale offshore projects has been completed
boom period provided a major export market for European with a capacity of 160 MW.
manufacturers, and did much to establish the credibility of
the industry. Since then, Europe has turned the tables 1.1.2 MARKET CHANGES
and consolidated its position as the global market leader.
Within Europe, certain countries are particularly strong: Structural changes to the industry have taken place in
the top five in terms of installed capacity being Germany, recent years, and new companies have arrived. The
Spain, Denmark, The Netherlands and Italy. increased size of wind farms, growth of business at
approximately 30% per annum, improved technology and,
1.1.1 NEW DEVELOPMENTS in particular, improved turbine availability, have all allowed
the wind energy business to be considered seriously by
There have been significant changes in the industry over main players in the conventional power industry: Shell has
the last five years, and it is still in a state of considerable formed a wind energy subsidiary, Shell Wind Energy; and
flux. Some major structural changes are taking place with- the Enron subsidiary, Enron Wind Corporation, was pur-
in the sector as it matures; these are significant for any chased by General Electric to form GE Wind Energy.
company wishing to participate in the wind industry over
the long term. SIIF, a French company 35% owned by Electricité de
France, is emerging as a major player with global aspira-
The unit size, both of individual turbines and of wind farms tions, recently purchasing the US operations and mainte-
has grown significantly. Five years ago, a wind farm of 20 nance provider and developer enXco. Substantial con-
MW would be considered large. Now, particularly as the struction by the Italian conglomerate EDENS, as well as
North American market reopens, the Spanish market con- ongoing activity by FPL and most of the Spanish utilities,
tinues to grow, new markets open, and the offshore mar- all underline the nature of today’s wind developers, as
ket takes off, very large wind farms are being introduced, compared with those of the previous decade, which tend-
in the scale of hundreds of MW. ed to be small and independent.
The definition of a wind farm is somewhat vague, but if a The last year has also seen the separation of Gamesa
wind farm is considered, somewhat artificially, as a single Eolica, the leading supplier in the Spanish market, from
114
financing, then the year 2001 saw a major breakthrough, its Danish partner, Vestas. This step has produced a
with a single financing in Spain of more than 1,000 MW major new competitor worldwide. The Indian company,
Suzlon, has also emerged on the world market as a tur- ket developed in the mid 1990s shows that there can be
bine supplier. a danger in too rapid development of a new technology.
Quality problems arose, both in initial manufacture, which
Over the past decade, the wind turbine manufacturing were later revealed in severe storms in Gujarat and also,
industry has become increasingly concentrated. This was perhaps more severely, in inadequate preparation of pro-
emphasised by the announcement of a merger in jects. In particular, inadequate measurement was carried
December 2003 between the world's largest and third
largest manufacturers Vestas Wind Systems and NEG
Micon. If approved, they will have a combined global mar-
ket share of approximately 35%.
out of the wind resource on project sites. Whereas in
industrialised countries it is normal to have several years’
wind data before a project is built, in India there was, at
best, only a few months’ and, in some cases, no data at
3
VOLUME
all. This is a dangerous position.
1.2 Demand Drivers Another major flaw in the Indian framework of the 1990s,
was the existence of a subsidy to wind turbine owners
1.2.1 MARKET TYPES which was based on the rated capacity of the wind tur-
bines rather than an incentive to optimise production of
The market may be split into two separate segments, the the renewable electricity. That proved problematic
“conscience” market and the “needs” market. because a subsidy was given whether or not production
was efficient. This incentive resulted in poor siting of wind
Conscience Market turbines, and manufacturers followed customer demands
to use very large generators, which improved project prof-
In the conscience market, the driver for development of itability but reduced production and also attracted manu-
wind energy has hitherto been a desire to produce elec- facturers with highly dubious products, which gave the
tricity by cleaner means. This has been supported by entire technology a bad name. India has since corrected
incentives, described in volume 5 chapter 1, and, hence, the inherent flaws of its incentive scheme and the market
has been essentially a political market. There are signs, has started to develop again.
however, that as the cost of wind energy continues to fall

there are some applications for which wind energy is com- With the experience of these mistakes, it is noticeable
petitive in its own right and the nature of these markets that the second round of incentives in the Indian market
will therefore change. is rather different, and the market is presently being much
more tightly controlled, both in terms of development qual-
Needs Market ifications and also in quality control of the turbines pro-
duced. As a result, wind energy in India now has a sus-
For the needs market, motivations are somewhat differ- tainable future. Other countries in this category presently
ent. Such markets are characterised by a growing, and considering serious development of wind energy include
unsatisfied, need for energy, and a limited amount of new Brazil, Tunisia, China, Egypt, Morocco, the Philippines,
generating capacity coming online. In these markets, wind Turkey and Vietnam.
energy is considered as one of several alternatives. Given
the relative ease with which wind technology adapts to dif- The key difference between the two market types is that,
ferent countries and requirements, and the relatively short for the conscience market, comparisons are always made
time between initiating construction and delivering power, between wind costs and, say, combined cycle gas costs
it has become the most attractive alternative for some. whereas, for the needs market, the comparison may be
the cost of having power rather than not having power.
115
A good example of this latter category is wind power’s ini- The conditions for commercial viability are, therefore,
tial development in India. The way in which the Indian mar- quite different.
1.2.2 POLITICAL RISK
There are many examples of political uncertainty in both cantly with increases in volume, and to create significant
the needs and the conscience markets: a few examples employment.
are given here.
The wind power industry has experienced the risk of rely-
The risks associated with establishing a wind farm in a ing too heavily on one single market. In the middle of the
developing country are similar to those encountered with 19980s, manufacturers were entirely relying on the
any other form of development. There is political risk, Californian market. In the 1990s, they found themselves
technology risk and financial risk. There is also, in the equally dependant on the Indian market for wind power
case of wind energy, an additional risk that the technolo- technology. When those two markets collapsed due to
gy will be copied and an indigenous product developed political and economic turmoil, many manufacturers went
without a license agreement. bankrupt. Today, the global wind power market is more
geographically spread. However, the three largest mar-
In both the conscience and the needs market there is kets still accounts for app. 70% of the global market
political risk. The cost of wind energy has declined sharply (2002).
over the last decade, but still requires some form of
incentive to encourage its widespread development. This
incentive is inevitably political in nature although it may be 1.3 The Onshore Market
drafted in any number of different ways, from tax credits
to premium prices, to tradable green certificates. If the 1.3.1 MARKET STATUS (CUMULATIVE MARKET)
political attitude changes in any one of the active coun-
tries, the market in that country can undergo radical alter- The progress of wind power around the world in recent
ation. This has, indeed, been seen on many occasions. years has been impressive. By the end of 2003 more than
For example, in Germany, there was some uncertainty in 39,400 MW of electricity-generating WTs were operating
1999 about the planning regime within which wind energy in 50 countries. Of these, more than 28,000 MW were
developments were built. Whilst that uncertainty was installed in the EU, enough to meet 2.4% of EU-15 elec-
being clarified, the market declined, but it has since tricity demand.
recovered and achieved three record years. When US
President Carter left office in the early 1980s, the market Europe dominates the global wind market, with European
stopped overnight. Early in 2003, the premium price in manufacturers controlling 90% of the global market in
Spain fell, although this was coupled with an increase in 2002. The most successful markets for wind power in
pool price and hence the composite kWh (kilowatt hour) recent years have been Germany, Spain and Denmark.
price was almost unchanged. Nevertheless, the risk of Outside the EU, India and the US are leading markets, but
price change was accentuated. over 70% of the market remains in the EU. Between 1992
and 2002, cumulative installed capacity multiplied 27
In the case of the EU, underpinning individual domestic times. 2002 itself was a record year for EU installations,
policies for renewable energy and environmental policy, with over 5,800 MW of new capacity.
there is strong support from both the European
Commission and the Parliament. Targets for renewable Wind energy is now established across parts of western
energy have been set and enshrined in EU law. Hence, at and southern Europe, and installations are beginning to
a higher level of policy, significant support exists for ongo- take off in the new member states of central and eastern
ing development of renewables, including wind. Wind Europe, for example in Latvia and Poland. Figure 1.1
116
energy is particularly well received in this context as it has shows installed MW capacity in the EU-15, compared with
demonstrated an ability both to reduce its price signifi- that installed worldwide.
The curve demonstrates the percentage of global installed grown by an average 30% per annum and, by the end of
capacity installed in the EU-15: in 1990 this figure stood 2002, EU-15 countries represented 74% of installed
at 25%. Over the last five years, the European market has capacity worldwide.
Figure 1.1: EU-15 and Global Cumulative Installed Wind Capacities (MW)
VOLUME
Table 1.1: EU-15 Cumulative Installed Capacities (MW) by Country
Country 1995 1996 1997 1998 1999 2000 2001 2002 2003
Austria 5 10 20 30 34 77 94 140 415

Belgium 4 4 4 6 6 13 31.6 35 68
Denmark 619 842 1,129 1,443 1,771 2,417 2,489 2,889 3,110
Finland 6 7 12 17 39 39 39 43 51
France 3 6 10 19 25 66 93 148 239
Germany 1,132 1,552 2,081 2,875 4,442 6,113 8,754 11,994 14,609
Greece 28 29 29 39 112 189 272 297 375
Ireland 7 11 53 73 74 118 124 137 186
Italy 32 70 103 180 277 427 682 788 904
Luxembourg 2 2 2 9 10 10 15 17 22
Netherlands 249 299 319 361 433 446 486 693 912
Portugal 8 19 38 60 61 100 131 195 299
Spain 133 249 512 834 1,812 2,235 3,337 4,825 6,202
Sweden 69 103 122 174 220 231 293 345 399
UK 200 273 319 333 362 406 474 552 649
EU-15 Total 2,497 3,476 4,753 6,453 9,678 12,887 17,315 23,098 28,440
117
Table 1.2: New Member State Cumulative Installed Capacity (MW)
1995 1996 1997 1998 1999 2000 2001 20021 20031

Cyprus 2
0 0 0 0 0 0 0 2 2
Czech Republic 3
11 11 11 11 11 6 7 3 10
Estonia 4
0 0 0 0 0 0 0 2 3
Hungary 5
0 0 0 0 0 0 1 3 3
Latvia 6
1 1 1 1 1 1 1 24 24
Lithuania n/a n/a n/a n/a n/a n/a n/a 0 0
Malta n/a n/a n/a n/a n/a n/a n/a 0 0
Poland 7
1 1 2 2 3 4 22 27 57
Slovakia n/a n/a n/a n/a n/a n/a n/a 0 3
Slovenia n/a n/a n/a n/a n/a n/a n/a 0 0
EU New Members Total 13 13 14 10 17 18 33 61 102
1
All data for 2002 - 2003 from EWEA
2
Data for 1995 - 2001 from Cyprus Institute of Energy
3
Data for 1995 - 2001 from Czech Society for Wind Energy
4
Data for 1995 - 2001 from Latvian Wind Energy Association
5
Data for 1995 - 2001 from Horvath Engineering, Hungary
6
7
Data for 1995 - 2001 from Vis Venti Association for Supporting Wind Energy, Poland
Figure 1.2: Top 10 Cumulative Global Market Shares in 2002 (MW)

Country Cumulative Installed % Share

Capacity End, 2002
Germany 12,001 41.4
Spain 4,830 16.7
USA 4,685 16.2
Denmark 2,880 9.9
India 1,702 5.9
Italy 785 2.7
Netherlands 688 2.4
UK 552 1.9
China 468 1.6
Japan 415 1.4
Total 29,006 100.0
118
1.3.2 MARKET GROWTH (ANNUAL MARKET)
Germany, Spain and Denmark accounted for almost 80% electricity needs from wind power. Spain followed with
of the wind power capacity installed in Europe in 2003 1,377 MW, to achieve a total of 6,202 MW. Denmark
(see Table 1.3). With 2,645 MW, Germany accounted for installed 243 MW to reach 3,110 MW, and the industry
49% of the installed capacity, reaching a total of 14,609 association expects that wind will meet approximately
MW by the end of 2003, enough to meet 6% of national 20% of the country’s electricity needs in 2004.
Figure 1.3: EU-15 and Global Annually Installed Wind Capacity (MW) 3
VOLUME
119
Table 1.3: EU-15 Annually Installed Capacity by Country
Country 1996 1997 1998 1999 2000 2001 2002 2003

Austria 5 10 10 4 43 17 46 276
Belgium 0 0 2 0 7 19 3 33
Denmark 223 287 314 328 646 72 506 243
Finland 1 5 5 22 0 0 4 8
France 3 4 9 6 41 27 55 9
Germany 420 529 794 1,567 1,671 2,641 3,247 2,645
Greece 1 0 10 73 77 83 25 78
Ireland 4 42 20 1 44 6 13 49
Italy 38 33 77 97 150 255 106 116
Luxembourg 0 0 7 1 0 5 2 5
Netherlands 50 20 42 72 13 40 222 226
Portugal 11 19 22 1 39 31 64 107
Spain 116 263 322 978 423 1,102 1,488 1,377
Sweden 34 19 52 46 11 62 52 54
UK 73 46 14 29 44 68 87 103
EU-15 Total 979 1,277 1,700 3,225 3,209 4,428 5,913 5,411
Table 1.4: New Member State Annually Installed Capacity (MW)
1996 1997 1998 1999 2000 2001 2002 20031

Cyprus 2 0 0 0 0 0 0 0 0
Czech Republic 3 0 0 0 0 -5 0 0 7
Estonia 4 0 0 0 0 0 0 2 1
Hungary 5 0 0 0 0 0 1 1 0
Latvia 6 0 0 0 1 0 0 20 0
Lithuania n/a n/a n/a n/a n/a n/a n/a 0
Malta n/a n/a n/a n/a n/a n/a n/a 0
Poland 7 0 1 0 1 1 18 5 30
Slovakia n/a n/a n/a n/a n/a n/a n/a 3
Slovenia n/a n/a n/a n/a n/a n/a n/a 0

EU New Members Total 0 1 -4 7 1 15 29 41
1
All data for 2003 from EWEA
2
Data for 1995 - 2002 from Cyprus Institute of Energy
3
Data for 1995 - 2002 from Czech Society for Wind Energy
4
5
Data for 1995 - 2002 from Horvath Engineering, Hungary
6
7
Data for 1995 - 2002 from Vis Venti Association for Supporting Wind Energy, Poland
120
Table 1.5: Offshore Installed Capacities
Location Country Installed Year Cumulative Installed

Capacity (MW) Capacity (MW)
Vindeby Denmark 5 1991 5
Lely Ijsselmeer, The Netherlands 2 1994 7
Tunø Knob Jutland, Denmark 5 1995 12
3
Dronton IJsselmeer, The Netherlands 17 1997 29
Bockstigen-Valor Gotland, Sweden 3 1998 32
Blyth United Kingdom 4 2000 36
Middelgrunden Copenhagen, Denmark 40 2000 76
Utgrunden Sweden 10 2000 86
VOLUME
Yttre Strengrund Sweden 10 2001 96
Samsø Denmark 23 2003 119
North Hoyle United Kingdom 60 2003 179
Horns Rev Denmark 160 2003 339
Nysted Denmark 158.4 2003 497.4
Arklow Bank Ireland 25 2003 522.4
Figure 1.4: Top 10 Global Annual Market Shares in 2002 (installed MW per annum)
Country Annual 2002 % Share 2002

Germany 3,247 50.2
Spain 1,493 23.1

Denmark 497 7.7
USA 410 6.3
Netherlands 217 3.4
India 195 3.0
Japan 140 2.2
Italy 103 1.6
UK 87 1.3
Norway 80 1.2
1.4 Offshore Market first major offshore wind farm installed at Horns Rev off
the Danish coast.
To date, only Europe has installed wind capacity offshore,
although projects are planned for the US. Fourteen projects Figure 1.5 demonstrates the great leap made in 2003,
are now operating in four EU countries with a total capacity chiefly thanks to the installation of the three Danish wind
of 522 MW. farms - Horns Rev, Nysted and Samsoe, while Table 1.6
121
below gives an impression of the potential for offshore
Since the first offshore turbines were installed in 1991, wind farms in the EU up to 2006 - a little under 9 GW
development has been gradual. 2003 saw the world’s (9,000 MW). Worldwide, the database prepared by
Douglas-Westwood show the current planned offshore Table 1.6: EU-25 Planned Offshore Installations up to 2006
windfarms to total around 50 GW up to 2010. It should be
Country Name/Location Date MW
noted that these figures are for identified projects alone
Germany Nördlicher Grund phase 1 2004 360.0
and those currently under development - and the potential
Belgium Thornton Bank 2005 216.0
exists for much more.
Finland Kokkola 2005 207.0
Germany Borkum Riffgrund phase 1 2005 231.0

As demonstrated by the existing level of its onshore
Germany Nordergründe 2005 266.0
installed capacity (57 MW) Poland is moving ahead with
Germany Sky 2000 2005 150.0
wind installations faster than the other new member
Ireland Arklow Bank phase 2 2005 216.0
states, and it is the only new member state to have an off-
Netherlands Near Shore wind park 2005 99.0
shore installation in the planning phase (see Table 1.6).
Netherlands Q7-WP 2005 120.0
Poland Bialogora 2005 120.0

Figure 1.5: Annual and Cumulative Offshore Capacity,
Sweden Fladen 2005 140.0
1991 - 2003
UK Barrow 2005 108.0
UK Cromer 2005 108.0
UK Gunfleet Sands 2005 108.0
UK Lynn 2005 108.0
UK Robin Rigg (Solway Firth) 2005 108.0
France Ile de Groix 2006 100.0
Germany Adlergrund phase 1 2006 320.0
Germany Amrumbank West 2006 288.0
Germany Arkona-Becken Südost phase 1 2006 195.0
Germany Beltsee 2006 249.0
Germany Borkum Riffgrund West phase 1 2006 280.0
Germany Butendiek 2006 240.0
Germany DanTysk phase 1 2006 400.0

Germany He Dreiht 2006 535.5
Germany Meerwind phase 1 2006 262.5
Germany Nordsee Ost phase 1 2006 400.0

Germany North Sea Windpower phase 1 2006 166.5
Germany Riffgat 2006 198.0
Germany Sandbank 24 phase 1 2006 360.0
Germany Uthland 2006 400.0
Ireland Arklow Bank phase 3 2006 263.0
UK Shell Flat 2006 324.0
Other Less than 100 MW in size - 1072.8
Total - - 8719.3
Source: The World Offshore Wind Database, Douglas-Westwood (2003).
122
1.5 Domestic Market the present beneficial tariff system. A notable achievement
of the wind industry in Spain is the very high level of
Europe continues to play a dominant role, both in the devel- Spanish manufacturing content, which is now starting to
opment and manufacture of WTs and in the exploitation of export elsewhere. A notable event in 2001 was the sepa-
its wind resource through the development of wind farms. ration of the market leader in Spain, Gamesa, from its pre-
The top three EU-15 markets - Germany, Spain and vious joint venture partner, Vestas of Denmark. Gamesa
Denmark - as well as promising new member states mar-
kets are discussed below.
Germany
Eolica is now competing on the international market.
Galicia, in the north west led the country in terms of

installed capacity in 2002 - installations in this region alone
3
VOLUME
amounted to 341.5 MW. Castilla La Mancha follows, while
Germany continues to be the leader in terms of cumulative Aragón, Navarra and Castilla y León are also thriving.
and annual MW installed. 2003 saw a notable drop in the Spain’s wind power boom has been spurred on by a thriv-
annual MW installed from 3,247 to 2,645 in 2003, howev- ing turbine manufacturing industry. Three of the country’s
er that rate of decline is not forecast for the coming years manufacturers, Gamesa, Made and Ecotecnia are among
(see Table 1.3). The wind sector is large, and the political the world’s top 10, with Gamesa achieving fourth position
commitment to wind energy remains strong - the revision of in the global league table, according to the latest report
the EEG is likely to reflect that, albeit with some tougher from BTM Consult (2003).
conditions, for example for lower wind speed sites at inland
sites. The decline of the onshore market will occur this The situation in Spain is rather different to that in Germany,
decade, and an increase in offshore wind farms and the since the winds are higher and there is more space avail-
rise of a repowering market. This changing market will cre- able. Commentators expect to see the present level of the
ate ups and downs in the figures, but Germany remains the market sustained in Spain until it is limited by other con-
major wind market in Europe this decade. cerns, possibly that of grid connection. This limitation is
already seen in some large-scale projects currently being
The presence of a strong German industry is also impor- developed.
tant, as is the potential for repowering. Furthermore,

Germany has significant offshore plans, the development of Denmark
which may continue to buoy up its domestic market, even
as the onshore market begins to decline over the next Denmark continues to dominate the manufacturing side of
decade. the industry worldwide, and has also benefited from a buoy-
ant home market. Denmark is a relatively small country with
Spain a high population density and already has a high level of
penetration: 20% of Danish electricity consumption is cov-
Spain has been the next most active market after Germany. ered by wind power. In 2003, some 243 MW were installed.
Now, more than 4% of Spain’s electricity is supplied by wind A broad majority of the Danish Parliament has agreed on a
power. As is to be expected, the most energetic sites were long-term strategy for wind power - ”Energy 21” from 1996.
largely used in the early days of development and, hence, The goal is that by 2030, 40-50 percent of the Danish elec-
those presently being developed do not have the same tricity consumption should be covered with power from wind
level of resource. Nevertheless, the cost of installation is turbines. Denmark is following a strategy on the one hand
reducing, and this has allowed the exploitation of less ener- to expand wind power offshore, and on the other hand to
getic sites to be undertaken with a similar return on capital. replace some of the smaller, less productive turbines that
Informed commentators consider that it is unlikely that the were installed during the 1980's (machines up to 150 kW).
123
Spanish government will wish to undermine the market for Early Danish development was based on individual tur-
what is now a considerable Spanish industry by withdrawing bines sited near their owners and, hence, there were
many such turbines scattered across the Danish country- Other New Member States
side. A repowering incentive is intended to tidy up this
arrangement, and to use bigger turbines to replace a larger There is no doubt that there will be activity in several new
number of small ones. It has also allowed the better use of member states, including the Czech Republic, the Slovak
the more energetic sites which, again, were the first to be Republic and several Baltic states, as well as accession
exploited. The Danish repowering programme has been run- countries such as Romania. Wind energy can be a key tool
ning for a couple of years, and has been extremely suc- in the process of cleaning up new member states’ elec-
cessful not least due to good planning procedures. Three tricity production systems which are, at present, heavily
offshore projects have been constructed in 2003. The reliant on nuclear power and fossil fuels.
framework for the future Danish offshore tenders is expect-
ed to be published in the beginning of 2004. Wind energy associations have been established in six new
member states: Estonia2, the Czech Republic, Hungary,
Poland Latvia, Poland3, and Slovakia4. The objective of these asso-
ciations is to establish greater wind capacities in their
Poland was the first eastern European country to make respective countries.
any real progress in the exploitation of its wind resource
(see Table 1.2). The reasons for this pioneering step are
three-fold. First, its immediate neighbour, Germany, has 1.6 Export Market
been the leading market for some years, and some com-
ponents for the German and, indeed, the Danish industry The export market is extensive and includes both countries
are made in Poland. Second, the wind resource is very falling into the conscience market and those constituting
similar to that of its immediate neighbour, with develop- the needs market. The figures in Table 1.7 from BTM
ments along the North Sea expected. And third, the polit- Consult are computed on the basis that an export market
ical and commercial position within Poland is generally is one outside the manufacturer’s base country. The Table
favourable for external investors. It would probably be pre- shows the total number of MW by manufacturer exported
mature to suggest that the level of development could be and installed domestically, as well as the resulting export
comparable to that in northern Germany, but commenta- percentage of their sales. As the report from BTM points
tors expect to see Poland undertaking large develop- out, these figures do not express actual cross border sales
ments, continuing to participate in the industry in terms of as some manufacturers are producing outside their base
manufacture and, perhaps, taking a leading role in the country and within the “export” market, for example Vestas
development of the infrastructure for some offshore wind manufacturing in Italy and GE Wind in Germany (national
farms. The activity associated with the planned offshore base of the now defunct Enron Wind bought by General
wind farm will result in a closer connection with the indus- Electric).
try as a whole. Several leading WT manufacturers are con-
sidering the possibility of establishing Polish factories. A The existing export market for the EU-25 as a whole
law exists requiring utilities to connect wind farms to the includes Norway, which recently installed two 40 MW wind
grid, but there is not yet any well functioning tariff struc- farms. Key non-European markets include the US, Canada,
ture set for the purchase of electricity. India, Japan, China and Australia. Other markets set to
emerge in the future are discussed in volume 5 chapter 2.
124
Table 1.7: Export of Manufacturers Worldwide in 2002 1.7 Industry Segments
Company Country Installed Domestic Export Export
1.7.1 MANUFACTURERS – MEGAWATT SCALE
2002 (MW) 2002 (MW) 2002 (MW) Share (%)
Vestas Denmark 1,605 266 1,338 83.4

Nine of the top 10 turbine manufacturing companies are
NEG Micon Denmark 1,033 166 867 83.9
European. Wind energy is an outstanding European suc-
3
GE Wind US 638 62 576 90.4 cess story, with European companies manufacturing more
Bonus Denmark 509 74 435 85.5 than 90% of the turbines sold worldwide in 2002. In terms
Enercon Germany 1,334 1,103 230 17.3 of electricity generation, in 2003 WTs generated 2.4% of
EU-15 electricity, in Denmark this figure is 20%, while in
VOLUME
Nordex Germany 504 284 220 43.7
Germany it is 6%, and in Spain in 2002 the figure stood at
Lagerwey Netherlands 114 4 111 96.9
4%.
Mitsubishi Japan 30 5 25 84
Gamesa Spain 854 839 15 1.8

Table 1.8 shows the top 10 megawatt scale Turbine sup-
Made Spain 247 234 13 5 pliers in 2002.
Dewind Germany 86 80 7 7.8
Repower Germany 223 221 2 0.7 1.7.2 MANUFACTURERS – SMALL TURBINES

Ecotecnia Spain 120 120 0 0
Smaller turbines may be installed in small wind farm con-
Suzlon India 60 60 0 0
figurations or as individual units. The vast majority of
Fuhrlander Germany 47 47 0 0
small WTs are less than 30 kW in capacity, with rotor
diameters from 1 m up to around 15 m. Small turbines
Source: BTM Consult (2003).
usually satisfy an individual power demand or property.
Table 1.8: Top 10 MW Scale WTG Suppliers 20025
Company Country Accumulated MW Installed MW Share MW Accumulated MW Share of Total MW

2001 2002 Installed 2002 2002 Installed %
Vestas Denmark 4,983 1,605 22.2% 6,588 20.6%
Enercon Germany 3,206 1,334 18.5% 4,540 14.2%
NEG Micon Denmark 4,510 1,033 14.3% 5,543 17.3%
Gamesa Spain 2,125 854 11.8% 2,979 9.3%
GE Wind USA 2,288 638 8.8% 2,925 9.1%
Bonus Denmark 2,306 509 7.0% 2,815 8.8%
Nordex Germany 1,473 504 7.0% 1,978 6.2%
Made Spain 783 247 3.4% 1,030 3.2%
Repower Germany 379 223 3.1% 602 1.9%
Ecotecnia Spain 362 120 1.7% 482 1.5%
Others 3,677 371 5.1% 4,048 12.6%
Total 26,092 7,436 103.00% 33,528 105.0%
125
Figure 1.6: Top 10 MW Scale WTG Suppliers 2002 Nuon Renewable Energy Projects The Netherlands
P&T Technology Germany
Renewable Energy Systems (RES) UK
SIIF Energies France
Windkraft Nord (WKN) Germany
Airtricity is developing wind farms in the Republic of

Ireland, Northern Ireland and Scotland. It is also develop-
ing the largest offshore wind farm in the world, off the
Arklow coast in Ireland.
Elsam’s offshore wind farm at Horns Rev comprises 80

wind turbines located 14-20 kilometres off the coast in
the North Sea. It is the largest wind farm of its kind and
produces enough electricity to supply 150,000 house-
holds, year-round.

At the end of 2002, Energia Hidroelectrica de Navarra
(EHN) had installed a total of 1,380 MW. This represent-
Included in this market sector are turbines that may be ed approximately 30% of Spain’s installed capacity. The
lowered in high winds, making them safe options for elec- group, at the time of writing, has plans for a further 1000
tricity generation in areas prone to storms. They may also MW to be installed in the next few years. The wind energy
be installed without the use of cranes - previously a limit- production of the EHN group in 2003 was 1,376 GW
ing factor as many developing countries lack access to hours, with a production share in Spain of 14%.
such hardware.
Italia Vento Power Corporation (IVPC) has 10 wind farms
The small WT sector can be divided into five segments: in the regions of Foggia and Benevento in Southern Italy,
• Individual use with an installed capacity of approximately 170 MW.

• Isolated communities and industries
• Connected to basic grids P&T Technology has primarily concentrated on securing
• Connected to distribution grids wind farm locations under leasehold agreements. In addi-
• Power source for water pumps tion, a range of wind farms has been constructed: since
2000, this amounts to approximately 210 MW.
1.8 Wind Farm Developers Renewable Energy Systems (RES) has projects in the UK,
Europe, North America, the Caribbean and Asia. At the
The Principal European Wind Farm Developers include: end of 2003 RES had over 790 MW of wind energy capac-
ity built and more than 6,000 MW under development.
Airtricity Ireland
Elsam Denmark SIIF Energies operates the largest wind farm in Portugal
Energia Hidroelectrica de Navarra (EHN) Spain (Pinheiro and Cabril), is selecting potential sites in Europe
Italia Vento Power Corporation (IVPC) Italy (France, Italy, Spain) and Latin America (Mexico and
National Wind Power UK Brazil), and has interests in the US and Scandinavia.
126
2 EMPLOYMENT IN THE WIND TURBINE SECTOR
2.2 Direct Employment in Wind Turbine

Assumptions Manufacturing
1. The following chapters relate only to employment Within the EU, employment in manufacturing is concen-
through manufacture, installation, operation and main- trated in a few countries, with Germany, Denmark and
tenance of wind turbines in EU countries. This excludes Spain accounting for more than 90%. The WT manufactur-
employment not associated with an input to the manu-
facture, installation and maintenance of turbines.
2. The total direct and indirect employment for the EU

ing sector’s share of total manufacturing employment is,
on average, approximately 0.1% for the EU-15, but for
countries with a large WT manufacturing sector, this share
may be much higher. For example, in Denmark in 2002,
3
VOLUME
is calculated based on the installation of turbines in the wind industry’s share of total manufacturing employ-
Europe. Employment in Europe and elsewhere associ- ment was 1.2%, more than cement and steel production.
ated with exports of turbines outside the EU is not
included. Figures for direct employment in Italy and the Netherlands
have not been included in this analysis due to lack of
data, although some WT manufacturing does take place in
2.1 Direct and Indirect Employment both these countries.
Effects
Employment throughout the manufacturing sector has
The production structures of wind turbine (WT) manufac- been increasing considerably in the EU since the beginning
turers vary considerably. Some, for example, manufacture of the 1990s, as exemplified by the Danish experience.
components, including blades, whereas others design and
assemble components purchased from different sub-con- Table 2.1: Direct Employment in WT Manufacturing in Europe
tractors. Because sub-contractors and component suppli- for 2002
ers are a key link in the WT manufacturing process - from
basic raw materials to the finished product - their inclusion Employment within EU Turbine Manufacturing Share
in this analysis provides a more accurate assessment of Total 30,946 100%

the employment effects of the WT manufacturing sector in Austria 720 2.3%
the EU as a whole. This measure is based on national Denmark 6,624 21.4%
account statistics (Eurostat, 2000) and input-output UK 1,150 3.7%
methodology. France 756 2.4%
Germany 10,439 33.7%
Employment in WT manufacturing includes both direct as Portugal 60 0.2%
well as indirect employment. Spain 11,197 36.2%
Source: EWEA (2003b)
127
Figure 2.1: Direct and Indirect Employment in Danish Turbine Table 2.2: Direct Employment in WT Installation in Europe for
Manufacture 2002
Employment within EU Turbine Installation
Austria 213
Denmark 1,500
UK 800
France 340
Germany 5,771
Greece 30
Portugal 100
Spain 4,500
Others 1,395
Total 14,649
Source: Danish Wind Industry Association (2002). Source: EWEA (2003b), own calculations.
The Danish Wind Industry Association estimates that direct employment per MW installed. The variation in employ-
and indirect employment in WT manufacturing in Denmark ment between the countries is shown in Figure 2.2, reflect-
increased from around 2,900 in 1991 to 21,000 in 2002, ing the same differences as those seen in the cost varia-
a relative surge not experienced by any other manufacturing tion for installation included in the volume on costs and
industry. It has brought WT manufacturing to the fore as an prices (Volume 2). As with costs, installation seems to
important sector for the Danish economy, and has con- require the highest employment figures in the UK.
tributed to reducing unemployment.
The UK figures are considerably higher than earlier Danish
In recent years, the growth of employment in WT manu- studies suggest (Krohn, 1998 and Danish Wind Industry
facturing in Germany and Spain has been higher than in Association, 1995 and 1997). In these studies, the global
Denmark with nearly a doubling in the numbers employed employment figures for installation were found to be in the
over the period 2000 to 2002. region of five individuals per MW in 1998. The largest part
of installation activity is construction. For this activity,
2.3 Direct Employment in Wind Turbine direct employment accounts for around two-thirds of the
Installation total employment related to the construction part of instal-
lation. The multiplier shows total employment in the EU
WT installation itself offers significant employment oppor- associated with €1 million of output in the construction
tunities, although there are differences in employment sector, including employment in all the sectors supplying
effect depending on the type of WT, the location and the inputs to the construction sector. Furthermore, the employ-
country of installation. An estimate of the employment ment content will be reduced as a result of cost reductions
impact of WT installation is given in Table 2.2. achieved from 1998 to 2002. The employment factor used
for countries not included in the Figure is three individuals
The employment effects of WT installation in other EU per MW. This figure is within the range of the averages
countries has been calculated, based on the average included in Figure 2.2.
128
Figure 2.2: Direct Employment Associated with WT Table 2.3: Direct Employment in Maintenance in Europe for
Installation in Selected European Countries for 2002 2002
Employment within EU Maintenance
Austria 60
Denmark 300
3
UK 50
France 44
Germany 1,010
Greece 90
VOLUME
Portugal 30
Spain 966
Others 218
Total 2,768
Source: EWEA (2003b), own calculations.
The high employment figure for WT installation in the UK The O&M employment calculation for the “other country”
can be partly attributed to the remote siting of wind parks group is based on a conservative value of 0.1 person per
that require quite extensive road construction and grid MW of installed capacity in each of these countries (see
infrastructure investment. Figure 2.3). The fluctuation in employment per MW capacity
in Figure 2.3 is quite high. This could partly be caused by the
difference in age of installed turbines or could be a function
2.4 Direct Employment in Maintenance of wind turbine size and wind park grouping.
Activities
Figure 2.3: Employment in Maintenance of WTs in Selected
Employment related to operation and maintenance (O&M) European Countries for 2002

will increase considerably as installed capacity increases.
However, present employment related to this activity is
still small compared to that associated with manufactur-
ing and installation.
It seems from Figure 2.3 that maintenance is more

employment-intensive in countries with the least installed
capacity, namely Austria, France and Greece. It is notice-
able that the UK has low maintenance employment fig-
129
ures compared to other countries, whereas its installa-
tion-related employment is the highest in the EU.
2.5 Indirect Employment in Calculations of Indirect Employment in Manufacturing
Manufacturing
All calculations in this section are based on installation
Individual assessments of indirect employment in WT of turbines in Europe and not the actual turnover figures
manufacturing have been made by the national associ- which were not available. Estimated turnover is based on
ations in Denmark and Germany. For Germany, an indi- the investment cost for wind turbines given in Volume 2.
rect employment total of approximately 24,000 people
for 2002 has been estimated and, for Denmark, Calculated direct and indirect employment can be com-
14,500. pared to direct employment figures reported by national
associations. Reported direct employment is aggregated
Alternatively, the estimate of indirect employment content at European level and constitute 60% of the total direct
within different sectors is based on national account sta- and indirect employment totals found in Table 2.5. This is
tistics from Eurostat’s input-ouput tables (2002). partly caused by omitting from the calculations the export
element of turbine manufacturing activity in Europe.
Input Structure for WT Manufacturing
For 2002, the figure for total European manufacturing
Firstly, we have to establish the composition of interme- and installation employment is closer to 12 individuals
diate inputs to the manufacture of WTs. The input struc- per MW installed. The global employment effect is some-
ture varies a great deal both for individual manufacturers what higher than 12 per MW, because of imports of raw
and for EU member states. The following is based on materials, etc. to Europe, which create employment
responses received from the national associations and abroad.
earlier Danish data:
The calculations here might understate employment
Table 2.4: Input Structure in the Danish WT Manufacturing slightly because of the high import quota implicit in the
Sector employment multipliers for Europe.
Input structure Denmark 1995 (%)

Generator 4 2.6 Indirect Employment in WT

Gearbox 12 Installation
Rotor 18
Again, the direct employment reported by the national wind
Tower 18
associations varies a great deal per MW installed capaci-
Brakes 1.5
ty, and has, in a number of cases, been difficult to esti-
Electronic 4
mate. Therefore, a calculation based on the different ele-
Nacelle (remainder) 42.5 ments of installation activities has been carried out in
Total 100 order to provide an alternative measure of employment.
The different elements of installation activities have been
estimated from national association data and other
The contributing sector of the 25 sector level data from sources (Danish Wind Industry Association, 1997).
the EU input-output table (see Appendix J) have been
associated with the production inputs by judgement and The question of divergence in installation cost composition
based on Danish studies (Krohn, 1998; Danish Wind between onshore and offshore wind turbines has not been
Industry Association, 1995 and 1997). Only limited data addressed. If the cost composition is identical for the two
130
for the input structure of WT manufacturing in different EU locations, the different level of installation costs has no
countries are available. effect on employment per € million of installation costs.
Table 2.5: Calculation of Direct and Indirect Employment for WT Manufacturing
Input Structure Contributing Sector Employment Employment Employment

Multiplier 1995 Multiplier 2002 2002
Generator 4% Electrical goods 14.22 10.81 1,836
Gearbox 12% Industrial machinery 13.6 10.33 5,268
Rotor 18% Rubber and plastic products 14.27 10.84 8,292
Tower
Brakes
Electronic
18%
1.5%
4%
Metal products
Industrial machinery
Office and data processing machines
19.84
13.6
10.72
15.08
10.33
8.15
11,528
659
1,384
3
VOLUME
Nacelle (remainder) 42.5% Industrial machinery 13.6 10.33 18,658
Total 100% 47,625
Source: Eurostat (2003).
In respect of the cost components, each one must be related 2.7 Total Direct and Indirect
to the national accounts sector that supplies the service. As Employment (Manufacturing,
construction is the major employment contributor during the Installation and Maintenance in
installation phase, there will only be minor differences in the Europe)
employment effect, even if the composition of cost differs
between the countries. The aggregate employment effect is, The calculations in sections 2.2 - 2.6 provide an estimate
however, also dependent on possible differences in the for employment in the wind industry in Europe, including all
employment content for each € cost in the different countries. activities such as sub-contractors, etc., associated with the
Labour productivity varies among EU member states, espe- installation activity in the EU.
cially for non-traded goods such as construction activities and
services. This aspect is not included in the calculations that The figures relating to maintenance employment have been
use EU level statistics for labour content in construction. adjusted to include indirect employment as it is estimated
that only around 25% of maintenance costs are related to

Calculation of Indirect Employment in Wind Turbine wages. If the manufacturing employment element relating
Installation in Europe to turbine export had been included, the total employment
figures would have been somewhat higher.
1995 EU data for employment as part of the national
accounts has been used with calculated multipliers as Finally, it must be stressed that total employment relating
shown in Appendix J. The multipliers reflect the difference to wind is considerably higher than the 72,000 given for
in indirect and direct employment for the various elements Europe in Table 2.8, not only as a result of production and
of WT installation. For example, construction of the foun- installation outside the European region, but also due to
dations directly and indirectly employed 13.78 individuals the indirect employment effects of imported inputs to
per € million in 2002. European WT manufacturing.
The basic assumption behind the calculation is that the

composition of installation costs as an EU average did not
change from 1995 to 2002. For the employment multipli-
er, an assumption of a 1.5% increase in labour productiv-
ity per year has been made. This implies that the employ-
131
ment multiplier per current cost unit (€) has decreased by
around 4% per year.
Table 2.6 Installation Costs for Wind Turbines 2002
2002 Foundations Infrastructure Electrical Installations, Grid Reinforcement Other Installation Total Installation Cost
Roads, etc. etc., Connections Costs Excluding Turbine
France 34% 14% 42% 0% 9% 100%
Denmark (1995) 16% 5% 55% 16% 9% 100%
Spain 23% 54% 23% 100%
Portugal 22% 22% 22% 33% 0% 100%
Table 2.7 Calculation of Direct and Indirect Employment for Turbine Installation in Europe for 2002
Average Share of Costs Contributing Sector Employment Employment Employment

2002 (Simple Average) Multiplier 1995 Multiplier 2002 2002
Foundations 24% Construction 18.14 13.78 4,706
Infrastructure Roads etc. 14% Construction 18.14 13.78 2,665
Electrical Installations etc., Construction/

Connections 40% Industrial machinery 15.87 12.06 6,790
Grid Reinforcement 16% Construction 18.14 13.78 3,210
Other Installation Costs 6% Other Market Services 57.61 43.78 3,780
Total 100% 21,150
Table 2.8: Total Direct and Indirect Employment Related to WT Manufacture in Europe for 1998-2002
Employment in WT Employment in Employment in Total Employment

Manufacturing (for Home Market) WT Installation WT Maintenance
1998 16725 7400 950 25,075

2002 47625 21150 3500 72,275
Growth 1998-2002 185% 185% 268% 188%
Source: Own calculations based on installation of WTs in Europe and Eurostat (2000).
132
3 EMPLOYMENT PREDICTION AND METHODOLOGY
The history of employment relating to wind energy in the One way of addressing these more complex economic
EU is very positive. Employment has been growing rapidly linkages is to use macroeconomic general equilibrium
in recent years, and the sector has thus contributed to models. However, these models very seldom include
reducing unemployment in the region. details of WT manufacturing.
A broader understanding of employment in the wind ener- Consequently, this work has focused on employment in
gy sector is, however, not straightforward as there is great
uncertainty about what this employment covers. Here, we
have chosen to use a notion of direct and indirect employ-
ment and to separately examine the manufacture of tur-
activities directly producing WTs and those supporting and
supplying to the wind sector. This is, therefore, not an
attempt to address the employment created by the WT
sector, or to give an overall figure for the net employment
3
VOLUME
bines, their installation and employment arising from effect. Instead, this chapter tries to estimate the employ-
O&M. The use of different terms and the applied method- ment directly or indirectly associated with the WT manu-
ology is described in the section below. facturing sector.
The objective has been to examine the magnitude of

employment in the sector directly producing WTs and 3.1 Direct and Indirect Employment
employment associated with the production of inputs to
turbine manufacturing. This is not a calculation of the Direct employment relates to employment within WT man-
employment created by WT manufacturing, as parts of the ufacturing companies and sub-contractors whose main
workforce in the WT sector today would certainly have activity is supplying WT components etc., for example
been employed in other activities had the wind sector not blade manufacturers. However, companies producing
existed. A similar approach would be to evaluate the net intermediates or components for the WT industry are
employment effect by deducting the employment associ- deemed as providing indirect employment if this is only a
ated with alternative electricity producing technologies minor part of their activities.
from the employment associated with wind-based elec-
tricity production. This approach would result in a net There is an important distinction between national and
employment lower than the gross employment associated global employment patterns in WT manufacturing. Direct

with wind-based electricity production. The net effect is, and indirect employment at national level does not
however, assumed to be of a considerable size as the include employment associated with imports. For small,
employment content is somewhat higher in WT manufac- open economies, this means that there will be a large dif-
turing and installation. ference between national and global employment content.
For a larger country, or for the entire EU, the difference
Some might argue that the input-output approach does between national and global employment content will
not address the wider employment effects of WTs. The lessen as trade flows between countries are consolidat-
argument is based on the fact that the income generated ed, with the effect being included in indirect employment
in the sector via wages paid and rents extracted will con- at EU level. If direct and indirect employment effects for
tribute to demand for other goods that, again, will gener- EU countries are amalgamated, this figure will be less
ate employment. However, such arguments should lead to than the direct and indirect employment effects calculat-
a consideration of how WT development is being financed. ed at the EU level.
The funds used for WT investment might have been
invested in other electricity producing equipment, or even
in totally different sectors. These investments would have 3.2 Input-Output Methodology
created jobs, as in the wind sector, and the net effect may
133
have been higher or lower depending on the labour inten- The calculation of employment effects is based on input-
sity of the activities in which the investment takes place. output methodology used by economists. The basic idea
is to include the effects from suppliers of inputs (raw sectors of the economy to produce an additional €1 mil-
materials, etc.) to obtain a better measure of the total lion of output for each of the sectors. These production
effect of the activity in question - in this case, WT manu- changes are then multiplied by the direct employment in
facturing. Direct employment in manufacturing activities each of the 20 sectors per € million of output.
having WT-related products as their main output is seen
as the first link in a chain of employment effects. To reach 2002 levels of employment multipliers, the cal-
Secondary links are employment associated with the pro- culated figures for 1995 must be projected forward. This
duction of components and raw materials used in the pro- can be done individually for the multipliers of all sectors,
duction of turbines. These secondary effects continue for example, by using trends in multipliers, or by assum-
with employment used for extracting raw materials need- ing “productivity” increases. The latter approach was
ed for products that are later used as WT components. chosen here.
The secondary effects diminish the further back in the pro-
duction chain we go. National Versus EU Employment
The employment used is the number of employees per One important aspect is the distinction between national
output unit measured in € million. and global employment effects. If considered nationally,
the employment effect would include the direct employ-
Example: ment effect, but all the indirect effects would exclude the
• 1 MW installed capacity at a price of €1 million part of inputs that are being imported. Thus, the smaller
• €0.75 million of wind turbine output in 2002 the country, the larger the imported share of production
(the rest is installation costs, etc.): inputs. In this way, the indirect employment effect is less
for small, open economies. For larger countries or for a
Direct employment + indirect employment: group of countries such as the EU, the indirect employ-
€0.75 million x 7.94 employees per € million ment effect is larger as a much higher fraction of the
inputs to the sector will originate within the EU. Then, if all
+ level 1 input: (input from sector “industrial machinery” the countries added the national (direct and indirect)
x industrial machinery employment coefficient per employment effects, the sum of these would still be less
€ million output) + input from sector 2 x employment than the direct and indirect effects for the entire EU, as
coefficient + input sector 3… this also includes employment effect of the intra regional
trade flows.
+ level 2 inputs: (input from sector 2 to sector “electric
machinery” x employment coefficient in sector 2 per Using Input-Output for Projections of Employment
€ million output in sector 2…
The input-output methodology used for projections of
The calculation could continue indefinitely but, instead, employment allows the use of different assumptions on
we use the input-output methodology for calculating the developments in productivity for different sectors, as well
inverse matrix and multiplying by the employment coeffi- as possible shifts in the composition of inputs from other
cients. The employment coefficient is the direct employ- sectors in the manufacturing of WTs.
ment per output in the sector.
The productivity and composition changes are closely
We have in this case calculated the direct and indirect linked to assumptions that can be made about the over-
employment effects from Eurostat’s national account sta- all cost reduction of WT manufacture. There must be
tistics for 1995. By using input-output methodology, we consistency between the assumptions made on cost
134
can calculate the necessary production increases in 20 reductions and on reduction in employment per MW.
3.3 Projection/Prediction Parameters Labour Productivity
In order to make consistent projections of employment, Apart from the cost reductions per MW, there will be
there are a number of parameters that have to be increases in labour productivity. In the long term, a 2%
addressed. These include activity parameters as well as annual increase in labour productivity (employment per
parameters of technological progress. output unit in fixed prices) is a reasonable assumption for
Turnover or Indicators for Output in MW
The total turnover for the WT manufacturing sector in the

the European economies. This also reduces the future
employment effect per MW of installed wind capacity.
Input Composition in WT Manufacturing

3
VOLUME
EU can be projected using a variety of methods. One
method is to use the installation forecast for MW glob- Finally, the composition of inputs in WT manufacturing can
ally and then add the European market share of manu- be addressed. In many cases, there will not be the nec-
facturing (see below). Then, there needs to be a conver- essary amount of information to separately project this
sion from MW installation to turnover in € where 1 MW parameter. Thus, only overall cost reductions will be pro-
installed capacity might equal €1 million at today’s jected and equally distributed on all inputs. This implies a
prices, but the cost reductions should reduce this figure constant technical coefficient in the input-output system.
by at least 2% a year (according to experience curve and
cumulated installation). These figures are in fixed price
terms, which means that the cost of WTs decrease by 3.4 Sensitivity – Main Parameters
2% annually, compared to the price of other goods and
services. For all projections, there is a need to identify the most
important parameters with respect to their possible varia-
However, this also includes an assumption of an tions and their impact on total employment in the sector
unchanged mix of WT categories between those with low in Europe.
investment cost relative to production and those with
higher investment cost per MW, but higher production per Wind Turbine Market Growth and Regional Distribution

installed MW. In some cases this can be observed as the
larger machines have higher investment cost per MW, but Future growth of the WT market is the main assumption
lower investment cost per produced kWh. for employment as it is clearly the driver for production.
However, it is associated with some uncertainty since the
Share of Production Taking Place in Europe market size in 15 years’ time has a possible variation of
a factor of 10.
As the European market dominates both in terms of annu-
al installed capacity, and in terms of manufacturing activi- Productivity/Cost Reductions
ty, the installation has been equal to European production
and then some additional production for exports could be The sensitivity of employment projections to assumptions
added. The share of worldwide production taking place in of cost reductions is less obvious than the sensitivity to
Europe will in the future be reduced and this development market projection. In the longer term, assumptions about
should be addressed by making explicit assumptions. technological progress and cost reductions are, however,
Even though market shares are still high for the European of a considerable accumulated size, and have a large
producers, a larger fraction of manufacturing will take impact on employment forecasts. With experience curves
place locally at the markets where they are to be installed. suggesting a 15% cost reduction per electricity output and
135
a 10% cost reduction for turbine costs for every doubling
of installed capacity, these cost reductions must partly
attribute to reductions in the use of labour input in the
production of turbines. Moreover, the installation of tur-
bines will become less labour intense due to productivity
growth within all sectors of the economy.
The link between cost reductions and accumulated instal-

lation makes the cost reductions sensitive to wind market
development, in addition to the uncertainty that relates to
the experience curve itself. Cost reductions that reduce
labour input in manufacturing are for installed MW and not
relative to electricity production. Therefore, it is the 10%
cost reduction mentioned above that is the relevant figure
here. If market growth corresponds to a doubling of
installed capacity in five years, then cost reductions per
installed MW will decline by approximately 2% per year.
This figure might just as well be 1% or 3% depending on
market expansion and the “real” experience curve.
European Production Share
As wind energy develops and becomes more widespread,

the European share of total installation will decline. Even
though European producers will maintain a high market
share, a larger part of their manufacturing activity will take
place outside Europe. These foreseeable changes are
very difficult to project, depending as they do on both mar-

ket forces/demand and strategic developments/reorgani-
sation of the WT manufacturing sector. The most likely
development is that a larger proportion of European man-
ufacturers’ activities will take place outside Europe. The
impact will be to reduce the growth of European employ-
ment within the WT manufacturing sector. European mar-
ket share and location of production facilities is thus an
important parameter for sensitivity analyses.
If long term projections are made, it is unclear how much

European companies’ manufacturing will take place in the
country of installation.
136
4 SCENARIOS FOR EMPLOYMENT IN THE WIND TURBINE SECTOR
Based on scenarios for the future development of wind in of which 75% will consist of WTs produced in Europe
Europe and globally, it is possible to identify correspon- (installation will not generate employment in Europe apart
ding employment scenarios. from that included below). WT manufacturing in Europe
will be €19,692 million.
4.1 Projection of Employment based on The employment multiplier for 2002 of 11.21 employees
Wind Energy Installation in Europe
and Globally
Based on assumptions for the parameters, etc. described

per € million of activity has to be reduced by the general
productivity increase of 2% annually resulting in a multi-
plier for 2020 of 7.79 person per € million. 3
VOLUME
in chapter 3 above, it is possible to develop a scenario for The resulting employment in WT manufacturing for the
European employment in the wind sector. Here, a very year 2020 will therefore be 7.79 x 19,692 = 153,400
simple scenario with the majority of composition parame- employees.
ters remaining unchanged will be presented.
Installation Employment
Some basic assumptions are identical to those in Wind
Force 12 (EWEA, 2003c). The scenario projects European In 2020, 15,000 MW will be installed in Europe.
and global installation activity in 2020 as follows: Installation employment from 2002 has to be adjusted for
• Annual global installation will increase to approximately the general increase in labour productivity and the
150,000 MW in 2020, of which 15,000 will be in Europe. reduced employment input that has contributed to lower-
• The European share of global WT manufacturing is ing the total costs of installing WTs. This means that the
assumed to decline to 25% by 2020. cost reduction is assumed to be at a similar level for both
• Turbine manufacturing input is assumed to have con- manufacturing and installation.
stant composition in Europe.
• Installation activity is assumed to have constant compo- Following on from this, the employment multiplier for
sition (no change from increased offshore expansion). installation in 2002 has to be adjusted. We assume that
• Cost reduction is assumed at a rate of 2% annually. the composition between the components of installation

activity is unchanged from Table 2.7, and that the employ-
An annual 2% growth in labour productivity is assumed for ment multipliers for the contributing sectors all follow the
both manufacture and installation. same trend with a 2% annual reduction.
In this scenario, where a great deal of the employment 15,000 MW installed in Europe at a total cost of €1 million
increase will come from expansion of markets outside per MW in 2002 reduced by 1/1.02^18 equals a total
Europe, it will be mainly increases in manufacturing employ- investment cost in 2020 of €10,502 million.
ment that are responsible for overall employment growth.
Installation cost constitutes an unchanged share of 25%
Manufacturing Employment of this amount and the employment multiplier for installa-
tion is 14.89 in 2002 reduced with the annual productivi-
Global installation of 150,000 MW in 2020 and a ty increase of 2% resulting in a multiplier for 2020 of
European share of 25%, including those for the European 10.42 per € million of installation activity.
market would require:
Direct and indirect employment in European
€1 million per MW in 2002 reduced by 1/1.02^18. This WT installation for 2020 would therefore be 27,400.
137
equals a total investment cost in 2020 of €26,256 million,
Maintenance However, it must be stressed that employment growth will
be even higher in the regions outside Europe where instal-
By 2020, an accumulated 230,000 MW will have been lation growth is also highest. The employment effect in
installed in Europe. With the conservative value of 0.1 these countries will probably be even higher than in
employees per MW in 2002, this would mean a consider- Europe due to lower productivity and wages.
able increase in employment for maintenance. The
employment content also has to be adjusted for the gen- European Scenario for Enployment up to 2010
eral productivity increase of 2% annually. In maintenance,
only the general productivity increase is assumed and To illustrate the employment effects of wind turbine instal-
there are no additional cost reductions. lation in Europe another scenario is included. Calculations
are performed according to the above scenario (see Table
With these assumptions, employment in O&M in 4.1). This scenario includes only employment effects in
Europe in 2020 would be 16,100. Europe from installation activity in Europe. The European
WTs are thus assumed to be entirely manufactured in
Employment Scenario Results Europe. Furthermore the employment effect from export of
European produced WTs is excluded.
For the scenario described above, employment in 2020
has increased considerably from that in 2002. Table 4.1 illustrates that even with moderate growth in WT
installation, employment will be stable around the present
Total employment related to manufacturing, installation and level if export activity is not included. The reduction in
O&M in Europe will be 196,900 in 2020 for this scenario, employment associated with manufacture of WTs is bal-
based on a considerable expansion of wind energy. anced by the increase in employment associated with
maintaining the already installed capacity.
This is more than a doubling of today’s employment in the
European sector, but it is not the full employment story.
Table 4.1: Direct and Indirect Employment Associated with European WT Installation
2003 2004 2005 2006 2007 2008 2009 2010

Europe Cumulative Installed MW 29,116 35,216 41,516 47,966 54,566 61,316 68,216 75,216
Europe Annual Installation 5,900 6,100 6,300 6,450 6,600 6,750 6,900 7,000
Employment Manufacturing 45,300 45,017 44,687 43,975 43,250 42,515 41,772 40,732
Employment Installation 20,520 20,799 21,060 21,139 21,206 21,263 21,309 21,194
Employment Maintenance 2,854 3,385 3,912 4,431 4,942 5,445 5,939 6,420
Total 68,674 69,201 69,659 69,545 69,398 69,223 69,020 68,346
138
5 DEVELOPMENT & INNOVATION
Development and innovation have contributed to the fast Ball Bearings

progress of the WT manufacturing sector. One of the rea-
sons for this is the ability of the sector to adapt technol- As mentioned above, components such as ball bearings
ogy from other sectors. used for WTs were, until recently, mainly standard
products. But with the development of 2-3 MW turbines it
Technology transfer from the WT industry to other sectors became necessary to produce special large ball bearings,
has been more modest. One example, however, from
Denmark, is the expansion of small-scale machine manu-
facturers to larger companies, based on their activities in
the wind industry. These small companies, often charac-
designed to the specific requirements of WTs.
Another niche area in this industry is the bearings that

support pitch-regulated blades. Because of the very small
3
VOLUME
terised as “smiths”, have expanded their expertise with rotation angle of these WT components compared to their
the technology developed for the wind sector. As these use in other kinds of machinery, the loads on these bear-
companies have often been located outside traditional ings are very high. To limit wear, specialised bearings are
business centres, this development has been seen as being developed that perform well within small rotation
very positive trend to broader industrial development. angles.
Along with the manufacturing of WTs, a range of spe- Blades

cialised service suppliers for transport, installation, main-
tenance and insurance has developed. In turbine up-scaling, the weight of components such as
blades, nacelles and towers is of the utmost importance.
One promising technology cross-over area is in aerody- To keep loads down it is necessary to keep weight down.
namics where the use of new materials and the cost This is especially important for the blades; the longer the
reductions associated with them has transferred from the blades, the more the need for lightweight materials.
aeronautic industry to WT blade manufacturers. Initially, blades were manufactured using glass fibre mate-
rials with weight reduced mainly through design improve-
An EU-funded research project (Neij et al., 2003) has ments. But to manufacture blades longer than 40-50 m
analysed the possibility of using experience curve fore- for 2-3 MW machines requires the use of reinforced com-

casts from the wind energy sector to predict possible posites. One of the most common new blade materials is
developments within other renewable energy sectors. This glass fibre reinforced with carbon fibre, but hybrid ver-
is based on the idea that some of the technological inno- sions using glass fibres and wood are also being used.
vations found for wind energy technology can be applied Finally, a number of new technologies are being intro-
to other fields. duced to blade production.
The turbines manufactured from the mid 1980s until the Gearboxes
late 1990s were mainly constructed using standard com-
ponents, the only major exception being the blades, which The gearboxes used in turbines were also, until recently,
were designed and constructed for specific turbine use. standard components. But the large numbers of turbines
But in the late 1990s the turbines had grown so large in produced nowadays, and the need for lighter materials to
size and were being manufactured in such large numbers, reduce the weight of the nacelle and thus the loads on
that special components started being designed and man- larger turbines have driven production of lighter and more
ufactured for turbine use only (see below). compact gearboxes designed specifically for WTs.
139
Installations of Offshore Turbines
Installation of offshore turbines presents a fresh chal-

lenge and demands a new approach. Specially designed
and constructed vessels have been developed that can
carry two or more turbines from the nearest harbour to the
offshore site and erect the turbine towers, nacelles and
rotors. These vessels are continuously being improved to
carry more turbine components, thus reducing the instal-
lation time.
These examples are all closely related to the WT sector

itself; technology cross-over to other sectors should not
be neglected, and deserves more research activity.
Endnotes
1 A form of financial support mechanisms instigated by some national governments to
assist in the development of renewable energy technologies.
2 See www.tuuleenergia.ee (for further information).
3 See www.visventi.org.pl (for further information).

140 4 See www.save.apis.sk (for further information).
5 It should be noted that BTM Consult and EWEA figures of installed capacity vary
slightly due to different methodologies in their compilation.
VOLUME 4
ENVIRONMENT
HOOFDTITEL
Acknowledgments
This volume was compiled by Professor Olav Hohmeyer,

and Scientific Assistants Florian Wetzig and David Mora of
the Chair of Energy and Resource Economics of the
University of Flensburg, Germany. Our thanks also to the
national wind associations around Europe for their contri-
butions of data, and to the other project partners for their
inputs.
WIND ENERGY - THE FACTS - ENVIRONMENT
142
1 INTRODUCTION TO VOLUME 4 - ENVIRONMENT
Sustainable development is an issue of prime importance this volume, showing the main elements affecting public
both now and in the future. As defined by the Brundtland acceptance along with the results of some recent surveys
Commission in 1987, sustainable development is “devel- from a selected number of EU countries.
opment that meets the needs of the present without com-
promising the ability of future generations to meet their own In the first part of this volume, the concept of the external
needs” (World Commission for Environment and cost of energy is introduced. As environmental and health
Development, 1987). costs caused by energy conversion processes are not taken
into account in the calculations of the producer or con-
Environmental pollution and emissions of CO2 caused by sumer of energy, economists call these costs “externali-
the use of fossil fuels constitute a significant threat to ties”. Analysis of these externalities enables the environ-
sustainable development. A major contributor to these mental and health benefits of wind energy compared to fos-
emissions is electricity generation based on fossil fuels. sil fuels to be expressed in economic terms.
4
The Intergovernmental Panel on Climate Change (IPCC)
predicted in its last report (IPCC, 2001) that human- Subsequently, the benefits of wind energy are discussed. In
induced greenhouse gas (GHG) emissions will lead to a contrast to fossil fuel fired power plants, wind energy con-
substantial increase in GHG concentrations in the atmos- verters cause virtually no operational emissions. There may
VOLUME
phere causing increased radiative forcing, with CO2 con- be minor losses of lubricants from the turbine gearbox but
tributing about 50% to this anthropogenic greenhouse these do not normally find their way into the environment.
effect. Without drastic emission reductions of CO2 and Being a clean energy source is the main advantage of wind
other GHGs a significant change in the world’s climate is energy when compared to conventional electricity genera-
inevitable unless energy systems and sources are tion. Indirect emissions, which result from manufacturing,
changed as soon as possible. In addition to the problem installation, maintenance and removal, do play a very small
of climate change, emissions of SO2, NOx and other pol- part in this equation. Nevertheless, these have been taken
lutants from energy conversion processes in conventional into account in our analysis.
electricity generation cause substantial regional damage
to human health and the environment. By means of external cost analysis, it is possible to quan-
tify the environmental and health costs of the different
As most renewable energy sources, such as wind power, electricity generation technologies. To compare the exter-
emit neither GHGs nor other pollutants such as SO2 or NOx, nal costs of wind energy and of the substituted conven-

they will be the basis of any long-term sustainable energy tional electricity generation, we need to analyse and cal-
supply system (Fischedick et al., 2000). The large-scale use culate them. The net avoided external costs of wind power
of renewable energy sources is essential if the necessary are the external monetary benefits of wind energy. Only if
reductions in CO2 and other emissions from electricity gen- we combine these with a comparison of the internal costs
eration are to be met and if sustainable development is to of wind energy and conventional electricity generation sub-
be achieved. stituted do we get a fair picture of the competitive situa-
tion of wind energy.
The following chapters provide a summary of our current
understanding of the direct and indirect environmental In chapter 1, a review of the external cost concept is given.
impacts associated with wind energy, as well as its eco- In chapter 2.1 a short description of the background for the
nomic (external) costs and those associated with avoiding calculations of avoided emissions and avoided external
the environmental and health impacts of conventional elec- costs from the use of wind energy in the EU and in new
tricity generation by substitution with wind energy. Public member states is presented. In chapter 2.2, a short
acceptance of wind energy is crucial for its successful intro- overview of electricity generation structure in each country,
143
duction. Thus, a public acceptance analysis is included in as well as a very brief description of the national environ-
mental policy frameworks is given. In this chapter the
total and specific emissions of CO2, NOx and SO2 are
given for each country.
Calculations of external costs of standard air pollutants

are performed by the EcoSense model, which has been
developed as part of a major European Commission
research effort on the analysis of external energy costs.
This model is briefly introduced in chapter 2.3, but a
short description of the input data and modelling
assumptions used are given here. Chapter 2.4 reports
on the emissions and external costs which can be avoid-
ed by extending the use of wind energy in the EU and in
the new member countries (Turkey, Romania and
Bulgaria are also included). These are reported as total
as well as specific values.
To facilitate a comparison of future and present calcula-

tions of emission and external cost reductions due to the
use of wind energy, a standard methodology for calculat-
ing emission reductions has been designed. This is report-
ed in chapter 3.
Based on the future diffusion of wind energy on the one

hand and on improvements in conventional electricity-
generating technologies on the other, mid- and long-term
emission reductions are forecast in chapter 4.
Chapters 5 and 6 report the public debate on wind ener-

gy, as far as this has been subject to scientific research
and as far as the results of this research are available.
The debate considers such issues as visual intrusion,
noise, and interference with birds, and their influence on

public acceptance.
144
VOLUME 4 - ENVIRONMENT: TABLE OF CONTENTS
CHAPTER 1 EXTERNALITIES 147

1.1 Introduction to Externalities 147
1.2 Definition and Classification 147
1.3 Importance of Externalities 147
1.4 Externalities and Electricity Production 149
1.5 Impacts of Wind Energy and Other Technologies. 152
1.6 Externalities of Wind Energy 154
1.7 Benefits of Wind Energy 155
1.8 Present State of Knowledge 156
CHAPTER 2 ENVIRONMENTAL BENEFITS OF WIND ENERGY

2.1 Background
158
158 4
VOLUME
2.2 Electricity Generation and Emissions in EU-25 and other European Countries 159
2.2.1 Electricity Generation Sector and Environmental Policy Framework 160
2.2.2 Emission Data 163
2.3 The Calculation of External Costs with the EcoSense Model 165
2.3.1 Software Description 165
2.3.2 Input Data to the Model 166
2.4 Benefits of Wind Energy - Results 167

2.4.1 Avoidable Emissions by the Use of Wind Energy 167
2.4.2 Avoidable External Costs by the Use of Wind Energy 168
CHAPTER 3 STANDARD METHODOLOGY FOR CALCULATION OF EMISSION REDUCTIONS 172

CHAPTER 4 ANALYSIS OF EMISSION REDUCTIONS 173
4.1 Avoidable Specific Emissions Through Wind Energy 173
4.2 Avoidable Total Emissions Through Wind Energy 174
4.3 Avoidable External Costs Through Wind Energy 177
CHAPTER 5 PUBLIC ACCEPTANCE ANALYSIS 179

5.1 Environmental Impacts of Wind Energy 179
5.1.1 Visual Impact 179
5.1.2 Noise 180
5.1.3 Land Use 182
5.1.4 Impact on Birds 182
5.1.5 Impacts of Construction on Terrestrial Ecosystems 184
5.1.6 Electromagnetic Interference (EMI) 184 145
5.1.7 Flickering 185
5.1.8 Consumption of Energy (Energy Balance) 185
5.2 Environmental Impacts of Offshore Wind Energy 186

5.3 Factors Affecting Public Acceptance of Wind Energy 188
CHAPTER 6 PUBLIC ACCEPTANCE IN THE EU 190

6.1 Attitudes of EU Citizens to Energy and Energy Technology Issues 190
6.2 Public Acceptance in Spain 192
6.3 Public Acceptance in the United Kingdom 194
6.4 Public Acceptance in Denmark 197
6.5 Public Acceptance in Germany 197
6.6 Public Acceptance in Sweden 198
6.7 Public Acceptance in Austria 198
6.8 Public Acceptance in Belgium 199
6.9 Conclusions 200
146
1 EXTERNALITIES
1.1 Introduction to Externalities are not included in the market pricing calculations and it
can be concluded that private calculations of benefits or
The economics of wind energy show that the capital costs, costs may differ substantially from society’s valuation if
O&M costs, taxes, insurance and other costs, along with substantial external costs occur.
the expected profit, comprise the price of a kWh of elec-
tricity. Depending on the market situation and, perhaps, Externalities can be classified according to their benefits
additional promotional measures, wind energy may or may or costs in two main categories: non-environmental and
not be competitive. It is generally appreciated that although environmental externalities. Table 1.1 lists examples of
wind energy and other renewable energy sources have envi- these externalities of energy conversion (European
ronmental benefits compared to conventional electricity Commission 1994; Centre for Energy, Policy and
generation, these benefits may not be fully reflected in elec- Technology, 2001):
tricity market prices. The question therefore is: “Do market
4
prices for electricity give an appropriate representation of Table 1.1: Classification of Externalities
the full costs to society of producing electricity?”
Environmental and Human Health Non–Environmental
The externalities of energy generation deal with these
VOLUME
• Human health (accidents, disease) • Subsidies
questions in order to estimate the hidden benefits/dam- • Occupational health (accidents, noise, • Research and
ages of electricity production not accounted for in the physical stress) development costs
• Amenity impacts (noise, visual impacts, odor) • Employment
existing pricing system. The costs are “external” because
• Security and reliability of supply • Effects on GDP
they are paid for by third parties and by future genera-
• Ecological impacts (acidification, eutrophication,
tions. In order to establish a fair comparison of the differ- soil quality)
ent electricity production activities, all costs to society, • Climate change (temperature rise,
both internal and external, need to be taken into account. sea level rise, precipitation changes, storms)
The following sections explain the basic concepts and

describe present knowledge about the external costs of The environmental and human health externalities can
electricity generation. Chapter 2 will report on specific additionally be classified as local, regional, or global, with
external costs, which can be avoided in the EU by the use the latter. referring to climate change caused by emis-
of wind energy. sions of CO2 or destruction of the ozone layer by emis-

sions of CFCs or SF6. Non-environmental externalities
1.2 Definition and Classification refer to hidden costs, such as those borne by tax-payers
in the form of subsidies, research and development costs
Looking at the foundations of externalities, the different or benefits like employment opportunities, although for
definitions and interpretations are based upon the prin- the latter it is debatable whether this constitutes an exter-
ciples of welfare economics, which state that economic nal benefit in the welfare economics sense.
activities by any party or individual making use of scarce
resources cannot be beneficial if they adversely affect 1.3 Importance of Externalities
the well-being of a third party or individual (Energy
Information Administration, 1995). As markets neither include external effects nor their
costs, it is important to identify external effects and to
From this, a generic definition of externalities is “bene- monetise the external costs of different energy systems if
fits and costs which arise when the social or economic these are of a similar order of magnitude as the internal
activities of one group of people have an impact on anoth- costs of energy, and if these external costs vary substan-
147
er, and when the first group fails to fully account for their tially between competing energy systems, like conven-
impacts” (European Commission, 1994). Externalities tional electricity generation and wind energy.
As markets do not internalise external costs, internalisa- Figure 1.1: An Illustrative Example of the Social Cost of Energy
tion has to be achieved by adequate policy measures like
taxes or adjusted electricity rates. Before such measures
can be taken, policy-makers need to be informed about
the existence and the extent of external costs of different
energy systems.
Analysing external costs is not an easy task. Science (to

understand the nature of the impacts) and economics (to
value the impacts) must work together to create analytical
approaches and methodologies, producing results upon
which policy-makers can base their decisions on appropri-
ate measures and policies.
As much of the costing of non-market goods includes

valuation procedures, for example by putting a value on a
person becoming ill as a result of a nuclear accident or
the cost of visual intrusion caused by a wind turbine
(WT), or the cost of future damage caused by a tonne of The question arises whether the internalisation of exter-
CO2, the externalities may pose uncertainties; include nalities in the pricing mechanism could impact on the
assumptions, risks and moral dilemmas. This sometimes competitive situation of different electricity-generating
makes it difficult to fully implement externalities by poli- technologies, fuels or energy sources. As Figure 1.1 illus-
cy measures. Nevertheless, they offer a base for politi- trates, a substantial difference in the external costs of
cians to improve the allocation processes of the energy two competing electricity generating technologies may
markets. Koomey and Krause (1997) in their introduction result in a situation where the least-cost technology
to environmental externality costs state that: “… to not (where only internal costs are considered) may turn out to
incorporate externalities in prices is to implicitly assign a be the highest-cost solution to society, if all costs (inter-
value of zero, a number that is demonstrably wrong”. nal and external) are taken into account.
148
1.4 Externalities and Electricity Production Figure 1.2: Generic Fuel Cycle
For the particular case of electricity production, the use of

energy sources may “cause damage to a wide range of Production of construction materials
receptors, including human health, natural ecosystems and

the built environment, and they are referred to as external Transport of construction materials
cost of energy” (European Commission, 1994).
Exploration of fuel
The externalities in the energy sector started to be quanti-
fied by pioneer studies in the late 1980s and beginning of
the 1990s (Hohmeyer, 1988, Friedrich et al., 1989, Extraction of fuel
Ottinger et al., 1990), which started the interest and gave a
4
first insight into the importance of externalities for energy
Transport of fuel
policy as a decision-making tool. The most outstanding pro-
ject on determining the external cost of energy is the
ExternE project, which developed a consistent methodology Transport of personnel
VOLUME
to assess the externalities of power generation in the EU.
For that reason, a brief introduction of its methodology and Power plant operation
an analysis of its results is provided in this chapter.
An important aspect in any analysis of the environmental Treatment of flu gases
externalities of electricity production is defining the activities

that can have an impact. In that sense, the impacts of power Generation of wastes and byproducts
production are not exclusively generated during the operation
of the power plant, but also in the entire chain of activities
Further treatment of waste removal of plant
needed for electricity production and distribution, such as
at the end of its service lifetime
fuel extraction, processing and transformation, construction
and installation of the equipment, as well as waste disposal.
These stages, which constitute the chain of electricity pro- Restoration of site after closure

duction and distribution, are known as the fuel cycle. Every
technology (wind, hydro, coal, gas, etc) has its own very dis-
tinct fuel cycle. A generic fuel cycle can be seen in Figure 1.2.
149
The impacts from any of the stages in the fuel cycle ated with the power technology. Subsequently, the fuel
depend on the particular location of an activity. Impacts chain burdens are identified. Burdens refer to anything
may vary greatly as a function of the sensitivity of the that is, or could be, capable of causing an impact of what-
surrounding ecosystem, the population density, and eco- ever type. After having identified the burdens, an identifi-
nomic and social aspects. In the case of renewable fuel cation of the potential impacts is achieved independent of
cycles like wind, the major impacts of the fuel cycle arise their number, type or size. Every impact is then reported.
from the activities required to produce and install a wind This process just described for the fuel cycle is known as
turbine and ancillary systems, while only minor externali- the Accounting Framework. For the final analysis, the
ties arise from wind turbine operation. most important impacts are selected and only their
effects are calculated.
The ExternE methodology is a bottom-up approach, which
first characterises the stages of the fuel cycle of the sys- Afterwards, the Impact Pathway approach developed by
tem in question (e.g. coal), defining the activities associ- ExternE proceeds to establish the effects and spatial dis-
tribution of the burdens to see their final impact on health
Figure 1.3: Impact Pathway Approach and the environment. Then, the economic valuation
assigns the respective costs of the damages induced by
a given activity.
The most important results of this study are found in its

final phase in which the ExternE methodology was imple-
mented in the EU in 1998 to take into account site-spe-
cific conditions, technologies, preferences, problems and
policy issues. The aim was to create an EU-wide data set
to assess the external cost. The results are shown in
Table 1.2.
150
Source: European Commission (1994).

Table 1.2: External Cost Figures for Electricity Production in the EU for Existing Technologies (c€/kWh*)
Country Coal&Lignite Peat Oil Gas Nuclear Biomass Hydro PV Wind
AT 1-3 2-3 0.1
BE 4-15 1-2 0.5
DE 3-6 5-8 1-2 0.2 3 0.6 0.05
DK 4-7 2-3 1 0.1
ES 5-8 1-2 3-5** 0.2
FI 2-4 2-5 1
FR 7-10 8-11 2-4 0.3 1 1
GR 5-8 3-5 1 0-0.8 1 0.25
IE
IT
NL
6-8
3-4
3-4
3-6 2-3
1-2 0.7 0.5

0.3
4
VOLUME
NO 1-2 0.2 0.2 0-0.25
PT 4-7 1-2 1-2 0.03
SE 2-4 0.3 0-0.7
UK 4-7 3-5 1-2 0.25 1 0.15
* Subtotal of quantifiable externalities (such as global warming, public health, occupational health, material damage)
** biomass co-fired with lignites
Source: European Commission (1999), data updated in 2003.
151
Table 1.2 is a summary of the national reports with the 1.5 Impacts of Wind Energy and Other
final results. The values vary between countries since spe- Technologies
cific peculiarities from every country have an influence on
the results due to a different range of technologies, fuels The assessment of externalities is the result of the
and pollution abatement options as well as locations. The economic valuation of impacts on the environment and
fossil fuel cycles demonstrate the highest values (coal and human health from all the activities required to
lignite, peat, oil and gas), of which gas is the least dam- produce a kWh of electricity. In order to provide an idea
aging. Renewable energy and nuclear show the lowest of the relevant impacts of wind energy and other
externalities or damages. technologies to assess the external cost, a broad
description of the impacts of wind energy and other
technologies is given.
Wind energy, a clean technology mainly due to the avoid-

ance of air pollutant emissions, is not totally free of
impacts on the environment and human health.
Wind energy has very few environmental impacts in its

operation stage, although it may cause some impact in its
direct vicinity in the form of aerodynamic noise.
Furthermore, the visual impact of large WTs on the land-
scape may adversely affect some people. Visual intrusion
of the turbines along with ancillary systems in the land-
In these results, the externalities for the nuclear cycle scape and noise are considered as amenity impacts of the
assume that waste and other hazardous impacts are well technology. Other impacts deal with indirect pollution from
managed. As the results on nuclear power plants are the production of components and construction of the tur-
based on calculations done for the ExternE project, and as bine. A brief description of wind energy impacts follows:
the calculation of the underlying accident probabilities and • Noise: coming from WT operation, installation of the
source terms have never been made available for third turbines at the wind farm site, turbine manufacturing
party analysis, these figures are not as credible as the processes, and transportation systems used in tur-
other estimates of external costs given in Table 1.2, where bine delivery and maintenance. The dominant issue is
all assumptions underlying the calculations are revealed. aerodynamic noise from the turbines. However, mod-
What is more, the numbers seem to contradict the results ern WTs are seldomly heard at distances further than
of the German reactor safety study phase B which give 300 m as background noise from wind in trees, for
rather more significant source terms and accident proba- example, will be higher.
bilities for severe core melt-down accidents with contain-
ment rupture (Gesellschaft für Reaktorsicherheit, 1989). • Visual intrusion of the turbines and associated equip-
ment in the landscape: the most difficult to quantify.
The ExternE results show that the damages vary substan- Nevertheless, the total costs are generally overesti-
tially between countries. At present these external costs mated, as the number of persons adversely affected is
are hardly ever internalised, although the EU ordinance on rather limited. In addition, since the beginning of the
subsidies for environmental measures (Official Journal of 1980s planners have become much more sophisticat-
the European Communities, 2001) states that proven ed. Today's wind power plants are erected in designat-
externalities may be compensated by public payments of ed areas, thus further limiting the number of affected
152
up to 0.05 €/kWh without being considered as subsidies. areas.
• Indirect atmospheric emissions: impacts of global warm- Table 1.3: Priority Impacts assessed in the ExternE Project
ing and acid deposition due to emissions from materials
processing and component manufacturing. Experience Fossil Fuel Technologies:
shows that these effects are in the range of less than 2% • Effects of atmospheric pollution on human health
of the emissions avoided if fossil fuels are substituted. • Accidents affecting workers and/or the public
What is more, they decline as the share of clean renew- • Effects of atmospheric pollution on:
- materials
able energy in the system increases.
- crops
- forests
• Accidents: affecting workers in manufacturing, con- - freshwater fisheries
struction and operation as well as accidents affecting - unmanaged ecosystems
the general public due to turbine operation and road trav- • Impacts of global warming
el by workers. So far, most accidents have affected work- • Impacts of noise
4
ers installing and maintaining WTs.
Specific for some Activities in Fossil Fuel Technologies:
• Impacts of coal and lignite mining on ground and surface waters
• Impact on birds: collision in flight with turbines and • Impacts of coal mining on building and construction
behavioural disturbance from blade avoidance. Although
VOLUME
• Resettlement necessary through lignite extraction
numerous studies show that birds rarely collide with rotor • Effects of accidental oil spills on marine life
blades this is an issue sometimes raised. • Effects of routine emissions from exploration, development and
extraction from oil and gas wells
• Impacts of construction on terrestrial ecosystems:

Nuclear Technologies:
long-term loss of land where turbines are placed and • Radiological and non-radiological health impacts (routine and acci-
impacts of erection activities together with electrical con- dental releases to the environment)
nections, buildings and access tracks. It has to be • Occupational health Impacts (radiological and non-radiological
noted, however, that only the access roads and a very exposures due to work accidents and radiation exposure)
• Impacts on the environment of increased levels of natural back-
small area around the tower of a WT are lost for other
ground radiation (major accident releases)
uses. The Danish and German examples show that agri-
culture goes on in wind parks, which are often used for Renewable Technologies:
grazing cattle. Wind
• Accidents affecting workers and/or the public

• Electromagnetic interference: the moving blades can • Effects on visual amenity
• Effects of noise emissions on amenity
affect radio waves and microwaves used for communi-
• Effects of atmospheric emissions (turbines’ manufacturing, on
cation purposes although this has proven to be less of
site construction and servicing)
an issue.
Hydro
These issues are explained in greater detail in the following • Occupational health effects
chapters. • Employment benefits and local economic effects

• Impacts of transmission lines on bird populations
• Damage to private goods (forestry, agriculture, water supply, ferry
In order to also give an idea of the sources of externalities
traffic)
for other fuel cycles, Table 1.3 lists the priority impacts • Damages to environmental goods and cultural objects
taken into account in the most important study available,
the ExternE project. This list only includes those impacts
which have been identified as having substantial impor-
tance. Other impacts such as land use by the installations,
153
visual intrusion and interference of transmission lines on
Source: European Commission (1999).
birds have not been included.
The nuclear fuel cycle in the ExternE project has eight With this information, it is possible to estimate the social
stages covering electricity production from the mining of cost of coal and wind power. Assuming that the cost of
uranium oxide. The impacts deriving from this fuel cycle producing a kWh with coal is around 3 c€/kWh on aver-
are caused by inhalation, external exposure and ingestion age, internalisation of the coal externalities increase
of agricultural products due to atmospheric emissions, costs by between 5 and 18 c€/kWh resulting in rather
liquid discharges and solid residues. high costs of electricity. Table 1.4 shows the social cost
of coal and gas power systems for Spain, Denmark and
The hydro power fuel cycle differs greatly from the fossil Germany in which the external cost range given for coal is
fuel cycles. The particular impacts of this cycle are the higher than the internal cost. For the case of gas the
intrusion of the infrastructure into the environment and the external cost is below the internal cost.
flooding of large areas in the case of large hydro dams.
Based on the figures given in Volume 2, the cost of produc-
1.6 Externalities of Wind Energy ing electricity with wind energy in coastal and inland sites
can be derived. These costs were based on constant 2001
Different studies and methodologies show that the exter- prices for Denmark. Taking the inland wind energy cost for
nalities of wind energy are far smaller than the external machines of 600 and 1,000 kW along with the externality
costs of fossil fuel based electricity generation. The exter- figures of Denmark from Table 1.2 the results are:
nality values shown in the final results of the national
implementation of the ExternE project (see Table 1.2) Table 1.5: Social Cost of Wind Energy
range from 0.05 to 0.25 c€/kWh.
Costs 600 kW WT 1,000 kW WT
Cost of Wind
Looking at a conventional power production technology
c€/kWh 4.4 4.1
such as coal, the values observed are of the same order
External Cost*
or double the magnitude of the internal electricity cost of c€/kWh 0.09 – 0.16 0.09 – 0.16
these technologies. In general the lower and upper levels Social Cost 4.49 – 4.56 4.19 – 4.26
are between 2 and 15 c€/kWh.
Note: *The external cost was not converted to € 2001 prices.
Table 1.4: Social Cost of Coal and Gas Powered Systems (Internal + External a)
Coal Gas
Costs Spain Denmark Germany b Spain Denmark Germany b

Internal cost c
c€/kWh 3.93 3.41 3.14 5.2 5.23 2.85
External Cost
c€/kWh 4.8 - 7.7 3.5 - 6.5 3.0 - 5.5 1.1 - 2.2 1.5 - 3.0 1.2 - 2.3
Total Cost 8.73 - 11.63 6.91 - 9.91 6.14 - 8.64 6.3 - 7.4 6.73 - 8.23 4.05 - 5.15
a
The external cost was not converted to € 2001 prices.
b
Germany coal and gas (combined cycle) cost is own calculation. Source: Hohmeyer et al. (2000).
c
Projected avoided cost of conventional power assuming 25% capacity credit for wind power (see Volume 2).
Source: Coal prices from IEA/OECD updated to € 2001 prices.
154
The social costs are practically unchanged by the inclusion 1.7 Benefits of Wind Energy
of the external cost of wind energy. Based on this total
cost comparison, the cost of wind energy is very competi- The benefits of wind energy are the avoided emissions
tive to the cost of conventional power plants as shown in and their impacts from fossil fuel electricity generation.
figure 1.1. The social cost of coal for Denmark as shown in The external costs avoidable through wind energy can be
Table 1.4 ranges from 6.9 to 9.9 c€/kWh. Figure 1.4 illus- calculated as shown in chapter 2.
tates the social cost estimated in the tables for coal, gas
and wind in Denmark. The evaluation includes damages from air pollutant emis-
sions like SO2 and NOx as well as costs of the anthro-
Figure 1.4: Social Cost of Coal, Gas and Wind in Denmark pogenic greenhouse effect resulting from CO2 emissions.
The analysis has been carried out based on a calculation
with the EcoSense model (air pollutants) on the one hand
4
and on the estimates of Azar and Sterner (1996)
concerning the adverse effects of climate change on the
other.
VOLUME
The calculations carried out for the EU-25, Turkey,
Romania and Bulgaria take into account the replaceable
energy mix of each country as well as the technological
standards. The possible ranges of reductions in external
costs due to the increased use of wind energy are shown
in Figure 1.5.
Figure 1.5: Avoidable External Costs by the Use of Wind Energy

in 2000 in c€/kWh, EU-25 and other European Countries
As was mentioned before, a precise estimation of dam-

ages is not an easy task. In addition, the results of the
national implementation phase of the ExternE project

have to be used with care since social and environmental
impacts are difficult to quantify and damages of the fuel
cycles are not fully quantified. For the case of wind ener-
gy the external costs are strongly influenced by local
factors. Thus, translating values to other locations is not
recommended. However, the results do show the order of
magnitude of the differences between clean energy tech-
nologies and conventional ways of producing electricity.
a
source of CO2 emission data: Ministry of Environment and Physical Planning:
"National Programme for the Reduction of Greenhouse Gas Emissions", Athens 2002.
b
no emission data available.
c
all data are from 2002, source: EWEA (2003b).
d
source of emission data: MVM, Hungary.
e
no data available. 155
Source: Eurelectric (2002), own calculations.
Figure 1.5 gives an overview of the avoidable external costs 1.8 Present State of Knowledge
by wind energy per kWh. It is observed that there is a notice-
able difference between the countries covered by this study. The current state of knowledge of external costs can be
Some new member states and accession countries, in par- described as a process that was mainly initiated in the
ticular, have very high emissions resulting in high external late 1980s, when the first studies were published
costs of electricity generation. attempting to quantify and compare the external costs of
electricity generation. The studies released at that time
By combining the avoidable external costs with the started a public interest in externalities, as they showed
amount of electricity produced by wind energy, the total for the first time that the differences in external costs are
amount of avoided external costs can be calculated. This of the same order of magnitude as the direct internal
is shown for the year 2000 in € millions for each country costs of generating electricity. Since that time more
in Figure 1.6. Only three countries (Denmark, Germany research and different approaches, better scientific infor-
and Spain) use substantial parts of their wind energy mation and a constant improvement of the analytical
resource to reduce external costs. This reduction is more methodologies used have driven an evolution of external-
than €1 billion per year in the case of Germany. ities research in Europe and the USA.
Figure 1.6: Total Avoided External Costs by the Use of Wind This development has resulted in a convergence of
Energy in 2000 methodologies, at least for calculating the external costs
of fossil fuel based electricity generation and wind energy.
This has induced policy-makers to adopt some measures
to attempt a first internalisation, as under the German
Renewable Energy Law.
Despite the uncertainties and debates about externali-

ties, it can be stated that with the exemption of nuclear
power and long term impacts of GHGs on climate change,
the results of the different research groups converge and
can be used as a basis for developing policy measures
aimed at a further internalisation of the different external
costs of electricity generation.
Finally, it is worth drawing attention to issues that have

not been mentioned in this chapter which may enhance
a
"National Programme for the Reduction of Greenhouse Gas Emissions",
the concept of external costs such as, for example, sus-
Athens 2002. tainability and security and reliability of supply.
b
source of emission data: TEAS, Turkey.
With respect to sustainability, the neoclassical definition
of externalities assumes that monetary valuation by man-
The ranges low, mid and high relate to the lower and ufactured and natural capital can be a substitute for envi-
upper bound and the central value of the specific exter- ronmental deterioration. This valuation is considered to
nalities per kWh shown in the Figure above. The precise be an indicator of weak sustainability (Rennings, 1996).
description of the calculations is given in chapter 2. In contrast, strong sustainability principles demand an
economic system that does not exceed the capacity of the
156
global ecological system and development that meets the
needs of the present without compromising the ability of
future generations to meet their own needs (WCED,
1987). The neoclassical definition of externalities and
sustainability principles should be linked to sustainable
development issues (Weinreich, 2002).
The security and reliability of supply and its conse-

quences for market risk is an aspect that can also
enhance the concept of externalities of electricity gen-
eration. The inclusion or accounting of market risk due
to supply disruption and, especially, fuel price volatili-
ty represents a security issue. This has an effect on
the economics of fossil fuel which is not recognised
4
in traditional analysis. Furthermore, renewable ener-
gies (e.g. wind and solar) are not subject to volatile
fuel prices. The inclusion of volatility in the private
costs equation could change the perception that
VOLUME
renewables are high cost (Awerbuch, 2003). This topic
needs further research.
157
2 ENVIRONMENTAL BENEFITS OF WIND ENERGY
2.1 Background storage (hydro) power plants are used to cover very short
load peaks, they cannot be replaced by wind energy
Emissions either, due to the latter’s intermittent nature. This
leaves electricity generation from the fossil fuels
The most important emissions concerning electricity (assuming average generation structure): hard coal, lig-
generation are CO2, SO2, NOx and PM10 (particulate mat- nite, fuel oil and gas. However, this assumption can lead
ter up to 10 micrometers in size). Emissions generally to an overestimation of the share of the replaced elec-
depend on the type of fuel used. CO2 emissions are tricity supplied by lignite, as this is predominantly used
related to carbon content. There is no realistic opportu- in the base load segment as well, and to an underesti-
nity of reducing carbon emissions by using filters or mation of substituted electricity from gas, which, due to
scrubbers, although techniques such as burning the dynamic characteristics of gas fired power plants,
fossil fuel with pure oxygen and capturing and storing lends itself perfectly to balance fluctuations in the
the exhaust gas may reduce the carbon content of emis- supply of wind energy. As we know the current mode of
sions (IPCC, 2002). For SO2, the quantity of emissions operation of conventional power plants, the rules of their
per kWh electricity generated depends on the sulphur dispatch based on the so-called “merit order” and the
content of the input fuel. Furthermore, SO2 emissions dynamic behaviour of the different types of conventional
can be reduced by filtering the exhaust gases and con- power plants, we can safely assume a replacement of
verting SO2 to gypsum or elementary sulphur. In gener- intermediate load by wind energy.
al, the sulphur content of lignite is rather high, fuel oil
and hard coal have roughly a medium sulphur content Figure 2.1: Load Curves for Lignite, Hard Coal, Fuel Oil and Gas
and natural gas is nearly sulphur free. In contrast, NOx
emissions are practically unrelated to input fuel. As NOx
are formed from the nitrogen in air during combustion,
their formation depends mainly upon the combustion
temperature. Thus, NOx emissions can be reduced by
choosing a favourable (low) combustion temperature or
by denitrifying the exhaust gases (by wet scrubbing).
Technology
Due to its intermittent nature, wind power can at present

only replace specific segments of conventional electricity
Source: based on VDEW (1998).
generation. And as it varies with available wind speed it
cannot replace conventional base load power plants. As
wind energy is a capital intensive technology, and Apart from nuclear energy, all conventional fuel types
because the fuel is free, it needs to be used as much as are more or less used to generate intermediate load
possible. Thus, it should be used to replace conventional electricity. These are: hard coal, lignite, fuel oil, natural
power plants in the intermediate rather than peak load gas and derived gas. For our analysis, the contributions
segment. of the different energy sources to intermediate load
electricity need to be specified. They probably differ sub-
Keeping these facts in mind, we can define a reference stantially in different countries and there are virtually no
system whereby wind farms may replace conventional national statistics available on their contributions.
power plants. Firstly, neither nuclear nor standard hydro Therefore, data for the German situation supplied by
158
power plants are replaceable by wind, as both almost Vereinigung Deutscher Elektrizitätswerke (VDEW, 2000)
exclusively operate in the base load segment. As pump are used as the basis of our analysis. The load curves
for one typical load day (Figure 2.1) have been derived 2.2 Electricity Generation and Emissions
for each relevant type of fuel and will be taken as the in EU-25 and other European
basis for the calculation of shares of intermediate load. Countries
The graphs show that the highest load variations during This section provides a short overview of the 28 countries
one day are displayed by fuel oil and gas. Hard coal covered by this study. The countries are divided into
shows some variation, while electricity production based groups according to their geographical location.
on lignite is almost constant. Although, these load
curves are based on the German electricity generation The EU-15 countries can be sub-divided into three groups,
structure, power plants have common fuel-specific tech- shown in Table 2.2.
nical and economic characteristics. Therefore, load
curves are assumed to have similar day-to-day variations Table 2.2: EU-15 Countries
4
in other countries.
North Central South
Based on these considerations, Table 2.1 sets out Denmark Austria Greece
assumptions for the intermediate load shares, with the
VOLUME
Finland Belgium Italy
percentage figures being based on the total volume of
Sweden France Portugal
electricity produced for each fuel.
Germany Spain
Table 2.1: Share of Intermediate Load Ireland
Luxembourg*
Fuel Type Share of Intermediate Load
Netherlands
lignite 10 %
hard coal 30 % UK
mixed firing 50 %
*data are not available for emissions in Luxembourg.
fuel oil 100 %
natural/derived gas 100 %
The 10 new member states, along with Turkey, Bulgaria

and Romania can be divided into three similar groups (see

Table 2.3).
Table 2.3: New EU Member States, Bulgaria, Romania and

Turkey
North-east East South-east
Estonia Czech Republic Bulgaria
Latvia Hungary Malta*
Lithuania Poland Romania
Slovakia Slovenia
Turkey
Cyprus
159
*data are not available for electricity generation and emissions in Malta.
2.2.1 ELECTRICITY GENERATION SECTOR AND source of statistical data for the 28 countries studied
ENVIRONMENTAL POLICY FRAMEWORK (Eurelectric, 2002) does not allow a full disaggregation
with respect to power plants suitable for more than one
The countries covered by this study differ substantially in fuel (“mixed firing”). To permit a good comparison
the volume and structure of their electricity generation. between electricity generation in all the countries, the
All data used have been taken from Eurelectric (2002). same scale is used in the two figures.
Therefore, the shares of input fuels for electricity gener-
ation vary strongly between different countries. The share As figure 2.2 shows, there are a few countries which use
of hydropower used is determined by the very different mainly hard coal and lignite for the fossil part of their
resources of the 28 countries, while the share of nuclear electricity production. These are Germany, Greece,
is a function of the nuclear energy policy of each country, Spain, Denmark, Finland and Portugal. Other countries
varying from a very strong reliance on nuclear energy in favour gas, for example the UK and the Netherlands.
the case of France to a policy of no nuclear energy in
countries like Denmark and Austria. As has been Some of the new member states and others mainly use
explained above, intermittent renewable energy cannot at hard coal and lignite for their fossil fuel based electricity
present replace nuclear or hydro power. Thus only fossil generation. These are Poland, Slovenia, Czech Republic,
fuels are replaced by wind energy in this study. The struc- Bulgaria, Slovakia and Hungary. Natural gas is favoured
ture of electricity generation by fossil fuel fired conven- by Latvia, Turkey and Romania. The majority of these
tional thermal power plants is shown in Figures 2.2 and countries use a substantial share of nuclear energy for
2.3. Unfortunately, the only available comprehensive electricity generation.
Figure 2.2: Total Electricity Generation in EU-15 Countries in 2000

Source: Eurelectric (2002).
160
Figure 2.3: Total Electricity Generation in the 10 New Member States, Turkey, Bulgaria and Romania in 2000
VOLUME
a
data are from 2002. Source: EWEA (2003b).
b
no data available.
Figure 2.4: Total Electricity Generation in the EU-15, EU-25 For a better orientation, the amounts of electricity genera-
and all 28 Countries in 2000 tion in the EU-15, the EU-25 and in all 28 countries are
shown in Figure 2.4. As this figure illustrates, the amount
of electricity generation is very low in most of the countries

outside the EU-15.
Figures 2.5 to 2.7 provide a detailed picture of the elec-

tricity generation fuel mix in the various countries.
Source: Eurelectric (2002), own calculations. Data for Estonia are from 2002,
source: EWEA (2003b). No data are available for Malta.
161
Figure 2.5: Segmentation of Fuels for Electricity Generation in EU-15 Countries in 2000 (%)
a
data are from 2002, source: EWEA (2003b).
Figure 2.6: Segmentation of Fuels for Electricity Generation in the 10 New Member States, Turkey, Bulgaria and Romania in 2000 (%)
162 a
no data available.
Figure 2.7: Segmentation of Fuels for Electricity Generation in Figure 2.8: Electricity Generation by Wind Energy
the EU-15, EU-25 and all 28 Countries in 2000 (%) in 2000 (TWh/a)
VOLUME
Source: Eurelectric (2002), own calculations. Data for Estonia are from 2002,
source: EWEA (2003b). No data are available for Malta. Source: Eurelectric (2002).
Electricity generation by renewable energies is widely spread 2003), Denmark’s wind generation figures for 2002
around the European countries. The specific amount which showed production at 4.9 TWh representing a share of
is covered by renewable energies in each depends on geo- 14.8% (Danish Wind Industry Association, 2003) while
graphical conditions and the country’s policies on renewable Spain’s production of 9.5 TWh for 2002 represents 4%
energies. Therefore, the use of renewables differs widely (IDAE, 2003).
between the 28 countries. Due to its relatively low internal
costs, hydro power is used by most, with Austria and Latvia Here, it is very important to point out that the figures for
producing more than half their electricity from hydro power. electricity generation by wind energy have increased dra-
matically in recent years. In terms of installed wind capac-
The use of wind energy differs very substantially across ity, Europe experienced a growth of 10,200 MW of total
the 28 countries, with Germany, Denmark and Spain pro- installed capacity from 2000 to 2002. This fact has an

ducing more than 4 TWh/annum (2000) and nine other impact on the quantification of the benefits of wind ener-
EU-15 member states producing up to 1 TWh/annum. Of gy. However, for the purposes of this study, the figures for
all the other countries, only Turkey was using any signifi- electricity generation by wind energy were taken from the
cant amount of wind energy in the year 2000. Countries year 2000 (Eurelectric, 2002).
generating electricity from wind energy (2000) are shown
in Figure 2.8. Although wind energy is only used in significant volumes in
just four out of the 28 countries, the use of wind energy in
If wind energy production is looked at in terms of share of the year 2000 has already resulted in significant emission
electricity produced, a somewhat different picture reductions, which are discussed in chapter 2.4 below.
emerges, as only Denmark produced more than 10% of its
electricity from wind in 2000 (12.8%), while Spain (2.1%), 2.2.2 EMISSION DATA
Germany (1.8%) and Greece (1.1%) were way behind.
Nevertheless, the situation is changing dramatically; for To be able to analyse the possible environmental and
example in Germany the installed capacity has more than health benefits of the use of wind energy we need to
163
doubled since the year 2000. For 2002, Germany pro- know the specific emissions of the electricity replaced
duced 23.1 TWh which represents a share of 4.7% (Ender, by wind. These can be derived by dividing the absolute
emissions produced by a type of fuel in kilotons of Average emissions per kWh were calculated to provide a
CO2/annum used for electricity generation in one country starting point for examining the relationship between elec-
by the electricity produced from this fuel in kWh/annum. For tricity production from fossil fuels and total emissions
clarity, the emissions statistics for each country are given from the electricity sector in the different countries (see
on the CD attached to this report. Most of the data used for Figures 2.9 to 2.11). The results include all fossil fuel
the calculations are from Eurelectric (2002). However, not based electricity not just the intermediate load segment
all the necessary data were available from this source, so (see chapter 2.4).
some calculations have been based on additional sources.
Figure 2.10: Specific Average SO2 Emissions in g/kWh from
As explained in chapter 2.1, wind energy is capable of Fossil Fuel Electricity Generation in 2000
replacing intermediate load conventional power production.
The emissions avoided by wind energy depend on three fac-
tors: the specific emissions from each type of
generation facility; the fuel mix in each country; and the per-
centage of each fuel replaced by wind energy. A detailed
calculation of avoidable specific emissions by wind energy
in all the countries studied is shown in chapter 2.4.
Figure 2.9: Specific Average CO2 Emissions in g/kWh from

Fossil Fuel Electricity Generation in 2000
a
b
c
d
no data available.
e
source of emission data: NEK, Bulgaria.
f
a
"National Programme for the Reduction of Greenhouse Gas Emissions", Athens
(2002).
b
c
d
e
no data available.
f
164 g
Figure 2.11: Specific Average NOX Emissions in g/kWh from countries. This is related to differences in fuel mix as well
Fossil Fuel Electricity Generation in 2000 as the fact that some countries have power plants with
very low efficiencies.
The distribution of SO2 emissions per kWh is very differ-

ent, as shown in Figure 2.10. This is related to the very
heterogeneous sulphur content of fuel and the use of
desulphurisation in only the most advanced countries.
NOx emissions differ between the countries according to

the combustion process used, the combustion tempera-
ture, which is not optimal in all the countries, and the
4
scrubbing technologies employed, as shown in Figure 2.11.
To determine avoidable emissions from the use of wind

energy, specific emissions from electricity generation for
VOLUME
the different fuels must be calculated. Specific emissions
have been evaluated based on total emissions from elec-
tricity generation and the amounts of electricity generated
in each country. For further information about this calcu-
a
b
all data are from 2002, source: EWEA (2003b). lation see Appendix G.
c
d
no data available.
e
f
2.3 The Calculation of External Costs
with the EcoSense Model
In summary, Figures 2.9 to 2.11 show that the southern In order to be able to calculate the external costs avoided
European EU-15 countries (Greece and Spain), as well as by wind energy, it is necessary to model the pathway of
all south-eastern countries (Bulgaria, Cyprus, Romania, emissions from conventional power plants to the different

Slovenia and Turkey), Hungary and Estonia have rather receptors, such as plants, animals and humans, which
high emissions from electricity generation. Two of the cen- may be located thousands of kilometres away. As air pol-
trally-located countries (Ireland and the UK), Italy and lutants can damage a number of different receptors, the
Portugal, Lithuania, and most of the eastern European new task of analysing the impacts of any given emission is fair-
member countries (Czech Republic, Poland and Slovakia) ly complex. To allow such complex analysis, a tool has
show intermediate emission levels. Countries with rather been developed during the last 10 years in a major co-
low emissions are mostly northern or central countries - ordinated EU research effort, the EcoSense model. This
Denmark, Finland, Sweden, Austria, Belgium, France, chapter explains the basics of the model, which is used in
Germany, the Netherlands, and Latvia. the calculations in chapter 2.4.
Due to this distribution there is a significant increase of 2.3.1 SOFTWARE DESCRIPTION

specific average emissions from electricity generation
from northern to south-eastern Europe. EcoSense is a computer model for assessing environ-
mental impacts and the resulting external costs of electric
165
Figure 2.9 shows that the difference in specific CO2 emis- power generation systems. The model is based on the
sions is more than a factor of three between the various Impact Pathway approach of the ExternE project and
provides the relevant data and models required for an tional emissions on these receptors can be calculated on
integrated impact assessment related to airborne pollu- the basis of so-called dose response functions. Important
tants. (For extensive information on the model as well as data on receptors included in the model database are, for
the approach used, see European Commission, 1994) example, and population density and land use patterns.
EcoSense provides the windrose trajectory model (WTM) • Monetisation of Costs

for modelling the atmospheric dispersion of emissions,
including the formation of secondary air pollutants. For The last step is to monetise the impacts per kWh caused
any given point source of emissions (e.g. a coal fired by the specific power plant. In this stage, the calculated
power plant) the resulting changes in the concentration physical damage to a receptor is valued on a monetary
and deposition of primary and secondary pollutants can scale based on the best available approaches for each
be estimated on a Europe-wide scale with the help of this type of damage.
model. Developed in the UK by the Harwell Laboratory it
covers a range of several thousand kilometres. The refer- 2.3.2 INPUT DATA TO THE MODEL
ence environment database, which is included in EcoSense,
provides receptor-specific data as well as meteorological As the EcoSense model requires a specified site as a start-
information based on the Eurogrid-co-ordinate system. ing point for its pollutant dispersion modelling we have
chosen one typical electricity generation site for each coun-
The Impact Pathway approach can be divided into four try to assess the impacts and calculate the costs caused
analytical steps: by emissions from fossil fuel fired power plants which may
be replaced by wind energy.
• Calculation of Emissions
The co-ordinates at each site are chosen in order to locate
the reference plants centrally in the electricity generating
activities of each country. Thus, it is assumed that the cho-
sen site represents approximately the average location of
electricity generating activities of each country has been
chosen. For more information about the input data see
Appendix H.
To control for effects caused by this assumption and to

prevent extreme data results, a sensitivity analysis was
The first step is to calculate emissions of CO2, SO2 and carried out by shifting the geographical location of the
NOx per kWh from a specific power plant. plant. This analysis showed a relatively high sensitivity of
external costs to the location of the electricity generation
• Dispersion Modelling facilities. This is due to the very heterogeneous distribu-
tion of the different receptors in different parts of a coun-
Then air pollutant dispersion around the site of the specif- try. For this reason, the specific external costs per kWh
ic plant is modelled. Based on meteorological data, may differ by a factor of two. Unfortunately, the area cov-
changes in the concentration levels of the different pollu- ered by EcoSense is limited to 29° east, so substantial
tants can be calculated across Europe. parts of eastern Europe are not included in the analysis
and the impacts of eastward emissions due to the prevail-
• Impact Analysis ing westwind drift are not fully accounted for. Thus, in
166
Based on data for different receptors in the areas with sig- countries located at the border of the area covered exter-
nificant concentration changes, the impacts of the addi- nal costs may be substantially underestimated.
In order to run the model, the capacity of the power emissions from lignite but low on SO2 emissions from oil.
plant, its full load hours of operation and the volume (The calculations are described in detail in Appendix G.)
stream of exhaust gas per hour are required. The
assumptions made for the calculations are shown in The specific emissions per fuel and the share of interme-
Table 2.4 for the different fossil fuels. diate load generated on the basis of each fuel are used
to calculate the specific emissions which could have been
Table 2.4: Technical Data of the Reference Facilities avoided per kWh of wind energy in each country in 2000.
Assumed for the Calculation Results are shown in Figures 2.12 to 2.14. Due to a lack
Fuel Type Capacity Full Load Volume Stream
of sufficient data there are no results for Luxembourg and
(MW) Hours per Year per Hour (m )
3 Malta.
Hard coal 400 5,000 1,500,000
Lignite 800 7,000 3,000,000 Due to the fact that wind energy replaces only part of the
4
Fuel oil 200 2,000 750,000
electricity produced by fossil fuels (intermediate load),
specific avoidable emissions are different from average
Natural gas, 200 2,000 750,000
derived gas emissions from fossil fuel electricity generation. In most
cases, avoidable emissions by wind energy are less than
VOLUME
Mixed firing, 400 5,000 1,500,000
not specified average emissions from fossil fuel electricity generation.
This is justified by the fact that intermediate load elec-
tricity generation by fossil fuels is based on fuel with rel-
For each country, calculations have been performed for a atively low emissions (see chapter 2.1 for more informa-
representative power plant location based on the specific tion on this point).
national emission data for each fuel and each
pollutant. As is to be expected, the specific emissions of intermedi-
ate fossil power which could be avoided by using wind
energy, are higher in most new member states than in
2.4 Benefits of Wind Energy - Results most of the EU-15. This is due to less efficient power
plants and a lack of SO2 and NOx scrubbers.
2.4.1 AVOIDABLE EMISSIONS BY THE USE OF Consequently, new wind energy plants in the countries
WIND ENERGY besides EU-15 countries could induce higher specific

emissions. Nevertheless, this may not hold in the long run,
As explained in chapter 2.1, electricity from wind energy as a convergence of technical standards is expected in the
can replace intermediate load from fossil fuel power next 20 years.
plants. Avoidable emissions by wind energy can be calcu-
lated based on specific emissions derived in chapter 2.2. Figure 2.8 reveals that some countries are already avoiding
Due to the fact that there are no data available on specific a sizeable amount of fossil fuel emissions through their use
emissions per fuel for most countries, specific emission of wind energy. Due to the different specific emissions
data have been estimated by splitting up the total emis- avoided per kWh in each country (Figures 2.12 to 2.14) the
sions from conventional thermal electricity generation total emissions are not directly proportional to the wind
based upon the shares of electricity generated by the dif- energy produced. For Spain, in particular, total emission
ferent fossil fuels. Different power plants running on the reductions for SO2 and NOx are comparatively high in rela-
same fuel are assumed to have the same specific emis- tion to the electricity replaced. This is due to the high spe-
sions in any one country. Furthermore, it is assumed that cific emissions of Spanish fossil fuel power plants.
the countries have attained the same relative emission
167
abatement level for each fuel type. That is to say, for In 2000, approximately 15 Mt CO2 were avoided by the
example, that one country would not rank high on SO2 use of wind energy as shown in Figures 2.15 to 2.17.
Figure 2.12: Specific Avoidable CO2 Emissions in g/kWh by Figure 2.13: Specific Avoidable SO2 Emissions in g/kWh by
Wind Energy in 2000 Wind Energy in 2000
a
(2002).
b
no emission data available. a
c
all data are from 2002, source: EWEA (2003b). b
d
source of emission data: MVM, Hungary. c
e
no data available. d
no data available.
f
source of emission data: NEK, Bulgaria. e
g
source of emission data: TEAS, Turkey. f
Source: Eurelectric (2002), own calculations. Source: Eurelectric (2002), own calculations.
2.4.2 AVOIDABLE EXTERNAL COSTS BY THE Figure 2.14: Specific Avoidable NOx Emissions in g/kWh by
USE OF WIND ENERGY Wind Energy in 2000
To calculate the external costs avoided by the use of

wind energy, the external costs resulting from air pollu-
tants such as SO2 and NOx (calculated by EcoSense)

have to be added to the external costs of the anthro-
pogenic greenhouse effect resulting from CO2 emissions,
which are not calculated by EcoSense.
As air pollutants can damage a large number of different

receptors, calculations of external costs will generally
include a large number of damages, which tend to be
restricted to the most important impacts to allow a calcu-
lation of external costs with a limited resource input. At
present, EcoSense includes the following receptors:
humans (health), crops, materials (in buildings, etc.),
a
forests and ecosystems, with monetary valuation only b
included for human health, crops and materials. c
168 d
no data available.
e
There are two approaches to evaluating effects on human f
health: value of statistical life (VSL); and years of life lost
(YOLL). The VSL approach measures a society’s willingness Figure 2.16: Total Avoided SO2 Emissions in kt/annum by
to pay to avoid additional deaths. This can be seen in spend- Wind Energy in 2000
ing on improved safety in the aircraft or car industry. In the
EU and the US, figures of between US$/€1 million and
US$/€10 million per life saved have been found in different
studies. Earlier versions of the ExternE project adopted a
figure of US$3 million per life saved for VSL calculations. In
these calculations a person’s age does not matter.
The YOLL approach takes age into account. In the case

of chronic disease leading to death in a very old person,
only the years of life lost due to the disease as com-
4
pared to average life expectancy are taken into account.
For each year of life lost approximately one-twentieth of
the VSL value is used.
VOLUME
Figure 2.15: Total Avoided CO2 Emissions in kt/annum by a
Wind Energy in 2000
Unfortunately, EcoSense does not provide a calculation

based on the VSL approach. As pointed out above, VSL
may lead to substantially higher external costs than the
YOLL approach which is applied by the EcoSense model.
Results of former ExternE studies estimate external
costs based on both approaches. These resulted in VSL
results approximately three times higher than those
found with YOLL (Umweltbundesamt, 2002).
As the present version of EcoSense does not calculate

VSL values, the EcoSense results on human health
effects based on the YOLL approach have been scaled.
This has been done with a factor of one for low damage
a
source of CO2 emission data: Ministry of Environment and Physical Planning: cost estimates calculated for human health, a factor of
Athens (2002).
two for medium cost estimates and a factor of three for
b
source of emission data: TEAS, Turkey. high estimates.
As EcoSense does not calculate long-term damage from

Using one or other approach may lead to substantially CO2-induced climate change, the estimates of Azar and
different results of monetised human health damages. Sterner (1996) are used. As CO2 remains in the atmos-
Deciding which approach to use is a value judgement, phere for about 100 years, most of the damage will occur
based on society’s underlying value system. Thus, cal- in the distant future. If these damages apply to human
culations of the external costs of human health dam- health or irreversible environmental damages Rabel
ages should always give both measures and leave it up (1999) has strongly argued that no discounting should be
169
to the reader or the policy-maker to decide which applied, as the valuation of the damage increases with
approach they think most appropriate. the discount rate. Based on a discount rate of 0%, dam-
Figure 2.17: Total Avoided NOx Emissions in kt/annum by age costs of global warming are calculated by Azar and
Wind Energy in 2000 Sterner (1996) to be € (2000) 87.51 - 607.41/ton of car-
bon (see Appendix I). Recalculation in terms of CO2 emis-
sions leads to costs of € (2000) 23.87 - 165.69/ton of
CO2 (Umweltbundesamt, 2002). The remaining large
range of the estimate is due to the time period taken into
account for the analysed damages (300 or 1,000 years)
and the way the question of damage in poor countries is
dealt with.
Based on the EcoSense calculation, the avoidable exter-

nal costs per kWh by wind energy have been evaluated.
Results are shown in Figure 2.18.
The total avoided external costs in 2000 are shown in

Figure 2.19.
As can be seen in Figures 2.19 and 2.20, an amount of

nearly €1.8 billion has been avoided by the use of wind
a
source of emission data: TEAS, Turkey. energy electricity generation in 2000. Most of this applies
to Germany (38%), Spain (31%) and Denmark (15%).
170
Figure 2.18: Avoidable External Costs in c€/kWh through the Use of Wind Energy in 2000, EU-25 and other European Countries
VOLUME
a
source of CO2 emission data: Ministry of Environment and Physical Planning: "National Programme for the Reduction of Greenhouse Gas Emissions", Athens (2002).
b
c
all data are from 2002, source: EWEA (2003b)
d
e
no data available.
f
g
Figure 2.19: Total Avoided External Costs in €million/annum Figure 2.20: Shares of Total Avoided External Costs by the
by the Use of Wind Energy in 2000 use of Wind Energy in Europe 2000
a
source of CO2 emission data: Ministry of Environment and Physical Planning: a
"National Programme for the Reduction of Greenhouse Gas Emissions", "National Programme for the Reduction of Greenhouse Gas Emissions", 171
Athens (2002). Athens (2002).
b
source of emission data: TEAS, Turkey. Source: Eurelectric (2002), own calculations. Source of emission data for Turkey: TEAS,
Source: Eurelectric (2002), own calculations. Turkey.
3 STANDARD METHODOLOGY FOR CALCULATION OF
EMISSION REDUCTIONS
As shown above, the calculation of emission and external specific emissions (in mg/kWh) and avoidable total emis-
cost reductions achieved by the use of wind energy in the sions (in kt/annum).
EU-15 and the 10 new member states along with Turkey,
Bulgaria and Romania can be based either on the As the future emission reductions due to the use of wind
EcoSense model on the one hand or the regular reporting energy cannot be calculated on the basis of present con-
of electricity generation and emissions by Eurelectric ventional electricity generating technologies and fuel mix,
(2002) on the other. a forecast of future fuel mix and conventional technolo-
gies must be made.
Forecasts of possible future emission reductions and
reductions in resulting external costs can be carried out Based on the specific avoidable emissions and the fore-
on this basis. Like the calculation of preceding emission casted amount of electricity generated by wind energy, the
reductions it can be divided into two parts: avoidable total amount of avoidable emissions can be calculated.
172
4 ANALYSIS OF EMISSION REDUCTIONS
The potential of future emission reductions has been car- Figure 4.2: Specific Avoidable SO2 Emissions in g/kWh by
ried out based on data for 2020. The year 2020 has been Wind Energy in 2020, EU-25 and other European Countries
chosen as the last available year in Eurelectric forecasts.
The options are combined with the volume of convention-
al electricity replaced by wind energy in Europe forecasted
for the year 2020 by the EWEA (2003a).
4.1 Avoidable Specific Emissions

through Wind Energy
Future avoidable specific emissions through the use of
4
wind energy are shown in Figures 4.1 to 4.3.
Figure 4.1: Specific Avoidable CO2 Emissions in g/kWh by

a
all data are for 2010.
Wind Energy in 2020, EU-25 and other European Countries
VOLUME
b
all data are from 2000.
c
d
e
f
g
no data available.
h
i
Figure 4.3: Specific Avoidable NOx Emissions in g/kWh by


a
b
c
Athens (2002).
d
e
f
g
h
no data available.
i
j
a
b
c
d
e
f
g
no data available.
h
i
173
The figures show that specific avoidable emissions are 4.2. Avoidable Total Emissions Through
going to decrease from 2000 to 2020. This is due to two Wind Energy
factors. Firstly, the fuel mix is going to change in coming
decades in most of the countries covered by this study. In Based on the expected amount of electricity generated by
many cases, high emission fuels will partly be replaced by wind energy, avoidable total emissions have been calcu-
those with relatively low emissions. Accordingly, the share lated (see Figures 4.5 to 4.7).
of fuel oil and, in particular, natural and derived gas will
increase significantly. Parallel to this, the amounts of Forecasts of electricity generation by wind energy are
electricity generated by hard coal and lignite are going to based on data from the EWEA (2003a) relating to total
decrease or stagnate. This will lead to a lower volume of electricity generation and on data from Eurelectric (2002)
specific avoidable emissions by wind energy in 2020 com- concerning the distribution of generation between the
pared with 2000. countries.
Secondly, there will be a significant improvement in the Figure 4.4: Electricity Generation by Wind Energy in
technology of fossil fuel based electricity generation. The TWh/annum in 2020
east-European states, in particular, will up-grade their
technology by fitting SO2 scrubbers and improving com-
bustion processes to reduce NOx emissions.
Source: EWEA (2003a), own calculations.
As shown in Figure 4.4, the amount of electricity generated

by wind energy will increase strongly from 2000 to 2020.
For 2020, a total of 425 TWh/annum is forecasted by the
EWEA (2003a) for the EU-25 countries. For all 28, this
would lead to a forecast of more than 450 TWh/annum in
2020, an increase of nearly 2,000% within 20 years.
As shown in Figures 4.5 to 4.7, total avoidable emissions

depend on the level of specific avoidable emissions in
each country. Therefore, the total avoidable emissions are
174
not only related to the amount of electricity generated by
wind energy.
Figure 4.5: Total Avoidable CO2 Emissions in kt/annum by Figure 4.7: Total Avoidable NOX Emissions in kt/annum by
Wind Energy in 2020, EU-25 and Turkey Wind Energy in 2020, EU-25 and Turkey
VOLUME
a
all data are from 2000. Calculation of electricity generation by wind energy
based on data for 2010 resp. 2020.
b
"National Programme for the Reduction of Greenhouse Gas Emissions", a
Athens (2002). based on data for 2010 resp. 2020.
c
all data are for 2010. Calculation of electricity generation by wind energy based b
all data are for 2010. Calculation of electricity generation by wind energy based
on data for 2010 resp. 2020. on data for 2010 resp. 2020.
d
source of emission data: TEAS, Turkey. c
Source: Eurelectric (2002), EWEA (2003a), own calculations. Source: Eurelectric (2002), EWEA (2003a), own calculations.
Figure 4.6: Total Avoidable SO2 Emissions in kt/annum by As can be seen in Figures 4.6 and 4.7, the potential for
Wind Energy in 2020, EU-25 and Turkey emission reductions is very high in Spain. This is again
explained by SO2 and NOx emissions, which are forecast
to still be relatively high in Spain in 2020 in comparison

with other countries.
Figures 4.8 to 4.10 show the shares of total avoidable

emissions in Europe in 2020. Again, avoidable emissions
in Spain are a lot higher than in the UK, for example, even
though wind energy generation in the UK will be at the
same level as that in Spain. A comparison of the wind
energy electricity generation capacity of Spain and
Germany shows that Spain, according to Eurelectric esti-
mates, will be producing twice the amount of wind gener-
ated electricity in 2020 than Germany. But the total
avoidable SO2 emissions in Spain will be 10 times higher
a
than Germany’s.
b
all data are for 2010. Calculation of electricity generation by wind energy 175
c
Source: Eurelectric (2002), EWEA (2003a), own calculations.
Figure 4.8: Avoidable CO2 Emissions in Mt/annum Wind Figure 4.10: Avoidable NOX Emissions in kt/annum by Wind
Energy in 2020 Energy in 2020
a
b
source of emission data: TEAS, Turkey. a
Source: Eurelectric (2002), EWEA (2003a), own calculations. Source of CO2 based on data for 2010 resp. 2020.
emission data from Greece: Ministry of Environment and Physical Planning: b
"National Programme for the Reduction of Greenhouse Gas Emissions", Athens Source: Eurelectric (2002), EWEA (2003a), own calculations. All data from
(2002). All data from Cyprus are for 2010. Calculation of electricity generation Cyprus are for 2010. Calculation of electricity generation by wind energy for
by wind energy for Cyprus based on data for 2010 resp. 2020. Cyprus based on data for 2010 resp. 2020.
Figure 4.9: Avoidable SO2 Emissions in kt/annum by Wind

Energy in 2020
a
b
Source: Eurelectric (2002), EWEA (2003a), own calculations. All data from
176 Cyprus are for 2010. Calculation of electricity generation by wind energy for
Cyprus based on data for 2010 resp. 2020.
4.3 Avoidable External Costs Through Figure 4.12: Total Avoidable External Costs in €Million/annum
Wind Energy by the Use of Wind Energy in 2020, EU-25 and Turkey
In line with the specific emissions, the avoidable specif-

ic external costs in c€/kWh decrease from 2000 to
2020, especially in south-eastern European states where
avoidable costs are significantly less than in 2000 (see
Figure 4.11).
Figure 4.11: Avoidable External Costs in c€/kWh by the Use of

VOLUME
a
b
(2002).
c
all data are for 2010. Calculation of electricity generation by wind energy based
on data for 2010 resp. 2020.
d
Total avoidable external costs in 2020 are shown in Figure

a
4.12. Spain will take over pole position from Germany in
b
c
source of CO2 emission data: Ministry of Environment and Physical Planning: avoiding external costs by the use of wind energy.
(2002).

d
e
f
g
h
no data available.
i
j
Nevertheless, total annual avoidable external costs in

2020 are much higher than in 2000. They are expected
to increase from €1.8 billion in 2000 to more than €25
billion a year in 2020 because of the expected increase
in electricity generation by wind energy, from 22 TWh/a
in 2000 to more than 450 TWh/a in 2020. While elec-
tricity generation by wind energy is expected to increase
177
by nearly 2,000% from 2000 to 2020, avoidable exter-
nal costs will increase by about 1,400%.
Figures 4.12 and 4.13 show that many more countries will
take part in avoiding external costs by the use of wind
energy in 2020 than in 2000. Each of the seven countries
shown in Figure 4.13 will avoid more external costs in
2020 by using wind energy than all the countries togeth-
er in 2000 (each more than €1.8 billion a year); some of
them are expected to avoid more than three times this
amount (e.g. €6.5 billion in the case of Spain).
Figure 4.13: Shares of Total Avoidable External Costs by the

Use of Wind Energy in Europe in 2020
a
b
Source: Eurelectric (2002), EWEA (2003a), own calculations. Source of CO2

emission data for Greece: Ministry of Environment and Physical Planning:
(2002). All data for Cyprus are for 2010. Calculation of electricity generation by
wind energy based on data for 2010 resp. 2020.
178
5 PUBLIC ACCEPTANCE ANALYSIS
Previous chapters have reported the environmental bene- 5.1.1 VISUAL IMPACT
fits of wind energy in the current European electricity sup-
ply system and shown its potential future benefits. How The siting of WTs affects the visual or aesthetic properties
much of this potential can be achieved? What are the of the surroundings, especially in locations where people
main obstacles to overcome, so that the sector continues place a high value on the landscape. This is referred to as
to grow? These questions are being addressed through the ‘visual impact’ of wind energy. Visual impact has a
policy instruments, financial mechanisms, national renew- direct effect on amenity, defined as resources available
able energy targets, R&D programmes, etc. Combined for people’s convenience, enjoyment and comfort, in this
with increasing public awareness of climate change and case a landscape.
sustainability, these are important drivers for a thriving
renewable energy sector. But the most important issue is A landscape attracts different perceptions since aes-
public acceptance, especially by local communities and thetic values such as beauty and diversity are subjective
4
individuals living at prospective wind farm sites. Whether (Schwahn, 2002), while its value will also be influenced
an installation goes ahead or not often relies on them. by use (e.g. national park, wildlife habitat, agricultural
land).
The purpose of this chapter is to give an overview of the
VOLUME
environmental impacts and other factors affecting public Protected areas of national or regional importance are
acceptance of wind energy. more sensitive to the visual impact of wind energy. In
addition, wind energy may compete with other public uses
such as recreation, agriculture, tourism, wildlife conserva-
5.1 Environmental Impacts of Wind tion, and others.
Energy
The perceptions of individuals in communities affected by
Although wind energy is a clean technology, it is not free wind energy will depend on their attitudes to scenery and
of impacts on the environment. Wind energy has a num- natural beauty, the existing level of visual amenity and
ber of special features, including: their general attitude to WTs (Manwell, 2002).
• More than one wind turbine (WT) is needed for
large-scale production. Modern turbines are becoming larger both in size and
• WTs are mainly located in remote and rural areas capacity, and hence more dominant in the landscape. At

where the wind resource is present. the same time, the spacing between turbines is increas-
• The turbines may be visible from a great distance. ing, thus lessening their density in a given area. The
• The movement of the blades (flickering) may draw development of the technology is therefore changing the
attention. visual impact of wind farms from high density groupings
with high rotational speeds to fewer, larger machines
As well as these visual impacts, wind energy is associat- operating at lower rotational speeds.
ed with other environmental issues such as noise, land
use and impacts during the construction phase. Some Other visual impacts of WTs are lighting and, in the
impacts, such as those on birds and flickering can be vicinity of airports for example, marking to reduce bird
measured quantitatively; others, such as visual intrusion collisions. Ancillary facilities such as stores, substa-
and noise require more subjective and qualitative criteria. tions, transmission lines and roads also impact on
amenity.
These impacts are considered in this section. In addition,
an analysis of the primary energy consumption of a WT In order to maintain public acceptance, wind farms need
179
compared with a coal fired power plant is given. to be designed in such a way as to minimise these vari-
ous aesthetic and amenity impacts (see Table 5.1).
Table 5.1: Aesthetic Guidelines for Wind Plants (residential, industrial, tourist). The interaction of these factors
lessens or enhances the perception of sound from WTs.
Ensure visual uniformity (direction of rotation, type of turbine and
tower, and height) Physically, sound is a pressure variation detected by the ear;
Avoid fencing It depends on the source and the medium through which it
Minimise or eliminate roads travels. The speed of sound is about 340 m/s in atmos-
Bury intraproject power lines pheric air. It is important to make a distinction between
Limit or remove ancillary structures from site sound power level and sound pressure level. The former is a
Remove inoperative turbines property of the source of the sound whereas sound pressure
Avoid steep slopes level is a property of the sound at a given observer location.
Control erosion and promptly revegetate Noise is measured in decibels (dB) and the scale employed
Remove litter and scrap (dBA) is weighted to the range perceived by the human ear.
Clean dirty turbines and towers Table 5.2 shows a comparison of different power and pres-
sure levels of sound to indicate what can be considered a
Source: Gipe (1995).
threshold of hearing or a pain threshold.
The use of a computer simulation to generate a virtual The most important factors affecting noise propagation
image of the proposed wind farm can help developers and are: type of noise source, distance from source, wind
planners assess its visual impact. speed, temperature, humidity, precipitation and the pres-
ence of barriers and buildings. The factors with the most
The visual impact of wind energy has a big influence on influence on noise propagation are the distance of the
public perception and acceptance of the technology. source from the observer and the type of source.
Efforts to integrate WTs aesthetically into the landscape
and the sharing of economic benefits with local communi- Table 5.2: Level of Sounds
ties may help to soften negative attitudes to wind energy. Source Distance from Sound Level Environmental Subjectivity/
the Source (dBA) Noise* Impression
These aspects are discussed later in this section. (m)
Civil defence 140-130 Threshold of
siren pain
5.1.2 NOISE
Jet take-off 61 120
110 Rock concert Very loud
Pile driver 15 100
Noise is defined as an unwanted sound. It can be meas- Ambulance siren 31 90 Boiler room
ured quantitatively, and regulations exist to limit noise Freight train 15 80
Pneumatic drill 15 80 Printing press Loud
levels, but it also has a subjective element. Manwell Motorway traffic 31 70 Moderately
(2002) classifies the effects of noise from wind energy Vacuum cleaner 31 60 Data processing loud
centre
into two main categories: Department
• Subjective effects including annoyance, nuisance, store/office
Light traffic 31 50 Private business
dissatisfaction. office Quiet
WT > 1MW 200 49
• Interference with activities such as conversation.
WT > 1MW 300 45
Large transformer 61 40
Soft whisper 2 30 Quiet bedroom
Noise from WTs comes from the sound produced by the turn- 20 Recording
ing blades and from the gearbox, generator and hydraulic sys- studio
10 Threshold
tems (although in modern WTs this mechanical noise has 0 of hearing
been reduced almost to zero). As with other impacts of wind
energy, perception of the noise depends on local features (e.g. WT data is an estimation for illustrative purposes (University of Flensburg).
180 * Environmental noise is shown as an equivalent noise source at the sound level given.
rural or urban area, topography), number and distance of resi-
Source: National Wind Co-ordinating Committee (2002).
dents from the WT site, and the type of community affected
From the table, it can be seen that distance plays an dBA L90 means that the sound level must not exceed
important role in the perceived sound level. The noise the level indicated 90% of the time.
from a WT can reach moderate sound pressure levels • Leq (equivalent sound level): The average A-weighted
(< 50 dBA) when the distance from the turbine to the sound pressure level which gives the same total energy
receptor is between 200 and 300 m. Typically, the sound as the varying sound level during the measurement
power level of a modern WT is between 100 and 106 dBA period.
depending on the type of turbine and the wind speed at • Ldn (day night level): The average A-weighted sound
which the sound is measured (typically 8 m/s). pressure level during a 24-hour period, obtained after
adding 10 dBA to levels measured in the night between
The decibel scale must be carefully interpreted when 10 p.m. and 7 a.m.
evaluating the number of turbines to be placed and their
effects. A WT with a capacity higher than 1 MW has a Table 5.3 shows the noise limits of sound pressure levels
4
sound power level of 104 dBA for example. The installa- in some European countries. State-of-the-art turbines with
tion of a second turbine with the same sound power level capacities higher than 1 MW generally have sound power
will cause an increase of only 3 dBA. Increasing the ener- levels of between 100 and 106 dBA. Thus, a modern tur-
gy of a sound by 26% raises the noise power level bine has to be placed at a distance of between 200 m
VOLUME
1 dBA. Tripling the energy of a sound yields an increase and 300 m from the receptor to reach a sound pressure
of 5 dBA. The dBA scale is a logarithmic scale. In other level of between 45 dBA and 50 dBA (see Table 5.2).
words, as the sound power is doubled (two turbines) the
index increases by approximately 3 dBA. A sound level of Table 5.3: Legal Noise Limits in dBA
100 dBA thus contains twice the energy of a sound level
Commercial Mixed Residential Rural
of 97 dBA. The sound level decreases with greater dis-
Germany
tance from the source by approximately 6 dBA every time
Day 65 60 55 50
the distance is doubled (Gipe, 1995).
Night 50 45 40 35
Netherlands
In summary, the total perceived noise is the relative sum
Day (Leq) 50 45 40
of the ambient or background noise and the WT noise.
Night 40 35 30
The ambient noise can mask the turbine noise complete-
Denmark (Leq) 40 45
ly if the turbines are located in an industrial or urban area.
UK

Trees may also mask distant WT noise.
High speed (L50) 45
Low Speed (L50) 40

Another important factor is time. WT noise can be present
for hours, days or for longer periods depending on the
wind resource. An excellent wind resource location (e.g. Source: Gipe (1995).
load factors of 40%) can cause the turbines to operate for

more than 3,000 hours a year. The frequency of the noise A noise assessment aims to determine how the turbines
will also affect sound pressure levels. affect the existing ambient background noise and also
what is an acceptable level of noise from the turbines
Regulatory standards for determining acceptable sound themselves. The assessment should be able to demon-
pressure levels take this time component into account. strate compliance with national noise regulations.
The standards are as follows (Renewable Energy Research
Laboratory, 2002):
181
• L10, L50, L90: The A-weighted sound levels that are
exceeded 10%, 50%, 90% of the time. For example 45
5.1.3 LAND USE 5.1.4 IMPACT ON BIRDS
Land use refers to any alteration of current and future The main impacts of WTs on birds are deaths caused by
uses that can be affected by the installation of WTs. the birds colliding with power lines and blades, and dis-
turbance to migration routes. The main causes are listed
The wind project developer must contact regional, nation- as follows (Manwell, 2002):
al and local agencies to check for any land use restric- • Death or injury caused by rotating blades.
tions in order to seek permission for the development to • Electrocution from transmission lines.
go ahead. Equally important is the need to assess the • Alteration of migration habits.
views of the local population so that any concerns they • Reduction of available habitat.
may have on land use are investigated and resolved. • Disturbance to breeding, nesting and foraging.
Given the diffuse characteristics of wind energy, it is nec- More sensitive areas are those on migration paths and with
essary to locate several turbines together to achieve the a high number of birds present. The impacts are variable
same capacity as conventional fossil fuel power plants. depending on the species, season and site-specificity
Thus, wind energy installations require larger areas than (BirdLife, 2002).
conventional power plants. This is due to aspects such as
turbine spacing, topography, location of power lines and According to the latest report by BirdLife (2003), the main
other associated facilities, in conjunction with other issues potential hazards to birds from WT sites are: disturbance
such as protected areas, access roads, land use objec- leading to displacement or exclusion, including barriers to
tives of the community and incompatibility in land-use. movement; collision mortality; and loss of, or damage to,
habitats. These aspects are further explained as follows.
However, only 1% to 3% of the total area is occupied by
the turbine (tower base area, the foundations are mostly Disturbance
underground). So up to 99% of the land on which the tur-
bines are sited will still be available for other uses. In The BirdLife report cites several studies showing that with-
Europe, most wind energy sites are located in remote, in 600 m from WTs bird numbers are reduced. However,
rural areas where livestock grazing is a common practice the report states that: “The scale of such habitat loss,
(see Figure 5.1). together with the extent of availability and quality of other
suitable habitats that can accommodate displaced birds,

Figure 5.1: Wind Energy in Rural Areas and the conservation status of those birds, will determine
whether or not there is an adverse impact.” (p.2)
Disturbance to bird populations may also result from

increased human activities around the site, for mainte-
nance purposes, etc., as well as WT noise and movement
(BirdLife, 2003). In intensively farmed areas, however, the
presence of WTs may have little effect on wild and farm-
land bird populations which will already be depleted due
to intensive agricultural practices.
182 Source: University of Flensburg (Lehbek in Gelting. Schleswig-Holstein, Germany).

Collision Risk and Mortality that 692 turbines located in 18 wind farms do not put any
species at risk from death by collision. 88 deaths of medi-
With respect to collision mortality, the two most critical um and large birds were detected, which represents an
examples are the Altamont Pass. California, USA and La annual mortality rate of 0.13 dead birds per turbine. In
Tarifa in Spain, both of which raised concerns over their other words, it takes more than seven years for one tur-
impact on birds (National Wind Coordinating Committee, bine to kill one bird.
2002). In the case of Altamont Pass, the issue arose in
the late 1980s when the California Energy Commission In a study for the Finnish Ministry of Environment,
recorded 99 dead birds in a four-year period from 1984 to Koistinen (2002) showed that 10 birds were killed by 60
1988 which had been killed by the WT, transmission lines WTs in a one-year period.
or other unknown cause (Gipe, 1995). The Altamont Pass
wind park is characterised by a high density of turbines The likelihood of bird collisions is determined by wind
4
and the coexistence of turbines of diverse types and size. speed, nature and height of flight, species, age of bird and
At Altamont Pass, the main losses were of raptors (birds stage in its breeding cycle. Most studies have been carried
of prey such as hawks and eagles) while at La Tarifa soar- out on smaller turbines (BirdLife, 2002); newer, larger tur-
ing birds (storks and vultures) were affected. Both areas bines may have different effects. Low bird fatality rates do
VOLUME
have high concentrations of birds (BirdLife, 2003). These not mean that efforts to reduce the impact of WTs on bird
wind parks are examples of how poor siting and out-moded populations are unnecessary; even a low collision rate in a
WTs and tower technology can adversely impact bird pop- sensitive area may be significant for some bird species.
ulations (Sagrillo, 2003). Subsequent experiences in
Germany and Denmark show that such effects can largely Habitat Loss or Damage
be avoided by responsible planning practice.
Loss or damage to habitats is caused by turbine bases,
In 2001, Western EcoSystems Technology Inc. was com- substations, access roads and transmission line corri-
missioned by the National Wind Coordinating Committee dors. This is not believed to be a major concern to birds
(NWCC) to study avian collisions with WTs and other struc- outside sensitive areas, such as designated sites of
tures. The study aimed “to provide a detailed summary of national and international importance (BirdLife, 2003).
the mortality data collected at windplants and put avian
collision mortality associated with windpower develop- Recommendations

ment into perspective with other significant sources of
avian collision mortality across the United States”. Proper siting of turbines is important if adverse impacts
on birds are to be avoided. The following criteria have
The study estimated that in 2001 in the US, 33,000 birds been proposed (Manwell, 2002):
were killed by the 15,000 turbines in operation, with the • Avoid migration corridors.
majority of these fatalities projected to occur in California • Avoid siting in specific microhabitats.
where approximately 11,500 operational turbines exist. • Use appropriate tower design (tubular towers or
Most of the California turbines are older and smaller lattice towers.
machines, with a capacity ranging between 100 to 250- • Route electrical lines underground.
kW (Western EcoSystems Technology Inc., 2001). The
results indicate that each turbine in the US accounts for These criteria could be incorporated into national or
2.19 avian deaths a year for all species combined and regional planning strategies. An EIA systematically exam-
0.033 raptor fatalities per turbine per year. ines the possible environmental consequences of imple-
menting projects, programmes and policies (United
183
In Spain, a study carried out in the state of Navarra (EHN, Nations, 2002). EU Directive 85/337/EEC requires an
2003) on the impact of wind parks on bird life showed assessment of the environmental effects of those public
and private projects which are likely to have significant 5.1.5 IMPACTS OF CONSTRUCTION ON
effects on the environment. The Directive was amended in TERRESTRIAL ECOSYSTEMS
March 1997 by Directive 97/11/EC which included, in its
Annex II, installations for the harnessing of wind power for These impacts consist of long-term loss of land from tur-
energy production (wind farms). For wind energy develop- bine installation and their associated electrical connec-
ments every EU member state shall assess the project’s tions, buildings and access tracks. It has to be noted,
environmental impacts on a case by case basis. however, that only the access roads and a very small area
around the tower of a WT are lost. Danish and German
Impacts on Birds in Context research shows that agriculture may continue in rural wind
parks, which are often used for grazing cattle.
The impact of wind energy on birds must be placed in
context (Youth, 2003). Virtually all threats to birds are 5.1.6 ELECTROMAGNETIC INTERFERENCE (EMI)
human-related (99%), with habitat loss as a result of
industrialisation, over-exploitation of natural resources, WTs or generation equipment can interfere with commu-
over-population (human), etc., being the biggest threat. nication systems that use electromagnetic waves (see
Other threats include hunting, the pet trade, unsustain- Figure 5.2). This is caused mainly by the turbine blades,
able fishing practices, oil spills, and oil and natural gas which sometimes scatter the signals as they rotate. The
exploration, extraction and transportation. Chemical and tower may also reflect signals, so interfering with the orig-
pollution threats such as pesticides, lead from spent inal signal arriving at the receiver (Manwell, 2002).
hunters’ shot or sinkers left by anglers are also signifi-
cant. Structures such as skyscrapers, communication
towers and transmission lines kill migrating birds, while Figure 5.2: Electromagnetic Interference
climate change poses a new threat to bird habitats.
With respect to wind park developments, location is a crit-

ical factor and there is a need for further research on the
new, larger, generation of turbines.
In the US, the Western EcoSystems Technology Inc. (1) Reflected signal
(2001) study found a range of between 100 million to 1

billion bird fatalities due to collisions with artificial struc- Electromagnetic
interference
tures such as vehicles, buildings and windows, power
lines and communication towers, in comparison to (2) Distorted main signal
33,000 fatalities attributed to WTs. The study reports
that, “windplant-related avian collision fatalities probably
represent from 0.01% to 0.02% (e.g. one out of every
5,000 to 10,000 avian fatalities) of the annual avian col- Receiver
lision fatalities in the United States, while some may per- Sender
ceive this level of mortality as small, all efforts to reduce

avian mortality are important”.
In Finland, Koistinen (2002) reports 10 bird fatalities from

turbines, and 820,000 birds killed annually from colliding
184
with other structures such as buildings, electricity pylons
and lines, telephone and television masts, lighthouses
Source: Manwell et al. (2002).
and floodlights.
EMI mainly affects television reception, aircraft navigation Figure 5.3: Shadow Calculation
and landing systems, as well as microwave links.
Interference with television reception is the most common
effect but it can be easily and cheaply corrected. Other
mentioned impacts are unlikely to happen unless the tur-
bines are placed in close proximity to the transmitter or
receiver. EMI effects on FM radio, cellular phones and
satellite services are very unlikely to occur.
EMI is a site-specific issue. It is recommended that an on-

site assessment is performed to identify any effects on
radio services in the area as well as the interference
4
zones.
5.1.7 FLICKERING
VOLUME
The rotating turbine blades cast moving shadows which
cause a flickering that can affect residents living nearby.
Similarly, gloss surface blades flash when they rotate. This
effect has been subject to analysis especially in northern
Europe where this effect is considered, although it is not 5.1.8 CONSUMPTION OF ENERGY (ENERGY
seen as an issue in the US (Gipe, 1995). BALANCE)
Figure 5.3 shows an example of the shadow flicker In a conventional coal fired fossil power plant, the fuel
effect. The figure has been constructed for Denmark. cycle consists of exploring, mining, processing and trans-
The results would vary for different countries due to dif- porting coal, as well as manufacturing and installing the
ferences in cloud cover and latitude. There are two hous- equipment, the power plant operation and the disposal of
es in the picture marked as A and B which are respec- waste. In the case of a renewable source like wind, the
tively six and seven hub heights away from the turbine in fuel cycle includes only the activities required to produce,

the centre. The diagram shows that house A will experi- install, maintain and decommission the turbine and its
ence a shadow from the turbine for five hours per year. ancillary systems. The activities in the fuel cycle require
House B will experience a shadow for up to about 12 an input of energy to make possible the production of
hours per year. Seasonal variation is also included in the energy from a wind turbine.
calculation but is difficult to show without undue com-
plication (European Commission, 1999). How long does a WT take to recover the energy spent in
its fuel cycle, and how much energy does it produce?
In Germany, a court has ruled that the maximum allow- According to a study by the Danish Wind Industry
able shadow flicker a year is 30 hours (Danish Wind Association (1997), modern WTs recover all energy inputs
Industry Association, 2003). Programmes exist that in three to four months (see Table 5.5) and will save
automatically shut the turbine down when conditions between 63-78 times the energy input required to operate
make flickering likely. a coal fired plant over a 20 year period.
The study estimated the energy requirements of a typi-

185
cal Danish 600 kW WT during its 20-year lifetime
(see Table 5.4).
Table 5.4: Energy Use During Life Cycle, 600 kW WT It should be noted that these are conservative estimates
since the primary energy consumption does not include
Gross energy use total (TJ)
coal fired plant construction and operation or indirect
Manufacture of turbine 1.9 energy use during the coal firing process. Furthermore,
Installation of turbine 0.495 the comparison assumes a thermal efficiency of 45%
O&M (20 years) 0.774 which is well above the average figure for coal fired plant
Total excluding scrapping 3.169 in the EU. In general, therefore, the WT energy recovery
Scrapping, energy use 0.522 period will be even shorter (European Commission, 1999)
Scrapping, recovered energy -0.733
Total including scrapping 2.958 Delivery of the WT to a remote site makes very little dif-
Total incl. scrapping MWh 821
ference to the above figures. For example, even if a 65
tonne turbine had to be shipped 10,000 nautical miles,
this would only increase its net energy use by 1.5%.
Source: Danish Wind Industry Association (1997).
R&D programmes continue to develop shorter WT energy

The study then estimated how much energy usage (pri- recovery periods.
mary energy consumption) is required in a coal fired plant
to produce the same amount of electricity as the turbine
producers in one year. 5.2 Environmental Impacts of Offshore
Wind Energy
Table 5.5 shows that to produce the same quantity of
electricity per year, the WT requires far less energy input This section introduces the environmental impacts of off-
(821 MWh) than the coal-fired power plant. (3,202 shore wind energy developments. It is based on a compre-
MWh/2,598 MWh). hensive study funded by the European Commission aimed
Table 5.5: Energy Recovery Time for a Wind Turbine
(A) (B) (C) (D) (E) (F)

Wind Turbine Wind Turbine Coal-Fired Plant Wind Turbine Wind Turbine Wind Turbine Wind Turbine
Site Roughness Electricity Primary Energy Energy Use Energy Recovery Energy Recovery Energy Saving
Class Production Consumption* Period (year) Period (months) Period**
(MWh/Year) (MWh/Year)* (MWh) (MWh) E = D x 12 F = (B x 20)/C
Class 1 1,393 3,202 821 0.26 3.1 78
Class 2 1,130 2,598 821 0.32 3.8 63
* Input of energy required in a coal fired plant to produce 1,393 and 1,130 MWh/year of electricity considering only coal mining, transportation, energy content of coal and
a plant efficiency of 45%.
** Wind energy saving over a 20 year period operation.
186
at gathering and distributing knowledge on all aspects of off- Table 5.7: Potential Negative Environmental Impacts
shore wind energy, including: offshore technology; electrical
integration; economics; environmental impacts; and political Birds:
aspects (Garrad Hassan and Partners et al., 2001). - Collisions with turbines.
• turbines acting as obstacles for migrating birds.
Questionnaires were sent to developers, utilities compa- - Disturbance to feeding/breeding areas due to:
nies, consultants, research institutes and universities in • noise from turbines in operation and vessels during construction,
different European countries in order to identify the rele- maintenance and dismantling;
vant issues and collect information on factors such as • movements of blades or serious disruption to the food chain, e.g.
public acceptance, environmental impacts, conflicts of due to new sediment structure or “unnatural” reef effect; and
interest and the political aspects of offshore wind devel- accidents.
opment. With respect to environmental impacts, the sur- Mammals:
4
vey found that birds, visual effects and impacts on recre- - Loss of habitat due to:
ation are the top European concerns. • noise;
• movement of blades;
Table 5.6: Ranking Environmental Impacts • food chain changes;
VOLUME
• electromagnetic fields and vibrations (affecting the animals’ sonar
Impacts Indicator* system); and
Birds 1.5 • accidents.
Visual effect 1.5 Fish:
Recreational areas 1.8 - Impacts on fish and fish larvae from sedimentation/turbidity, under-
Noise 2 water noise, vibrations and electromagnetic fields.
Hydrography 2.1 - Effects from unnatural reefs.
Fish 2.2 - Accidents.
Marine biology 2.3 Benthic fauna and flora:
Sea mammals 2.4 - Changes in sediment structure.
Sea currents 2.4 - Direct loss from foundation and cable footprints.
Marine archeology 2.4 - Impact from foundations/hard substrates and electromagnetic
Seabed 2.5 fields.
Water quality 2.5

- Disturbance/destruction of the seabed due to accidents with
Raw Materials 2.6 ships/aircrafts.
Coastline:
* 1 = high importance - Impact on coastline due to current/sediment changes caused by

2 = medium importance
cables.
3 = low importance
Source: Garrad Hassan and Partners et al. (2001). - Impact on coastline due to accidents.
Visual impact:
Table 5.6 illustrates the average ranking of environmental - Intrusive artificial obstacles in an otherwise pristine landscape.
impacts. The scale used is from 1 to 3, where 1 corre- Noise impact:
sponds to an issue of high importance and 3 to one of low - Increased blade tip speed and the ability of sound to propagate
importance. The potential environmental impacts listed in more efficiently on sea surface may lead to noise impacts.
Table 5.7 are the expected impacts identified by the study - Impact on birds, sea mammals and fish from underwater noise.
according to current knowledge. Further research is need-

ed to improve understanding of these impacts, and iden-
187
tify others.
Source: European Commission (2002a).
In comparison with onshore wind energy development, the 5.3 Factors Affecting Public Acceptance
identification and understanding of offshore wind devel- of Wind Energy
opment impacts and their respective mitigation measures
are still in their infancy. Easy assessment of potential The environmental impacts of wind energy are often seen
impacts substantially facilitates development. as amenity issues which are mainly borne by local com-
munities; but are they the only factors that need to be
Table 5.8 lists some general recommendations for mitigating considered?
the expected impacts of large-scale offshore wind energy
developments (Garrad Hassan and Partners et al., 2001). Society as a whole has a general understanding and
awareness about the importance of environmentally
Table 5.8 General Recommendations for Offshore Wind friendly technologies, not only as a means of generating
Energy Developments cleaner electricity but also to conserve natural resources
and minimise waste. This recognition of renewable energy
Fish, birds and other groups. sources and other issues such as climate change, deple-
- Identify and avoid sensitive areas.
tion of the ozone layer and the sustainable use of energy
- Avoid site works during sensitive time periods.
is often the result of government information programmes
Birds:
and campaigns, energy saving initiatives run by utility
- Design to accommodate migratory flight paths.
companies, media reports, etc.
Sea mammals:
- Minimise noise levels during construction, operation and dismantling.
Fish:
Where such understanding and awareness is absent, get-
- Minimise effects of structures and cabling on stocks. ting the message across about environmentally friendly
Seabeds: technologies, such as wind energy, is much more difficult.
- Minimise sedimentation and turbidity. An informed society, on the other hand, will drive demand
Hydrographic, currents and water quality: for environmental technology.
- Use appropriate foundation design.
- Avoid use of pollutants when protecting the foundation, tower and There is no guarantee that wind energy projects will be
turbine from the marine environment. successfully implemented. The reasons lie in the distance
Visual:
between the costs (impacts borne by local communities)
- Early assessment to take account of distance from shore, marking
and benefits (for general society). This is the so-called
lights and nature of viewpoints.
NIMBY syndrome (“not in my back yard”), which is some-

- Well-balanced marking lights to take account of safety issues and
times a response to unknown technology and impacts.
visual impacts.
The support of local communities is essential if a wind
Noise:
- Ongoing public relations work to counter poor publicity.
project is to go ahead. Support is more likely to be forth-
- Maintain good standards of noise emissions despite increases in coming where there has been a clear assessment of the
turbine size and tip speed. impacts and the mitigation measures have been properly
Social conflicts: explained. Other factors are also relevant, however.
- Promotion of openness and local involvement.
Risk management: Local community participation is an essential element in
- Develop risk management methods and emergency procedures in project development in order to secure public acceptance.
order to reduce risks of ship collision and minimise consequences of Such participation has the following advantages:
collisions.
• Information is shared with the community and
feedback becomes part of the planning process.
Source: European Commission (2002b) • Decision-making and control stays local.
188
• The permitting process is facilitated.
Local participation does not have to be limited merely to
passing information between the parties but can also
include economic involvement through:
• share ownership;
• cooperative association ownership;
• electricity bill discounts; or
• tax rebates.
Public involvement and investment has been a decisive

factor in the successful expansion of wind energy in
Denmark and Germany. The next chapter provides a sum-
mary of research aimed at exploring the public acceptance
4
of wind energy in Europe, which builds on some of the
ideas introduced here.
VOLUME
189
6 PUBLIC ACCEPTANCE IN THE EU
Some EU countries have carried out surveys on public Table 6.1: Perception on Electricity Energy Sources
acceptance of wind energy. Despite differences in
methodology and focus, these surveys give an indication More than half of the electricity used in the EU comes from coal
of the degree of acceptance of wind energy. This section Yes, it is the case 21%
summarises results from different studies. The findings No, it is not the case 49%
of a pan-EU public opinion survey on energy-related issues Do not know 31%
are reported, followed by research from a number of More than one quarter of electricity produced in EU comes
EU member states on public acceptance of wind energy. from nuclear power stations
Yes, it is the case 55%
No, it is not the case 16%
6.1 Attitudes of EU Citizens to Energy Do not know 29%
and Energy Technology Issues More than a quarter of the electricity produced in EU coming
from renewable energy sources, such as hydroelectric energy,
The pan-EU survey “Energy, Issues, Options and wind or solar power
Technologies” was commissioned by the European Yes, it is the case 30%
Directorate for Research. The aim of the survey was to No, it is not the case 43%
gather information on the public view of energy-related Do not know 27%
issues, including scientific and technological aspects, and

prospects for the future. Over 16,000 people were inter-
viewed across the EU-15 during February and April in 2002. When asked about energy sources for the production of
The survey did not focus on wind energy, but it does reveal electricity (see table 6.1), there is an inaccurate perception
general perceptions on issues such as climate change and regarding coal usage for electricity production in the EU.
renewable energy technologies, including wind energy. 49% of the respondents do not think that more than half of
the electricity used in the EU comes from coal and 31% do
The study analyses the perceptions of Europeans about not know. However 55% are correct in that more than one
energy sources. In general, the responses reflect the current quarter of electricity produced in the EU comes from nuclear
situation for oil, coal and gas, but overestimate the use of power stations; and 43% also rightly believe that it is incor-
both nuclear and renewable energy sources (see Figure 6.1). rect that more than a quarter of the electricity produced in
EU is generated from renewable energy sources.
Figure 6.1: Europeans’ Perception of Energy Sources

One of the survey’s main findings is that the public sees cli-
To what extent is each of the following used to produce mate change as a serious issue (88% of respondents).
energy in your country? Fossil fuels are recognised as one of the main causes of cli-
mate change (75%), along with transport emissions (74%).
With regard to energy dependency, 37% of respondents

agree that this is an urgent issue and around a half think that
more energy sources should be developed combined with
greater encouragment for energy efficiency. A quarter want to
see a reduction in imports of fossil fuels and uranium.
When asked about the future, environmental protection

and low prices are the top priorities (72% and 62% respec-
190
tively); 30% believe that ensuring uninterrupted energy
supply should be a priority.
Europeans would like to know more about: how to save sources will be renewables like solar, wind and biomass,
energy at home (53%); the use of renewable energy followed by hydroelectric power (24%) and natural gas
sources at home (42%); alternatives to petrol and diesel (21%). Moreover, 27% consider that renewables will
in vehicles (39%); nuclear safety and radioactive waste provide the greatest amount of useful energy and 67%
(36%); new energy options such as fuel cells (27%); EU think that renewable energy sources are the best environ-
activities in energy-related research and development mental option. The report concludes that “overall, the per-
(23%); and how to save energy at work (13%). The study ceptions Europeans have of energy options in 20 and 50
concludes that “energy, and in particular aspects of ener- years from now is clearly influenced by their own instinctive
gy affecting them personally, is thus a subject on which preferences for renewable energy sources”. EU citizens
Europeans appear to want to be better informed”. expect that energy research will bring significant environ-
mental benefits, more diverse energy sources (69%) and
When asked about what will happen in 2050, 40% of cleaner transport (51%). The following Figures illustrate
4
respondents predict that the least expensive energy these results.
Figure 6.2: Energy Resources Perception in 2050 Figure 6.3: Energy Resources Perception in 2050
– Least Expensive Source – Source with Greatest Amount of Useful Energy
VOLUME
In 2050, which energy resource In 2050, which energy resource will provide
will be least expensive? the greatest amount of useful energy?
191
Figure 6.4: Energy Resources Perception in 2050 Figure 6.5: Energy Resources Perception in 2050
– Best Source for the Environment – More Research in the EU
In 2050, which energy resource will be the best In which areas should there be more energy-related
for the environment? research in the EU?
Finally, the survey found that attitudes vary according to The development of wind parks has an important environ-
country. For example, new energy sources and clean mental component. Environmental impact assessments
transport were chosen as priorities more often in Sweden, maximize the use of existing roads; and allocate existing
the Netherlands and Denmark. and new infrastructure, and restoration of areas impacted
during construction and installation. During the first five
years of wind park operation, potential impacts – espe-

6.2 Public Acceptance in Spain cially with regards to birds and other fauna - were evalu-
ated. Great attention was paid to the integration of wind
Although there has been no national assessment of pub- parks with the existing architecture, surroundings and the
lic acceptance of wind energy in Spain, regional informa- landscape. For example, substations were designed with
tion is available. The APPA (Spanish Renewable Energy the same facades as existing infrastructure.
Association) has provided valuable information on one of
the most important developers, EHN, which is responsible
for 30% of the wind capacity installed in Spain (36 wind
parks as of December 2001). On behalf of the developer,
CIES, a member of the Spanish Association of Opinion
and Market Studies, carried out a survey on public accept-
ance of EHN’s wind farms in different regions, with par-
ticular emphasis on the regions of Navarra (see Figure
192
6.6) and Castilla – La Mancha.
Figure 6.6: Substation, Ibargoiti Wind Park in Navarra (22 MW) Guerinda: 76% see the wind park as beneficial and 4%
consider it damaging. With regard to effects in the land-
scape, 56% say it makes no difference while 42% think it
does have an effect.
Alaiz-Izco: 81% give positive support to wind parks where-

as 6% are negative. With regard to effects on the land-
scape, 45% say it makes no difference, 29% think it
spoils the landscape, and 19% think it improves it.
In Albacete province in the state of Castilla – La Mancha

where 600 WTs are installed, a public acceptance assess-
Source: EHN (2001).
4
ment in 2001 found that 79% were positive about the WTs
and 1% were negative.
The development of wind parks in the states of Navarra and
Castilla has generated 2,000 jobs. In 2001, 400 MW wind In October 2002, a study carried out by CIES on behalf of
VOLUME
energy capacity was installed in the state of Navarra. CIES Energías Eólicas Europeas (EEE) showed that 79% of
carried out 1,369 interviews in Navarra and found a very respondents consider wind energy to be a benefit and 1%
high acceptance for WTs (85%) (see Table 6.2). Even though think it is damaging. The study also found that 62% think
the number of WTs increased dramatically over the period that wind parks make no difference to the landscape while
1995 to 2001, the level of support has remained stable. 23% think they do have an effect.
Table 6.2: Public Acceptance in Navarra The high acceptance of wind energy is due to environ-
mental, energy-related and socio-economic reasons.
Year 1995 1996 1998 2001
Acceptability values higher than 70% were found in all
Number of turbines* 6 40 217 600
areas surveyed, with 88% seeing wind as a clean energy
Positive/very positive 85% 81% 81% 85% source and 48% considering that it creates wealth and
Negative 1% 3% 3% 1% jobs. 69% of those surveyed thought that wind energy was
the best energy source to produce electricity. This com-

* Mostly turbines of 660 kW capacity. pares to 17% who support hydro, 2% thermal power and
Source: EHN (2001).
1% nuclear.
A previous study carried out by CIES for EHN in 1998 on A recent study carried out in Tarragona province (CERES,
public perceptions in wind park areas in Navarra found: 2002) in the region of Cataluña also shows a majority
favouring wind power (83%) over nuclear or fossil fuel
El Perdón: 82% see the wind park as a step forwards technologies. Another interesting finding is the link
whereas 2% think it is a step backwards. With regard to between educational attainment and perceptions of wind
effects on the landscape, 41% say it makes no difference, energy. The higher the level of education, the greater the
32% think it spoils the landscape, and 24% think it reluctance to accept certain aspects of wind energy such
improves it. as visual intrusion. The Centre for Sociological Research,
an autonomous state agency attached to the Office of
Leitza-Beruete: 74% think the wind park is beneficial, 8% the Presidency, carries out regular public opinion sur-
find it acceptable and 7% consider it damaging. With veys. The last survey, in March 2003, showed that 65.4%
193
regard to effects on the landscape, 56% say it makes no backed further research on clean energy sources and
difference while 36% think it does have an impact. 1.2% wanted to see more work on nuclear power.
6.3 Public Acceptance in the United living within a 20 km radius of a wind park. Its main
Kingdom findings are:
• 20% of respondents think their local wind park has a
The UK government aims to generate 20% of the UK’s broadly positive impact on the area while 7% felt that it
electricity from renewable energy sources by 2020, with has a negative impact. The majority are neutral.
the major focus on offshore wind energy. The latest poll • Before the construction of the wind park, 27% of
to measure public support for this target shows that respondents were concerned about landscape
74% support both the 20% goal and increasing the use changes, 19% were concerned about traffic during con-
of wind power. struction and 15% about noise during construction.
During the construction phase and afterwards, these
Aggregating data from 42 surveys carried out between figures fell to 12%, 6% and 4% respectively.
1990 and 2002 shows, on average, that 77% of the pub- • 54% would support a 50% increase in the number of
lic are in favour of wind energy with 9% against (British turbines at their local wind farm, 9% would not.
Wind Energy Association, 2003). • With respect to other technologies, respondents want
to see a decrease in nuclear, coal and oil power. Clean
A summary of research on attitudes to wind power from electricity production technologies are strongly sup-
1990 to 1996 (Marie et al., 1996), concludes that an ported with 69% in favour of wave energy and 82% in
“overwhelming majority of residents in areas with a wind favour of wind energy.
project are pro-wind, both in theory as a renewable energy
source and in practice in their area, with an average of
eight out of 10 supporting their local wind farm”.
A survey of people living within 20 km of four wind farms

in Scotland was carried out in 2000 for the Scottish
Executive (System Three Social Research, 2000). The
sample was divided into three zones of 5 km from the
farm, 5-10 km and 10-20 km. The main results are as fol-
lows:
• 67% of respondents said there was something they
liked about the wind park, this proportion increased to

73% for those living in the 5 km zone.
• With respect to visual impacts, 21% liked the look of the
wind park whereas 10% thought it spoiled the view.
• Regarding future developments, there was a positive
attitude towards wind parks; 14% of respondents
would be concerned if extra turbines were added to the
existing park.
A recent survey conducted for the Scottish Executive by

MORI in 2003 (EWEA, 2003f) shows that people living
close to Scotland’s 10 largest wind parks strongly support
wind energy, with 82% wanting an increase in electricity
generated from wind, and more than 50% supporting an
194
increase in the number of turbines at their local wind park.
The MORI poll (see Table 6.3) covered 1,800 residents
Table 6.3: Results of MORI/Scottish Executive Poll
Question: What effect, if any, would you say the presence of the wind Question: To what extent would you support or oppose increasing the
farm has had in your local area? Would you say it has had… number of turbines at the wind farm by 50%? Would you…
(Base: All)
%
%
A completely positive effect 7
Strongly support 26
A generally positive effect 13
Tend to support 28
Neither positive nor negative effect 51
Neither support nor oppose 25
A generally negative effect 5
Tend to oppose 5
A completely negative effect 2
Strongly oppose 4
Don’t know/ No opinion 23
Don’t know 11
Question: Was the wind farm already here when you moved in, or has
it been built since then?
Wind farm already here

%
10
To what extend would you support or oppose increasing the number
of turbines at the wind farm by 100%? Would you…
(Base: All) 4
VOLUME
%
Built since moved in 77
Strongly support 19
Don’t know 14
Tend to support 23
Question: I would like to know what you anticipated it might be like Neither support nor oppose 24
having a local wind farm (before it was built/ before you moved here), Tend to oppose 11
and then I will ask you about what it has actually been like.
Strongly oppose 10
Don’t know 14
Question: Which of the following problems, if any, did you think having
a wind farm in the area might cause?
Question: I am going to read out some different ways of generating

Question: And which, if any, have actually turned out to be problems
electricity. For each one, I would like you to tell me whether you think
caused by having a wind farm in the area?
Base: All who have lived in the area before the wind farm was built (1,547) the proportion of electricity generated in Scotland should increase,
reduce or stay at about current levels over the next 15 years? First
of all…
Thought might Have been (Base: All)
be problems % problems %
Increase Keep same Reduce Don’t know

Noise from the turbines 12 2
% % % %
The look of the landscape being spoiled 27 12
Coal fired power 9 24 60 6
Interference with TV and radio reception 6 1
Oil fired power 9 32 48 10
Damaging effect on local business 3 1
Nuclear power 7 17 68 9
Damage to plants or animals 12 3
Wind energy 82 11 2 6
Noise or disturbance during construction 15 4
Wave energy 69 11 3 17
Extra traffic during construction 19 6
A reduction in house prises 7 2 Source: EWEA (2003f).
None of these 54 82
195
Two opinion polls have been carried out in Scotland to from visiting the area in the future” (MORI Scotland,
determine the effect, if any, of wind farms on tourism 2002). The second concluded that the visual impacts
to the region. The first poll, of visitors to Argyll and of wind farms are a concern, especially in protected
Bute, revealed that wind parks “are not seen as having areas such as Areas of Outstanding Natural Beauty
a detrimental effect… and would not deter tourists (NFO, 2002).
Table 6.4: Results of MORI Scottish Tourist and Visit Scotland Poll
Results of MORI Scottish Tourists Poll: Results of Visit Scotland Poll:
Question: What effect, if any, would you say the presence of wind
Views of development of wind farms
farms has had on your impression of Argyll as a place to visit?
as a means of generating power (%)
Good idea - ecologically friendly 39

A completely negative effect 1%
Good idea generally 17
A generally negative effect 7%
Good idea - save digging fossil fuels 11
Equally positive and negative effect 43%
Good idea - need different sources of power 5
A generally positive effect 28%
In favor of them 5
A completely positive effect 15%
Necessary evil - better alternative to nuclear 4
Don’t know 6%
Necessity but an eyesore 4
Against - can’t generate enough power 2
Good idea - natural resource 1

Question: Has the presence of wind farms in Argyll made you any more
likely to visit the area in future, made it less likely, or has it made no If not too many - good idea 1
difference? Other positive 1
Other negative 7
Less likely 2% Don’t know 5
No difference 91%
More likely 4% Impact on further holidays in the Scottish countryside if

Don’t know 3% the number of wind farms was to increase (%)
Would make no difference 63
Steer clear of the area 15

Less likely to come back 10
Depends on the area 6
Minimal impact 2
Other 2
Don’t know/not stated 5

Source: EWEA (2003f).
196
6.4 Public Acceptance in Denmark 6.5 Public Acceptance in Germany
In Denmark, public opinion of wind energy over the last 10 The northern state of Schleswig-Holstein in Germany had,
years has been positive (Danmarks Vindmølleforening, as of December 2002, 1,800 MW of installed wind capac-
2002). ity, with a share of nearly 30% of the state’s energy
consumption (DEWI, 2003). An analysis of wind energy in
A survey (see Figure 6.7) carried out in 2002 shows that Schleswig-Holstein was prepared for the state’s Energy
59% would buy electricity from a renewable source, while Ministry in 2002. Some issues of relevance to public
24% would not. Results from a study in 2001 illustrate acceptance are summarised here (Eggersglüß, 2002).
that 86% of the population support wind energy with 68%
wanting Denmark to install more WTs and 18% thinking Germany’s approach to wind energy has changed dramat-
that existing capacity is sufficient. ically over the years. Initially, individuals who were inter-
4
ested in using wind energy, such as farmers, could install
Figure 6.7: Energy Sources Preferences in Denmark a WT on their own land. Then, the growing interest of non-
local investors made it possible to develop wind parks on
designated areas. In the meantime, many “citizen’s wind
VOLUME
parks” have emerged funded by companies who offer
shares to local small-scale investors. These have proved
very popular.
In general, the siting of a wind park is accepted by most

people in a particular area when the following principles
are followed:
• Sufficient distance from residential areas.
• Quiet turbines are chosen.
• The population is kept properly informed.
Source: Danmarks Vindmølleforening (2002).
• There is some sort of financial benefit for the local
community.
A 1997 study carried out in the municipality of Sydthy, • The developer has its head-quarters and

where 98% of electricity supplied to the 12,000 inhabi- administration situated in the area.
tants is generated by wind, found that people with a high . Land owners’ views are sought when choosing a site.
degree of knowledge about energy generation and renew-
able energy in particular tend to be more positive about Although wind energy is seen as a clean way of producing
wind power. In addition, 58% of householders in Sydthy electricity and preserving natural resources, concerns
have shares in their local wind park (Damborg, 1998). have been raised about changes to the landscape, noise,
flickering and effects on birds. Other worries include high-
In Denmark, 150,000 families are involved in wind energy er electricity prices and financial rewards for a few land
projects due to the possibility of receiving financial bene- owners and WT operators.
fits and/or a positive stance regarding the environmental
benefits of wind energy (EMU and Hammarlund A study to assess the effects of onshore and offshore wind
Consulting, 2003). parks on tourism was undertaken by the Schleswig-Holstein
tourist board (Günther, 2002). This concluded that the wind
industry does not affect tourism in the region. Visitors are
197
aware of the increasing number of turbines in the
landscape, but they do not influence visitors’ behaviour.
Research carried out by the EMNID Institute for the 6.6 Public Acceptance in Sweden
German science magazine P.M. in 2003 found that 66%
of Germans are in favour of further construction of wind A recent survey in Sweden (see Figure 6.8) shows that
farms. The institute also conducted a survey in 2002 (see wind power would be the preferred electricity production
Table 6.5) showing that 88% supported the construction option, with 64% support (SIFO, 2002).
of more wind parks in Germany, with 86% agreeing that
the share of wind power in the energy mix should increase Figure 6.8: Energy Sources Preferences Sweden
(EWEA, 2003f).
Energy sources preferences
Table 6.5: Results of EMNID Poll, 2002
Question: Which statements would convince you to support the erection

of further wind turbines?
Wind turbines produce green electricity 89.2%
Wind power is a new technology that creates thousands of jobs 62.1%
Wind turbines are easy to control and secure 66.3%
Question: Gas, oil and coal are limited resources, whilst the use of wind
power constitutes an alternative. Do you believe that the share of wind
power in the energy mix is already sufficient? Source: SIFO (2002).
Yes, it is sufficient 9.5%
No, it should be increased 86% Wind energy was the second choice for 25% of those
Don’t know 4.5% respondents who placed nuclear power as a first option.
Question: Wind farms must fulfil certain criteria, such as there have to
The SIFO survey also found that 73% of respondents
be minimum wind speeds, there has to be enough space between single
wind turbines, additional standards in residential and nature conserva- thought that Sweden should increase its proportion of
tion areas have to be met. Do you support the construction of further electricity generated from renewable sources.
wind farms when those criteria are fulfilled?
Yes 88.3% With regard to tourism, a public acceptance study from

No 9.5% 1988 to 2002 found that tourists have a negative attitude

Don’t know 2.2% to onshore wind farms, especially in rural landscapes, but
are more positive about offshore developments (EMU and
Question: If we assume that offshore wind farms also have to meet
Hammarlund Consulting, 2003). In some areas, opposi-
strict standards, would you support their construction far away from the
coast? tion to onshore wind farms was mainly from tourists and
non-permanent residents who place a greater value on
Yes 88.3%
landscape amenity than do permanent residents.
No 9.5%
Don’t know 2.2%

6.7 Public Acceptance in Austria
In July 2003, a national poll was conducted to determine

the Austrian public’s knowledge about renewable energy
sources, their acceptance of renewables and their future
198
energy preferences. The Gallup Institute conducted the Figure 6.9: Future Preferred Energy Sources
poll of 1,500 people for the Austrian Utilities Association
Verband der Elektrizitätsunternehmen Österreichs (VEÖ).
When asked about renewable energy, 45% of respondents
knew what it was. Solar, hydro and wind were the most
recognised renewable energy sources at 39%, 33% and
30% respectively.
Hydro, solar, wind and micro hydro are the most popular
energy sources when scored on a scale of 1 to 5 where 1
is very popular and 5 is very unpopular (see Table 6.6).
With respect to future energy sources, 68% of respon-
4
dents prefer solar, followed by hydro, wind and biomass.
Fossil fuel has little support and nuclear none, as can be
Source: VEÖ (2003).
seen in Figure 6.9.
VOLUME
Table 6.6: Popularity Energy Sources 6.8 Public Acceptance in Belgium
Source Indicator*
A survey of residents living on the Belgian coast, a popu-
Hydro 1.27
lar tourist area, was carried out in 2002 by the West
Solar 1.31
Flemish Economic Study Office. It revealed that 78% of
Wind 1.41
those surveyed have a positive or neutral attitude towards
Micro Hydro 1.44
the construction of a wind farm 6 km offshore. However,
Geothermal 1.92
the survey also found that 30% of the residents disap-
Biomass 1.97
prove of wind farms in their surroundings. Table 6.7 sum-
Natural Gas 2.88
marises the attitudes of different groups of residents and
Oil 3.21
tourists towards offshore wind parks in their immediate
Coal 3.27
surroundings.
Nuclear 4.53

* 1 = very popular and 5 = very unpopular Table 6.7: Public Perception of Near Shore Wind Farms at
Source: VEÖ (2003).
6 km from the Shore
Very to Neutral to
Group Moderately Very Positive
Negative
Residents 31.3% 66.5%
Second residence 10.2% 88.8%
Frequent tourists 18.7% 81.3%
Occasional tourists 19.5% 80.5%
Hotel, restaurant, pub with view of sea 6.8% 89.3%
Other 15.3% 84.7%
Total 20.7% 78.3%
199

6.9 Conclusions
The countries mentioned here account for 88.8% of the

total wind energy capacity in Europe (22,558 MW in
2002). Germany, Spain and Denmark accounted for
84.3% of that capacity in 2002 (EurObserv’ER, 2003).
The attitudes of EU citizens to renewable energies, and

their awareness of climate change impacts, indicates that
the environmental benefits of renewable energy sources,
including wind, are understood. Fossil fuel and nuclear
energy sources have less support, as shown by the data
from Denmark, Sweden, Austria and Spain.
The surveys cited in this report point to very positive sup-

port for wind energy, with acceptance depending on per-
ceptions of the technology and the way in which develop-
ers deal with local communities.
In Denmark and Germany, where many local citizens are

financially involved and decisions are taken at the local
level, there is a high public acceptance of wind energy.
Efforts to minimise impacts and integrate wind parks into
the landscape in an aesthetic way, combined with local
participation, have yielded good results in Spain.
200
VOLUME 5
MARKET DEVELOPMENT
HOOFDTITEL
Acknowledgments
This volume was compiled by the European Wind Energy

Association with the assistance of the project partners.
EWEA would particualy like to thank the National WInd
Associations for their valuable imputs.
WIND ENERGY - THE FACTS - MARKET DEVELOPMENT
202
VOLUME 5 - MARKET DEVELOPMENT: TABLE OF CONTENTS
CHAPTER 1 MARKET INCENTIVES 205
1.2 Environmental Taxes 208
1.3 Payment Mechanisms 209

1.3.1 Voluntary Systems and Green Marketing 210
1.3.2 Fixed Price Systems 213
1.3.3 Investment Subsidies 213
1.3.4 Fixed Feed-in Tariffs 214
1.3.5 Fixed Premium Systems 215
1.3.6 Tax Credits 216
1.4 Fixed Quantity Systems 216

1.4.1 Tendering Systems 217
1.4.2 Tradable Green Certificate Systems 219
1.5 Renewables in the New Member States 225
1.6 The EU Legal and Political Framework 225

1.6.1 The Electricity Directive 226
1.6.2 The Renewables Directive
1.7 Concluding Remarks

226
228
5
VOLUME
CHAPTER 2 FUTURE MARKETS 230
2.2 Conventional Scenario 230

2.2.1 Conventional Scenario Results 231

2.3 Other Forecasts 232
2.4 Wind Force 12 - the Advanced Scenario 232

2.4.1 Methodology 232
2.4.2 Regional Update of Wind Force 12 Advanced Scenario 233
2.5 Overview of Non European Markets 234

2.5.1 North America 234
2.5.2 Central and South America 234
2.5.3 Australia and New Zealand 235
2.5.4 Asia 236
2.5.5 Africa 237
2.5.6 The East 238
203
CHAPTER 3 NEW TARGETS 239
3.1 EWEA Targets – Onshore and Offshore 239

3.1.1 European Commission – Historical Targets 240
3.1.2 European Commission – New Targets 240
3.2 Increasing Wind Power Targets for Europe 241
3.3 Targets for the EU-15 in 2010 243

3.3.1 How Much Electricity Will This Provide? 244
3.3.2 What Proportion of Total EU Electricity from Wind? 245
3.3.3 What Share Will Wind Have of Total New Capacity Installed? 245
3.4 International Energy Agency Scenarios 247

3.4.1 OECD Alternative Policy Scenario 247
204
1 MARKET INCENTIVES
1.1 Introduction
As mentioned in the section on externalities, the full costs The European Commission’s ExternE project on external
to society of electricity production are not reflected in costs estimates that the cost of producing electricity from
electricity prices. Those costs are paid by taxpayers and coal or oil in the EU would double and the cost of elec-
society as a whole in the form of increased health care tricity production from gas would increase by 30% if exter-
and environmental costs such as climate change. nal costs, in the form of damage to the environment and
health, were taken into account (European Commission,
Wind energy is becoming increasingly competitive with 1999). Currently, average electricity production costs in
conventional sources. However, it is likely that some form the EU are 0.04 € per kWh. The study further estimates
of incentive will be required for the foreseeable future, at that the external costs amount to 1-2% of EU GDP or
least until environmental costs are fully internalised or between €85 billion and €170 billion, not including the
increased economies of scale and technological develop- cost of climate change.
ment makes wind power fully competitive with conven-
tional sources, such as coal and gas, without the need to Table 1.1 summarises the incentives for wind power and
consider externalities. other renewables available in the EU-15 as of 2003. The
table only includes conditions for new installations. Other
There are currently five main systems to support elec- conditions may apply to existing renewable energy capacity.
5
tricity from renewable energy sources in the EU member
states: investment subsidies, fixed price systems, fixed
premium systems, auctions, and certificates systems.
The idea behind the mechanisms is to offset at least
VOLUME
some of the competitive disadvantage for renewables as
a consequence of electricity markets neglecting the envi-
ronmental cost of production from conventional technolo-
gies. Low electricity prices are of little benefit if they lead
to high costs to society as a whole through higher health
care costs and environmental costs levied on current and

future taxpayers and citizens.
If the environmental costs of power production were reflect-

ed in European power prices, wind power and many other
renewable energy technologies would not need support, as
pointed out in the European Commission’s Green Paper on
Security of Supply (European Commission, 2002a)
The Green Paper states that wind energy can fully com-
pete with combined cycle gas if externalities are taken
into account. Furthermore, both wind energy, biomass,
small hydro, photovoltaics (PV) and geothermal are signif-
icantly cheaper for society than coal if externalities are
included. Coal is almost twice as expensive as wind and
biomass (1998 figures) according to the Green Paper.
205
Table 1.1: Support Mechanisms
RES-E TECHNOLOGIES CONSIDERED
Major Large Hydro Small Hydro ‘New’ RES (Wind On- & Offshore, PV, Solar Thermal Electricity, Biomass, Biogas, Municipal
Strategy Landfill Gas, Sewage Gas, Geothermal) Solid Waste
Austria FITs No Renewable Energy Act 2003. (Ökostromgesetz). FITs guaranteed for 13 years for plants which No
get all permissions between 1st of January 2003 and 31st of December 2004 and, hence,
start operation by the end of 2006. Investment subsidies mainly on regional level.
Belgium TGC + No Federal: The Royal Decree of 10 July 2002 (operational from 1st of July 2003) sets minimum prices for RES-E.
guaranteed Except for offshore wind it will be implemented by the regional authorities: Wallonia: Quota obligation (based on TGCs)
electricity Wallonia: Quota obligation (based on TGCs) on electricity suppliers– increasing from 3% in 2003up to 12% in 2010.
purchase Flanders: Quota obligation (based on TGCs) on electricity suppliers– increasing from 3% (no MSW) in 2004 up to
6% in 2010. Brussels region: No support scheme yet implemented.
Denmark Partial Tax No Act on Payment for Green Electricity (Act 478): Max combined price for wind power No
Exemption + and partial tax exemption of 4.4 c€/kWh. Minimum price of 1.33 c€/kWh. Exemption of CO2 tax
tender (max 1.33 c€/kWh) depends on electricity market price. Plans for offshore wind tenders.
Finland Tax No Tax refund Mix of tax refund and investment subsidies: From January 2003: Tax refund of No
Exemption 0.44 c€/kWh 0.73 c€/kWh for wind and of 0.44 c€/kWh for other RES-E. Investment subsidies
(plant <1MW) up to 40% for wind and up to 30% for other RES-E.
France FITs No FITs for RES-E plant < 12 MW guaranteed for 15 years (20 years PV and hydro). Tenders for plant No
>12 MW. FITs in more detail1: biomass - 4.9 c€/kWh; biogas - 4.6 c€/kWh; geothermal -
7.62 c€/kWh; PV2 - 15.25-30.50 c€/kWh; landfill gas - 4.50-5.72 c€/kWh;
wind3 - 3.05-8.38 c€/kWh; hydro4 - 5.49-6.10 c€/kWh. Investment subsidies for PV, biomass and
biogas (biomass and biogas PBEDL 2000 - 2006).
Germany FITs No German Renewable Energy Act: FITs guaranteed for 20 years5. In more detail, FITs for new No
installations in 2003 are: hydro - 6.65- 7.67 c€/kWh; wind6 - 6-8.9 c€/kWh; biomass - 5,8-10 c€/kWh;
landfill gas, sewage gas and mine biogas - 6.65-7.67 c€/KWh; solar PV and solar thermal electricity -
45.7 c€/kWh; geothermal - 7.16-8.95 c€/kWh.
Greece FITs + No FITs guaranteed for 10 years (at a level of 70-90% of the consumer electricity price)7 No
investment and a mix of other instruments: a) Law 2601/98: Up to 40% investment subsidies combined with
subsidies tax measures; b) CSF III: Up to 50% investment subsidies depending on RES type.
Ireland Tender No Tendering scheme – currently AER VI with technology bands and price caps for small wind (<3 MW), No
large wind (>3 MW), small hydro (<5 MWp), biomass, biomass and biogas. In addition, tax relief
for investments in RES-E.
Italy TGC Quota obligation (based on TGCs) on electricity suppliers: 2% target, increasing annually; TGC issued for all (new)
RES-E (inc. large hydro and MSW) – with rolling redemption8; unclear penalty enforcement and market distortions9.
Investment subsidies for PV (Italian Roof Top programme).
Luxembourg FITs No No FITs10 guaranteed for 10 years (PV 20 years) and investment subsidies for wind, PV, No
biomass and small hydro. FITs for wind, biomass and small hydro - 2.5 c€/kWh;
for PV - 50 c€/kWh11.
Portugal FITs + No FITs (Decree law 339-C/2001 and Decree law 168/99) and about 40% investment subsidies
investment small hydro and wave. FITs in 2003 - wind12 - 4.3c€/kWh - 8.3c€/kWh; wave - 22.5c€/kWh;
subsidies PV13 - 22.4c€/kWh - 41c€/kWh; small hydro - 7.2c€/kWh.
Spain FITs Depending FITs (Royal Decree 2818/1998): RES-E producers have the right to opt for a fixed price or for a 1.7 c€/kWh
on the plant premium tariff15. Both are adjusted annually by the government according to the variation in
size14 the average electricity sale price. In more detail (only premium, valid for plant < 50 MW16):
wind - 2.7 c€/kWh - PV17 - 18-36 c€/kWh - small hydro - 2.9c€/kWh - biomass - 2.5 - 3.3 c€/kWh.
206 Moreover, soft loans and tax incentives (according to “Plan de Fosento de las Energías Renovables”)
and investment subsidies on a regional level.
FIT: Feed-in Tarifs

TGC: Tradable Green Certificates
RES-E TECHNOLOGIES CONSIDERED
Major Large Hydro Small Hydro ‘New’ RES (Wind On- & Offshore, PV, Solar Thermal Electricity, Biomass, Biogas, Municipal
Strategy Landfill Gas, Sewage Gas, Geothermal) Solid Waste
Sweden TGC No Quota obligation (based on TGC) on consumers: Increasing from 7.4% in 2003 up to 16.9% in 2010. No
For Wind Investment subsidies of 15% and additional FITs (“Environmental Bonus”18) in size of
1.9 c€/kWh are available.
Netherlands FITs + tax Mixed strategy: green pricing, tax exemptions and FITs. The tax exemption for green electricity amounts No
exemption 2.9 c€/kWh and FITs range from 2.9 c€/kWh for mixed biomass and waste streams to 6.8 c€/kWh for
wind, PV, tidal, wave and small hydro.
United TGC + No Quota obligation (based on TGCs) for all RES-E: Increasing from 3% in 2003 up to 10.4% by 2010 No
Kingdom investment – penalty set at 3.51 £/kWh. Optional to the TGC-system, eligible RES-E are exempted from the
subsidies Climate Change Levy certified by Levy Exemption Certificates (LECs), which cannot be separately
traded from physical electricity. The current levy rate is 0.43 £/kWh. Investment grants in the frame
of different programmes (e.g. Clear Skies Scheme, DTI’s Offshore Wind Capital Grant Scheme,
the Energy Crops Scheme, Major PV Demonstration Programme, and the Scottish Community
Renewable Initiative).
1
Without efficiency premiums.
2
30.5 €/kWh for Corsica and Overseas Departments.
3
Stepped FIT: 8.38 c€/kWh for the first 5 years of operation and then between 3.05 and 8.38 c€/kWh depending on the quality of site.
4
Producers can choose between four different schemes. The figure shows the flat rate option. Within other schemes tariffs vary over time (peak/base etc.).
5
5
The law includes a dynamic reduction of the FITs (for some RES-E options): For biomass 1% per year, for PV 5% per year, for wind 1.5% per year.
6
Stepped FIT: 8.9 c€/kWh for the first 5 years of operation and then between 6 and 8.9 c€/kWh depending on the quality of site.
7
Depending on location (islands or mainland) and type of producer (independent power producers or utilities)
8
In general only plant put in operation after 1st of April 1999 is allowed to receive TGCs for their produced green electricity. Moreover, this allowance is limited to the first
8 years of operation (rolling redemption).
9
GRTN (Italian Transmission System Operator) influences strongly the certificates market selling its own certificates at a regulated price – namely at a price set by law as
VOLUME
the average of the extra prices paid to acquire electricity from RES-E plant under the former FIT-programme (CIP6).
10
Only for plants up to 3 MW except up to 50 kW for PV systems.
11
For plants commissioned in 2004 the FIT will be in the range of 45 c€/kWh
12
Stepped FIT depending on the quality of the site.
13
Depending on the size: <5kW - 42 c€/kWh or >5kW - 22.4 c€/kWh
14
Hydropower plant with a size between 10 and 50 MW receive a premium depending on the farm size according to the formula: Premium (c€/kWh) = 2.9 * (50-‘plant size
in MW’) / 40. For plants >50MW the premium tariff is set to 0.6 c€/kWh.
15
In case of a premium tariff, RES-E generators earn in addition to the (compared to fixed rate lower) premium tariff the revenues from the selling of their electricity on the
power market.
16
For Small Hydro the plant size is limited to 10 MW.

17
Depending on the plant size: <5kW: 36c€/kWh or >5kW: 18c€/kWh
18
Decreasing gradually down to zero in 2007
Source: EWEA, Rexpansion Project, forthcoming.
207
1.2 Environmental Taxes from the face of the earth just because they do not
appear on the electricity bill or because they are not
Energy taxes that reflect the actual environmental impacts included in electricity producers’ costs of generating
of each technology constitute an effective means to inter- energy. They are being paid for by society as a whole
nalise external costs. Taxes could make the full produc- through taxes on households and companies and through
tion costs of electricity generation transparent, level the environmental degradation such as that caused by
playing field in the future internal electricity market and climate change.
introduce fair competition between renewables and con-
ventional power technologies. This is recognised by the Meaningful environmental taxes are an effective way to
European Commission. In a Communication to the level the playing field in the electricity markets, but are
Council, the European Parliament, the Economic and difficult and time consuming to agree upon at EU level.
Social Committee and the Committee of the Regions in The same is true for removal of state aid to conventional
February 2001, the Commission states: power production technologies. Efforts should be made
to remove harmful subsidies to mature electricity tech-
“Environmental taxes and charges can be an appropri- nologies based on fossil fuel and nuclear, as suggested
ate way of implementing the ‘polluter pays’ principle by by an OECD study on improving the environment through
including the environmental costs in the price of goods reducing subsidies. The higher the subsidies to polluting
and services and by this means internalising external technologies, the higher the costs to society of introduc-
costs. The White Paper emphasised that the environ- ing clean technologies.
mental benefits of renewable energy justify favourable
financing conditions, e.g. through tax exemptions in The OECD argues that “support is seldom justified and
products from RES.” generally deters international trade, and is often given to
ailing industries”. It further argues:
After six years of negotiations, an EU Directive setting
1
minimum tax rates for energy products came into force “This policy [state aid] is often both costly and inef-
on 1 January 2004. As a result of numerous compro-
st
fective in the long run. Technological change and the
mises between the member states, the level of the min- development of new product markets will generally
imum energy tax rates is close to being the lowest com- lead to an even further loss in the competitiveness
mon denominator for the Community and is considerably of the supported industry. As a result, larger
lower than originally proposed by the Commission in amounts of support will be required in order to main-
1997, and by the Parliament in 1999. For electricity, the tain the industry... . In many cases, support is used
Directive introduces minimum taxes of 0.5 €/MWh for to prop up declining industries, merely postponing
business and 1 €/MWh for non-business. Due to the low their certain demise at the expense of taxpayers and
minimum tax levels, the many general exemptions and consumers.”
the lack of mandatory exemptions for renewables, the
effect of the Directive on wind power will be insignificant The OECD also argues “that support may be justified if it
in the short term. However, the importance of reaching a lowers the long-term marginal costs to society as a whole.
final agreement cannot be underestimated, as it empha- This may be the case with support to ‘infant industries’,
sises the political will in the EU to contribute to the “pol- such as producers of renewable energy.”
luter pays” principle set out in Article 174 of the Treaty
establishing the European Community. The problem with subsidies is that, once introduced, they
are difficult to remove. The existence of environmentally
Electricity generators are not financially penalised for the damaging state aid to mature industries such as coal
208
pollution they cause and the associated costs that soci- and nuclear will inevitably lead to higher environmental
ety has to bear. Environmental costs do not disappear policy costs.
Removing state aid to fossil fuels, nuclear and other Table 1.2: Tax Incentives
mature and environmentally damaging industries has
Country Tax Incentives
many attractions. Not only would it contribute towards a
more level playing field in the electricity markets and cre- Austria Private investors get tax credits for investments in
ate less biased market conditions, it would also save using renewable energies (personal income tax).
The amount is generally limited to 2.929€ per year.
large amounts of money currently spent on unproductive
Belgium 13.5–14% of RES-investments deductible from
state aid schemes and, finally, make it considerably company profits, regressive depreciation of
cheaper to develop the environmental technologies that investments. Reduced VAT on building refurbishing
if energy efficiency is included (6% instead 21%).
are a precondition to securing the EU’s indigenous supply
Denmark The first 3,000 DKK of income from wind energy are
of electricity. Removing environmentally harmful subsidies
tax free.
should ideally be supplemented by energy taxes. Taxation
France Deduction of 15% investment costs with a
can be an effective tool in energy policy if it aims to inter- maximum of 3,000 € per person. Reduced VAT
nalise the costs to society of environmental degradation, (5.5%) on renewable equipment (not applicable to
installation costs).
and contribute to the polluter pays principle.
Germany Losses of investments can be deducted from the
taxable income. This fact increases return on
Tax Incentives investments into wind projects.
Greece Up to 75% of RES-investments can be deducted.
Several EU countries have introduced specific tax incentives Ireland Corporate Tax Incentive: Tax relief capped at 50% of
5
all capital expenditure for certain RES-investments.
for renewable energy. These are summarised in Table 1.2.
Portugal Up to 30% of any type of investments on RES can
be deducted with a maximum of 700€ per year.
Reduced VAT (12%) on renewable equipment.
1.3 Payment Mechanisms –
VOLUME
Spain Corporation Tax: 10% (up to 20% in some
the “Second Best” Solution autonomous regions) tax liability instead of 35% for
investments in environment friendly fixed assets.
The Netherlands EIA scheme: RES-investors (most renewable energy

With environmentally harmful subsidies still in place, and systems) are eligible to reduce their taxable profit
in the absence of environmental taxes that fully reflect the with 55% of the invested sum.
internal costs of energy production, a second best solu-

Lower interest rates from Green Funds:
tion to create a level playing field in the electricity markets

RES- investors (most renewable energy systems)
is for member states (and potentially the EU) to provide can obtain lower interest rates (up to 1.5%) for their
investments. Moreover dividends gained are free of
frameworks that create adequate incentives to increase income tax for private investors.
renewable electricity’s share of the electricity consump-
tion. Source: EWEA (2003).
Usually, the level of the incentive depends on the production will continue until generating companies close down the
cost of wind power compared to other technologies and the least competitive part of their generation capacity and elec-
market prices for electricity. As a result of the gradual liber- tricity demand increases (the European Commission
alisation of electricity markets, competition is increasing in expects electricity consumption in the EU to grow by 1.3%
the European electricity sector. There is some concern, how- annually up to 2030 2).
ever, that Europe is moving from a situation of national elec-
tricity monopolies to private monopolies or oligopolies, Several mechanisms can be applied to promote the
rather than perfect competition. Increased competition, in increased deployment of wind power. These can be
combination with the present over-capacity in European elec- grouped into three main categories:
tricity generation, will probably in the short term make con- • Voluntary systems where the market determines the
209
ditions more difficult for wind power and other renewables price and the quantity of renewable energy (green mar-
as wholesale electricity prices decrease. The price reduction keting).
• Systems where the government dictates the electricity quantities. What matters is that control is achieved in a
prices paid to the producer and lets the market deter- rational and effective manner.
mine the quantity (fixed prices).
• Systems where the government dictates the quantity of The main difference between quota-based systems and
renewable electricity and leaves it to the market to price-based systems is that the former introduces com-
determine the price (renewables quotas). petition between the electricity producers (WT operators).
Competition between WT manufacturers, which is crucial
Fixed price systems and renewables quotas are both ways in order to bring down production costs, is present regard-
of creating a protected market, separate from the open less of whether government dictates prices or quantities.
electricity market where electricity from new renewable
energy sources would have difficulties competing with 1.3.1 VOLUNTARY SYSTEMS AND GREEN
existing, already depreciated nuclear and fossil based MARKETING
power plants. They are also ways of offsetting (fully or
partly) the competitive disadvantage arising from mar- In theory, voluntary demand could provide a market for
kets’ neglect of the environmental effects of conventional wind power and other renewable energy technologies inde-
energy production. pendently of government policy. However, experience with
voluntary systems or “green marketing programmes” to
It is sometimes argued that systems where the govern- date clearly suggests that voluntary green power schemes
ment fixes the quantity of renewable electricity demand purely based on customers’ willingness to pay extra for
(e.g. renewables quotas with green certificate trading) is green electricity (ie. without additional measures), have
more “market oriented” than systems where governments had no noticeable impact on the deployment of wind
fix the price. However, a system where the government energy and other renewable energy sources.
fixes quantity and leaves it to the market to determine the
price is unlikely to be more “market oriented” than a sys- A survey by European Opinion Research Group from 2003
tem where the government fixes the prices and leaves it shows that some willingness exists among Europeans to
to the market to determine the quantity. pay more for energy produced from renewable energy
sources (see Figure 1.1).
Few would argue that the oil cartel OPEC is a market
oriented mechanism because the members have

chosen to control the market through quantities rather
than prices. The reason is that quantities are easier to
administer. In the WTO, however, quantitative restric-
tions are generally banned while tariffs are accepted to
some degree because quotas are regarded as more
market distorting.
The main purpose of the wide range of available econom-

ic measures to support wind energy and other renewable
energy technologies is to provide incentives for techno-
logical improvements and cost reductions of environmen-
tal technologies, in this case the production costs of wind
turbines (WTs). The aim is to ensure the future availabili-
ty of cheap, clean technologies as a competitive alterna-
210
tive to conventional power sources. It is less important
whether markets are controlled through prices or through
Figure 1.1: Results of the Survey Relating Willingness to Pay More for Energy Produced from Renewable Sources
VOLUME
Source: European Opinion Research Group (2003).

However, the number of customers signing up for green tricity product. For comparison, in a survey by Ramboll, a
marketing programmes cannot be directly translated into majority of 58% of Danes answered yes to the question:
support for renewables, as most products contain less than “Would you consider buying more environmentally friendly
100% renewables. In Pennsylvania, USA 60,000 out of electricity, when it becomes possible?”
80,000 customers signed up for a “green” electricity prod-
uct that had a renewable energy content of less than 1%. The Lawrence Berkeley study suggests that the collective
impact of green marketing schemes on renewable elec-
Much research into voluntary green electricity systems tricity generation has been very modest. The study con-
has been conducted in the US where approximately 40% cludes that there is a considerable difference in con-
of households have access to a green power product. One sumers’ stated attitudes toward environmental products
study conducted by Lawrence Berkeley National and the actual demand for them3.
Laboratory at the University of California shows that 0.6%
of the residential customers with access to voluntary Schemes referred to as “shareholder programmes”, “con-
green electricity products are signed up (2000). In tribution programmes”, “ethical trusts”, “green electricity
211
Denmark, only 0.5% of the customers of a Copenhagen- tariffs” or “green electricity labels” are frequently referred
based supply company have decided to buy its green elec- to as voluntary schemes, because customers subscribe
to a service of their own free will. However, in most cases Finally, the system was problematic for Dutch wind power
where voluntary schemes are perceived to be successful developers as they had to compete with cheap production
(in terms of subscriber numbers), the driving force behind from existing, already depreciated renewables plant
the increase in these so-called “green” customers is the abroad, making them reluctant to build domestic capaci-
politically determined framework for investments in wind ty. As a consequence of these shortfalls, the Dutch
power and other renewables rather than high voluntary Ministry of Economic Affairs has decided to change the
demand for clean power. framework.
The Dutch “Voluntary” System “Green” Marketing in Denmark
One example of this is the Netherlands. Following the An important issue relating to green marketing is that it
opening of the Dutch retail market for electricity in July can be difficult for the consumer to make informed choic-
2001, the number of renewable electricity customers es between different suppliers. Determining whether a
increased from 250,000 to 1.4 million in January 2003. product is “green” and how to define “green” requires
However, the main reason behind the rapid increase in time and effort for electricity customers, suppliers and
customers signing up to green power schemes was not regulators.
the population’s willingness to buy green. Exemption from
a 6 c€/kWh ecotax on electricity in combination with a A Danish electricity supplier was marketing a green elec-
production incentive (2 c€/kWh in 2002) was the main tricity product, Naturstrom (Natural Power). In 2002 the
driver. By surrendering a guarantee of origin (not to be Danish Energy Regulatory Authority (Energitilsynet) noti-
confused with tradable green certificates), supply compa- fied the company that it was not allowed to charge its cus-
nies could claim exemption from the ecotax. That made it tomers a premium price for Naturstrom since the compa-
possible for suppliers to sell green electricity as cheap or ny could not prove that the product had any effect on the
cheaper than conventional power. environment. Ironically, for the Danish Energy Regulatory
Authority it was not a problem that the company actually
The Dutch system had obvious flaws, mostly related to sold the product (that is up to the ombudsman). The only
the possibility of importing green electricity from abroad. problem was that it charged a premium price for it. In
The high level of support for renewables in the 2003, the company decided to pay back the surcharges it
Netherlands made it a highly attractive market for foreign had collected from its “green” customers.
renewable electricity producers, which led to high costs in
the form of large avoided tax revenues. That would not be The case illustrates the challenge facing the electricity con-
so serious a problem had the incentive increased the sumer who wants to make an informed choice. In the
renewable electricity production abroad. But most imports Naturstrom case, Danish customers were paying sur-
came from existing plant, e.g. Danish wind power and charges for renewable electricity that would have been pro-
Swedish hydropower that would have been produced and duced regardless of the green marketing programme. No
sold domestically in the absence of the export opportuni- additional green electricity was produced. Furthermore,
ty to the Netherlands. Danish customers were paying extra for green electricity
that Swedish taxpayers had already paid for once through
In short, Dutch taxpayers were paying for renewable elec- investment subsidies and feed-in tariffs to wind power.
tricity production that would have been produced anyway
and, thus, did not increase EU production of renewable That the transparency of the green electricity product mar-
electricity by one single kilowatt hour. Furthermore, it was ket leaves much to be desired is underlined by the fact
unclear whether the Netherlands or the exporting country that Naturstrom was endorsed by the Danish Society for
212
could claim the production towards meeting their renew- the Conservation of Nature and certified by the Swedish
ables Directive targets. Society for Nature Conservation which elaims to have “the
world’s toughest environmental label”. However, for In France, Germany, Greece and Portugal, the tariff is
renewable electricity supplies it is not a requirement to related to the siting of the turbine. In high wind areas, the
obtain the Swedish label that more renewable electricity tariff is lower than in low wind areas. This is to avoid con-
production takes place, says the organisation. centrating the development of wind energy in very windy
areas of a country. In Spain, WT operators can choose
Obviously, not all green marketing schemes are flawed. between a fixed tariff per kWh or a premium above a fluc-
Following a decision by the Irish government to open the tuating electricity price.
electricity market to suppliers of green power, Irish com-
pany Airtricity started supplying green electricity, predomi- 1.3.3 INVESTMENT SUBSIDIES
nantly from wind. The company builds, owns and operates
its own wind farms and sells the electricity output directly In the early days of wind power development, investment
to end customers. Due to Ireland’s enormously rich wind subsidies were often used as an incentive to investors, nor-
resource, it is possible for the company to deliver elec- mally given on the basis of the rated power (in kW) of the
tricity to its customers below or at the same price as the generator. It is generally acknowledged that systems that
national electricity monopoly ESB. 26,000 Irish business- relate the amount of support to the size of the WT rather
es have signed up to Airtricity. According to the company, than the production of the electricity are not ideal because
the production from its wind farms saved the release of they lead to less efficient turbines. The incentive should be
502,968 tonnes of CO2 into the atmosphere in 2003, related to efficiency of electricity production rather than to
5
equivalent to taking 119,754 cars off the road for a year. completing the construction phase of a project.
1.3.2 FIXED PRICE SYSTEMS In the 1990s, India gave a subsidy to WT owners based
on the rated capacity of the turbines. This proved prob-
VOLUME
Figure 1.2 shows the level of feed-in tariffs for onshore lematic because the subsidy was given whether or not
wind energy in the EU-15 as of mid 2003. production was efficient. The scheme resulted in poor sit-
Figure 1.2: Feed in Tariffs for Onshore Wind Plants (c€/kWh)
213
ing of WTs, and manufacturers followed customer power demand. At times, this has led to the somewhat
demands to use very large generators, which improved paradoxical situation that owners of coal power plants
project profitability but reduced production and also receive higher prices for the electricity they supply than
attracted highly dubious products. India has since cor- owners of WTs.
rected the inherent flaws of its incentive scheme and the
market has started to develop properly. In Germany, as a rule of thumb, the additional cost of the
FIT adds approximately 1 € to the average household elec-
For wind energy, the global trend is to reject investment tricity bill per month but, as indicated above, the level is
subsidies as the only means of encouraging investments, difficult to establish when power prices fluctuate. Large
because it is considered economically inefficient as illus- German electricity users receive a discount on the tariff
trated by India’s experience. contribution.
However, investment subsidies can be effective if com- FIT systems have been highly effective at attracting wind
bined with other incentives, as is seen in the UK. In order power investments in Denmark, Spain and Germany.
to take account of the higher cost of offshore wind power Other countries with FITs in place are Austria, France,
compared to onshore, the UK government offers invest- Greece, Luxembourg, the Netherlands and Portugal.
ment grants to offshore projects to complement the
Renewables Obligation (RO), a renewables quota system. The main determinant of whether a FIT model is success-
In the absence of such investment grants either onshore ful at attracting investments is the level of the tariff. Of
development only would be possible or it would be nec- course, the payment mechanism has to be supplemented
essary to create two separate RO markets – one for by adequate grid connection conditions and a well func-
onshore and one for offshore, assuming that both are pri- tioning planning framework. Good planning and grid con-
orities for the government. nection frameworks are a precondition for any mechanism
to be successful.
1.3.4 FIXED FEED-IN TARIFFS
The relatively high level of the FITs in Denmark, Spain
Mechanisms based on fixed feed-in tariffs (FITs) have and Germany is the reason for their success. In con-
been widely adopted throughout Europe. Operators of trast, Belgium, Norway and Sweden have all been run-
wind farms are paid a fixed price for every kWh of elec- ning FIT systems that did not contribute much to wind
tricity they feed into the grid. The cost of the system - power or other renewable energy development.
defined by the difference between the level of the tariff Profitability, rather than the system itself, is what deter-
and the market price of electricity - is borne by the tax- mines success, together with effective planning and grid
payers or the electricity consumers. connection regulation.
The structure of the mechanism makes it impossible to The main benefit of a FIT is that it is simple and often
predict the level of support per kWh. If the level of the tar- encourages better planning. A FIT is not associated with a
iff remains constant, the level of support will change as a formal power purchase agreement (PPA) and has no defi-
result of changing electricity prices. The level of support nite term of existence. In principle, therefore, the level of
per kWh could become negative if electricity prices were the tariff can be changed at any time or removed by
to rise above the level of the tariff. Such a situation has repealing the law. The main disadvantage of a FIT is the
occurred in Scandinavia recently. In 2002/2003, electric- political risk inherent in the system.
ity prices on the Nordic power exchange Nord-Pool has
periodically increased dramatically as a result of low lev- The political risk of the FIT in Spain, seen from an
214
els of water (low electricity supply) in the Norwegian and investor’s point of view, is perceived to be somewhat low-
Swedish hydropower reservoirs combined with increasing ered, since the government has established some degree
of assurance that changes in the tariffs will not bankrupt longer sufficient to attract investments under the overall
existing projects built under previous conditions. economic climate.
However, the risk of political change is not eliminated in
Spain, and investors can only guess how long the tariff Fixed payments for wind power supplied to the grid in
will continue and at what level. Investors therefore have Denmark and Germany, combined with technology
to include a risk premium when planning the financial improvements, falling interest rates and low inflation
soundness of projects. have undoubtedly added to the profitability of investing
in wind turbines (WT) over the past decade. On the
Germany has been able to reduce much of the political other hand, higher profitability has been somewhat off-
risk by guaranteeing payments for 14 to 20 years. If the set by the increasing use of sites with lower wind
tariff is believed to be reduced it will have a negative speeds.
effect on the market for new wind power capacity in
Germany, as was seen in 2003. But those who have But what if interest rates had gone up instead of down?
already invested will not be affected – that is unless the Under a fixed price system, turbine owners would not
government decides otherwise. Some political risk is receive any compensation for the higher cost of finance
therefore still inherent in the German system as - and the outcome would have been lower profitability -
investors generally consider it less risky to enter into unless the fixed price had been adjusted upwards. WTs
long-term PPAs enforceable under civil law rather than are capital-intensive investments with low operating
5
rely on the good will of a government or parliament. costs, so the cost of finance can have profound impact
on project profitability. Had inflation added further to the
Greece is a good example that a sufficiently high FIT does decline in profits, a situation could occur where technol-
not guarantee development of wind energy. The FIT of ogy improvements and economies of scale in WT manu-
VOLUME
90% of the consumer price or approximately 5.75 facturing would not be sufficient to offset the higher
c€/kWh (around 7 c€/kWh if there is no grid access) is costs of finance and lower inflation-adjusted income.
supplemented by up to 40% capital grants. That level
would be sufficient to develop wind energy taking into Such a situation has not occurred, so there is no evi-
account Greece’s wind resources. However, wind power dence of politicians’ willingness to increase tariffs to
development was not taking off in Greece. The main bar- reflect higher wind power production costs. However, it

rier was in the planning system rather than the level of is probably fair to assume that it would be more difficult
the tariff. to convince governments to raise the tariffs – at least in
established markets – than to lower them. That test of
France is faced with a similar problem to Greece. The political will has yet to be seen. The main point is that a
financial incentives in the form of FITS for projects small- fixed price system is rather rigid when it comes to
er than 12 MW and auctions for larger projects seem adjusting tariffs – whether up or down – to reflect
adequate, but little wind development is taking place. changes in the production costs of wind power. It should
The main problem in the past has been grid and, espe- be mentioned that inflation risk can be avoided by
cially, planning barriers. However, the French government including an automatic inflation adjustment to the mech-
seems determined to overcome these following a nation- anism, as is done in the US production tax credit (see
al energy debate in 2003. section 1.3.6).
The political risk of FITs is usually understood as the risk 1.3.5 FIXED PREMIUM SYSTEMS
that a government will progressively lower the tariff to
reflect the fact that wind power becomes cheaper as the A “fixed premium” or “environmental bonus” mechanism
215
technology develops. But there is also the potential risk is another variant of the fixed price system. Rather than fix-
that a government will take no action when a FIT is no ing the price, the government fixes a premium to be added
to the electricity price. The cost per kWh of the system is, a cash payment does not make a big difference from a
contrary to the fixed FIT, predictable, although the total socio-economic or investor perspective. Politically, howev-
costs to society depends on the level of development. er, it can make a difference whether an incentive is paid
From the perspective of a WT owner, the total price by the electricity consumer or by the taxpayer.
received per kWh (electricity price plus the premium) is
less predictable than under a FIT because it depends on The largest wind power market to make use of a tax cred-
a changing electricity price. it is the US. Canada is also considering introducing a tax
driven system. The US market is driven by the federal pro-
In theory, a mechanism that is based on a fixed premi- duction tax credit (PTC), which is worth approximately 1.8
um/environmental bonus that reflects the external costs c/kWh. It is adjusted annually to take inflation into
of conventional power generation could establish fair account.
trade, fair competition and level the playing field in the
internal electricity market between renewable energy In recent years, there have been three separate phases of
sources and conventional power sources. Together with the PTC. The first phase ended on 30th June 1999 and
taxing conventional power sources in accordance with was not renewed until 1st January 2000. The second PTC
their environmental impact (see volume 4), fixed premium expired on 31st December 2001. Again, there was a gap
systems are, theoretically, the most effective way of inter- before its extension was announced in March 2002, with
nalising external costs. the third PTC continuing until December 2003. As at
December 2003, the tax credit has not been extended, as
From a market development perspective, the advantage of it is included in a controversial energy Bill on which
a price premium is that it allows renewables to penetrate Congress has not yet reached.
the market very quickly if their costs drop below the elec-
tricity price plus premium. If the premium is set at the As a result of the relatively short lifetime of each individ-
“right” level (theoretically at a level equal to the external ual PTC, the market has been very volatile and charac-
costs of conventional power), it allows renewables to com- terised by “boom and bust” cycles. Activity usually picks
pete with conventional sources, without the need for up dramatically prior to the end of a PTC. There was an
politicians to set quotas. enormous amount of activity in late 1998 and early 1999,
almost no activity in 2000, and a great deal of activity
In practice, however, basing the mechanism on the envi- again in 2001. Activity was picking up again prior to the
ronmental benefits of renewables is challenging. Very December 2003 deadline. For both investors and manu-
ambitious American and European studies (such as the facturers, these boom-bust cycles are highly problematic
European Commission’s ExternE project) on the external because it makes planning very difficult. Most European
costs of power generation have illustrated that establish- WT manufacturers have plans to start up local production
ing the exact costs is very complex. How do we account for in the US, but are reluctant to execute them until long-
lost homes on Pacific islands if the icecap melts, or put a term stability is secured.
price on deteriorating health? In reality, fixed premiums for
wind power and other renewable energy technologies, such 1.4 Fixed Quantity Systems
as the Spanish model, are based on estimated production
costs and comparisons with the electricity price rather In fixed quantity or “renewable quota” systems (“renew-
than the environmental benefits of renewable energy. able portfolio standards” in the US), the government
sets a quota for the level of renewable energy that
1.3.6 TAX CREDITS should be produced. It is then up to market forces to
determine the price. Two types of renewable quota sys-
216
A tax credit is another variant of the fixed price system. tems have been employed in national wind power markets:
Whether an incentive is given in the form of a tax credit or “tendering systems” and “green certificate systems”.
1.4.1 TENDERING SYSTEMS power purchasing contracts that bidders compete for are
enforced under civil law. From an investor risk perspec-
Tendering systems or competitive bidding has been or is tive, a long contract is very attractive since it minimises
being used to promote wind power in Ireland, France (for risk. A second attraction of a well-designed tendering sys-
wind farms larger than 12 MW) and the UK. Scotland and tem is that the government (as well as electricity users
Northern Ireland have also made use of the mechanism. and taxpayers) does not have to make best guesses
Developers of wind farm projects are invited to bid for a about the cost of producing wind power. Through the bid-
limited wind energy capacity in a given period. The com- ding process, the market sends a clear signal to the gov-
panies that bid to supply electricity at the lowest cost win ernment about the cost of wind power production. The
the contracts. Usually, 15-year PPAs are entered into. The political risk of tendering systems is therefore lower than
difference in price between these contracts and the price that of fixed price systems. However, investors are faced
of conventional power represents the additional costs of with another risk element under tendering. All developers
producing green electricity. that enter a bid risk losing the planning costs if the bid is
not accepted or if planning permission is not given on the
One of the major drawbacks of the tenders made so far is location in question.
that they have encouraged gaming of the system. Wind
energy is a technology that gets cheaper with time. Following the NFFO experience, most countries have dis-
Therefore, a contract holder will wait as long as possible regarded tendering procedures. At present, only Ireland
5
before building a project. Partly because of this inherent continues its competitive bidding procedure through the
flaw, the UK’s non-fossil fuel obligation (NFFO) tender sys- AER (see table 1.3), although it is considering changing
tem did not result in many projects being built. Another its system. The overall objective of the AER is to secure
flaw of the NFFO model was that it did not penalise devel- 500 MW of new renewable energy capacity in the period
VOLUME
opers if they failed to install the capacity for which they 2000 to 2005. The winners of the tender are awarded
had secured a power purchase contract. In principle, any- PPAs for 15 years.
one was free to make an unrealistic low and unprofitable
bid, win the contract and not develop the project. The inef-
fectiveness of the UK NFFO system led the government to
abandon the model and introduce a new system based on

tradable green certificates (see below).
The NFFO was heavily criticised for its failure to deliver,

and the UK experience has discredited tendering systems
substantially. Although the NFFO had obvious flaws, as
described above, that does not mean that tendering sys-
tems cannot function. They need to be better designed.
The problem with falling production costs over time could
have been overcome by introducing deadlines.
Furthermore, the model should be combined with a per-
formance bond and meaningful penalties for failing to
meet the contract. Finally, poor planning procedures in the
UK must also take their share of the blame for the disap-
pointing performance of the NFFO.
217
If designed correctly, tendering systems can work. One of
the main attractions of the model is that the 15-year
Table 1.3: Status of AER Contracts expected to be published in December 2003 but had not
AER No. Launched Technology Supported Cap been published at the time of writing. However, at a meet-
Capacity Price/KWH ing arranged by the Danish Wind Turbine Owners’
Amount
Association on 27th November 2003, a leading civil
(MW)
servant from the Danish Energy Authority revealed a few
AER I 1994
details of the forthcoming tender. The information cur-
Wind 73,1 IR £0,04
rently available indicates that the competitive bidding will
Hydro 4,3 IR £0,04
be combined with a price cap of 4.8 c€/kWh. Both the
Biomass 11,8 IR £0,04
Danish Wind Industry Association, which represents
CHP 22,6 IR £0,04
Danish WT manufacturers, and the Danish Wind Turbine
AER II 1995
Owners’ Association have already expressed some con-
Biomass/waste 30 IR £0.036
cerns about the price cap.
AER III 1997
Wind 137,3 ≤IR£0,039

The main criticism is that the cap will make the tender
Hydro 4,4 ≤IR£0,039
meaningless as bids will not be made. Potential develop-
Biomass/waste 17 ≤IR£0,039
ers are asked to take the risk of low electricity prices on
AER IV 1997
the power exchange, but will not get any benefit if electric-
CHP 49,6 ≤IR£0,03
ity prices increase above the cap. The Danish organisa-
AER V 2001 Large scale
tions suggest that the tender is conducted as an auction
wind (>3MW) 318,3 MW c€ ≤4,812
over the lowest “environmental bonus”, defined as the pre-
Small scale
mium above the market price that is required to build the
wind (<3MW) 35,795 c€ ≤5,297
power plants. The cap should be removed in order to cre-
Biomass 8 MW c€ ≤5,916
ate competition and attract a sufficient number of bids.
Hydro 0,949 c€ ≤6,475
Only then will the price signal about the cost of offshore
AER VI 2003
wind power be effective and able to secure the lowest cost
Large scale
to the consumers. Furthermore, the organisations argue, it
Wind (<3MW) 259,82 c€ ≤ 5,216
is neither market compatible nor in accordance with the
Small scale
polluter pays principle that wind power developers, as the
wind > 19,6 c€ ≤5,742

only electricity producers in Denmark, are asked to bear
Offshore Wind 50 c€ ≤8,4
the downside risk of low market prices and at the same
Hydro 5,483 ≤
time face income restrictions if market prices go up.
Landfill gas 1,309 c€ ≤7
Although the final proposal had not been published at the
Biomass
time of writing, the debate is already taking place and is
(anaerobic
illuminating some of the issues that are relevant to the
digestion) 2,022 c€ ≤6,412
design of tendering mechanisms.
Biomass CHP 26,83 c€ ≤ 7,018
If designed correctly, tendering systems could probably

Source: Ener-Iure Project Phase III (2002), Sustainable Energy Ireland (2003).
function adequately, as have offshore oil and gas leases.
However, it still remains to be proved that the system can
Denmark is planning to introduce a tendering procedure be effectively applied to wind power investments.
for its future offshore wind power development. The coun-
try, which gets some 20% of its electricity from wind power The sunk planning cost risk described above will also
in 2004, is following a strategy that future development of have an effect on the ownership structure of the wind
218
wind power should be offshore combined with repowering energy market. As projects increase in size the sector is
of onshore wind energy. The tender conditions were witnessing a shift in ownership away from individuals
towards larger developers. The popular element of the electricity (it is of little importance who has the obli-
the early days of wind power cooperatives and individual gation). The generators (producers), wholesalers, retailers
ownership will probably vanish unless new collective risk- or consumers (depending who is obliged in the electricity
sharing project development institutions are developed. supply chain) are obliged to supply/consume a certain
Another effect of the tendering system would be to con- percentage of electricity from renewable energy sources.
centrate development in the windiest areas. That is desir- At the settlement date, they have to submit the required
able from an economic efficiency perspective, but may number of certificates to demonstrate compliance.
have implications for planning and public opinion. Obligated parties obtain certificates in three ways:
• they can own and operate renewable energy plant;
Finally, the model is probably better suited to large off- • they can purchase certificates from another renewable
shore wind farms than onshore projects. But the planning energy generator; or
issues must be dealt with, deadlines must be in place • they can purchase certificates from a generator or bro-
and there should be meaningful penalties for not building. ker by purchasing certificates that have been traded
Imposing price caps appears incompatible with the basic independently of the power itself.
idea of tenders – to get the market to provide price sig-
nals for the production of wind power. The (gradually increasing) obligation creates a demand for
TGCs. It is left to the market to deliver the supply of and
1.4.2 TRADABLE GREEN CERTIFICATE establish a price for certificates. TGCs are issued to pro-
5
SYSTEMS ducers of renewable electricity in proportion to the volume
of green electricity they generate. A TGC serves as evi-
Tradable green certificate (TGC) systems are similar to dence that a specific amount of green power has been
tendering. The main difference is that the price for the produced and fed into the grid. If demand for certificates
VOLUME
power and certificate is settled daily on the electricity mar- exceeds supply - the amount of renewable electricity pro-
ket and there is a separate market for tradable certifi- duced is lower than the government quota - then the price
cates (tendering systems are typically based on 15-20 of certificates will rise. It will continue to do so until the
year PPAs). With daily setting of prices, the TGC model is price satisfies investors’ requirements for return, whereas
more risky for the investor unless an effective market for new capacity will be installed to meet the quota.
long-term certificates contracts (probably in the form of

financial futures or options) is developed. Currently, the introduction of tradable renewable certificate
systems has been proposed or is being implemented in
If a TGC market works effectively, the price of a certificate Denmark, UK, Belgium Sweden and Italy (see Table 1.4).
will reflect the difference between the market price of
electricity and the generation costs of new renewable gen-
erating capacity. The value of a certificate thus represents
the additional cost of producing renewable electricity com-
pared to conventional sources. That value will only by coin-
cidence be equal to the environmental benefits of wind
power and other renewables.
In theory (we have not yet seen a well functioning TGC

market for wind power), the mechanism should work as
follows: the government sets a specific and gradually
increasing quantity – or minimum limit – for the amount of
219
renewable electricity in the supply portfolio. An obligation
is placed on either the electricity suppliers or end users of
Table 1.4: Overview of Various Certificates Models
Denmark UK Belgium Belgium Italy Sweden

(Flanders) (Wallonia)
Period start 2002 start 2002 start 2002 start 2002 start 2002 start 2003
Obligation 20% by end 2003 3% in 2002; 4.3% 1.2% (2003), 2% 3% in 2003 2% in 2002 and 7.4% in 2003,
(proposal aban- in 2003; 10.4 % in (2004) increasing increasing up to will be increased 16.9% in 2010
doned) 2010; 15% in 2015 up to 6% in 2010 12% in 2010 annually by 0.35%
From September between 2004 and
2010 onward, the 2008
quota will be multi-
plied by a factor of
1.01
Obligation on end user supplier supplier supplier producers and end user
importers
Technology bands no no no no no no
(baskets) within
overall quota
Involved small hydro, wind, small hydro, all renewables, no all renewables and all renewables (incl. small hydro (<1.5
technologies bio-mass, solar, wind, biomass, solid municipal high quality CHP large hydro), facili- MW), large hydro
geothermal energy, solar, geothermal waste ties not older than (only in some
no waste energy, no waste eight years cases), wind, bio-
mass, geothermal,
wave energy
International trade no no no no yes, but only in no
allowed exchange for physi-
cal electricity
Price restrictions min= € 0.014/kWh not planned, max max price according max price defined n.a. min prices in the
(min/max price) max= € 0.037/kWh price according to to penalty, by penalty. introductory phase:
penalty min at federal level. Min - producers of in 2003 € 6
From 1.7.03 RES-E may in 2004 € 5.5
onward the grid exchange their TGC in 2005 € 4.4
operator has the for a subsidy at a in 2006 € 3.3
obligation to buy fixed price of € 65. in 2007 € 2.2
TGCs issued any- At federal level, in 2008 € 0.
where in Belgium from 1.7.03 onward max price according
for the minimum the grid operator to penalty
prices per TGC of: has the obligation
€ 90 (offshore to buy TGCs issued

wind); € 50 anywhere in
(onshore wind) Belgium for the
€ 50 (hydro); minimum prices per
€ 150 (solar); TGC of: € 50
€ 20 biomass (onshore wind);
€ 50 (hydro);
€ 150(solar);
€ 20 biomass
Penalty € 0.037/kWh The buy- out price € 75 per certificate from 1.4.03 n.a. 150% of the market
is £30.51/kWh (1,000 kWh) in onward: € 100 per price, but with a
(for 2003/2004) 2003, € 100 in TGC (1,000 kWh) maximum of a SEK
(~€ 45/kWh) 2004 and € 125 in 175 for certificates
2005 that should have
been surrendered
during 2004, and
SEK 240 for 2005
Trading scheme n.a. stock exchange stock exchange open trading and free or in the power open
direct support pool
Comments: Denmark: the proposed TGC was abandoned. UK: recent developments have shown that the certificate price is higher than the buy-out price. This development
is due to a shortage of certificates due to limited RES-E construction and the fact that the buy-out revenues for non-compliances are recycled to the suppliers in proportion to
220 the certificates they have used for complying with the obligation. Belgium: the green certificate system is run by regional regulators. Only certificates for offshore wind
energy (non-existent as yet, but expected) will be issued by the federal government. Italy: GRTN (Italian Transmission System Operator) strongly influences the certificates’
market by selling its own certificates from old CIP6 plants at a regulated price, namely that set by law as the average of the extra prices paid to acquire electricity in the CIP6
programme that year. The Italian system has little to do with TGC systems because both the price and the quantity of certificates issued are fixed.
The Dutch system (see section 1.3.1) is also sometimes In theory, under a Renewables Portfolio Standard (RPS) all
erroneously promoted as a TGC system, due to some con- changes - or rather, all expected changes - in the cost
fusion about the difference between TGCs and guaranties determinants of wind power investment will be immedi-
of origin (GoO). A GoO can be compared to content ately reflected in the combined price of electricity and the
descriptions in the supermarket. They are not tradable in price of renewable energy certificates. Likewise, a fall in
themselves unless the products are homogenous in electricity prices will be accompanied by an equal rise in
nature and a financial market is established. The exis- the price of the certificate.
tence of GoOs in the Netherlands does not therefore
make it a TGC system. As with tendering mechanisms, the rationale behind the
TGC system is to reduce the costs of expanding renew-
Complex Design able capacity by introducing competition between produc-
ers of renewable energy. Price competition will be trans-
The TGC mechanism is more complex in nature than other ferred further down the system and renewable energy pro-
payment mechanisms. WT operators will have to be active ducers will seek to bargain with turbine manufacturers
in two interrelated financial markets: one for TGCs and and land owners for lower prices to a larger extent than
one for power. One of the problems is that there seems under the fixed price system.
to be an asymmetry between the demand and the supply
side in the markets. WT owners would prefer to have as The role of the TGC market, as any other market, is to
5
long a contract as possible to minimise risk, while the establish a price according to the laws of supply and
electricity companies on the demand side seem to prefer demand. But determining a price is problematic when sup-
short contracts. It is essential that the certificate market ply and demand are fixed in the short term (the problem
is able to attract financial arbitragers and speculators that of vertical demand and supply curves). A price cannot be
VOLUME
can allocate risk. determined if a situation where demand equals supply is
an exception. The effect will be that the price will tend to
Ideally, there should be no floor or cap on the price of cer- be banging either against the price cap created by the
tificates. However, there will need to be a penalty for not penalty or the price floor (if there is one), rather than float-
complying. As with any other penalty, this should be set ing in the mid-range.
at a level so high that it will never be enforced. A high

penalty is one reason for the success of the US SO2 trad- In other words, if there is one certificate more in circulation
ing market. If the penalty is set too low (too close to the than is needed to comply with the obligation, nobody will
expected market price of the certificates) it will act as a want to buy it and it has no value. This is a problem arising
price controlling factor, which will distort the market. from fixed demand and fixed short-term supply - it causes
the price to become very volatile, fluctuating between zero
In the ideal market, the price of the certificate and the or infinity (or zero and the level of the fine for not complying
expected price of electricity will always add up to what with the obligation).
economists refer to as the “marginal cost” of producing
wind power; that is, the costs of adding one more unit - a In order to eliminate price fluctuations caused by the fixed
WT - to the generating base. In reality, any change in costs demand and to secure flexibility in payment, a system of
associated with wind power production will be compen- “banking” must be available. Certificates will be issued at
sated for by an equal change in the combined income the time of production of renewable energy and will be
from selling the electricity and its accompanying certifi- destroyed, in accordance with the requirements of the
cate. If, for example, interest rates rise, so will the com- obligation, on delivery to an independent authority. But
bined payment. If sites with poorer wind resource are there will most likely be an imbalance between actual pro-
221
used the combined payment will also rise. And if technol- duction of wind kilowatt hours and the quota obligation for
ogy improves it will fall. any given period. The market must be structured to cope
with the imbalance. A banking system could be a solution. certificate for each technology is the answer. One certifi-
Such a system gives consumers the option of buying cate, however, only ensures development of the cheapest
future production - and WT owners the option of selling renewable technology, while several certificates will result
future production by trading borrowed certificates. This in a market with dangerously low liquidity, at least in the
stabilises fluctuating prices by creating a basis for long- beginning of development.
term certificate purchase contracts. The system thus
allows participants in the market the option of hoarding One way to deal with the problem is to accept, say, that
certificates in the expectation of future price changes, and PV is 10 times more expensive than wind power and issue
WT owners the option of borrowing certificates in case 10 times the number of PV certificates. But such a solu-
their turbines do not produce enough electricity to meet tion brings us back to the fixed price problem - there is no
their long-term delivery contracts. easy way of estimating the true production costs of the
various technologies, which makes it difficult to determine
Because of the expected imbalances between actual pro- the proportional relationship between the costs of wind
duction and the renewables quota, and the problems of and solar. What if the cost of solar drops 10% from its
volatile certificate prices, consumers or electricity suppli- current level, and the certificate proportion is not changed
ers (who bear the weight of the obligation to buy renew- politically to reflect the drop? Investment in solar would
ables) must be able to hand in contracts on future delivery soar and nobody would invest in other renewables
by sending a larger number of certificates to destruction. options, even though they may be several times cheaper
This can be viewed as the “interest” on a certificate. With in real terms - hardly a cost effective way of meeting
no interest element, the obligation will be met at the latest renewable energy targets. Offshore wind, being more
possible juncture, causing disturbance in the market. expensive than the onshore variety, gives rise to the same
problem; a way needs to be found of stimulating its devel-
For owners of WTs (and their sources of finance) it is of opment if politically desirable.
paramount importance that any payment system allows a
fair amount of certainty to be attached to cash flow pro- Furthermore, issuing certificates in proportion to estimat-
jections. In support systems based on fixed price this ed production costs requires constant evaluation of the
tends to be less of a problem. But when selling power and costs of technologies as well as political intervention in
certificates on spot markets with fluctuating prices it the form of changed certificate proportions. The political
could become a problem. It increases the risk and there- risk in such a market would be substantial. One certificate
by the cost of producing wind power. for all technologies would also make it impossible to
determine the price of pollution abatement in relation to
Financial long-term contracts could limit this problem the individual technology and determine when a technolo-
through the establishment of a futures or options market. gy will no longer need support: an “umbrella certificate”
By selling electricity and certificates on long-term futures will still have a positive value when the least costly tech-
or options contracts, the risk (and the price) can be nology becomes competitive. The risk is that support will
reduced. Futures and options contracts make it possible be given to technologies that no longer need it.
to sell or buy certificates for delivery some time in the
future at a price that is agreed upon today. Such a market On the other hand, the liquidity problem of having several
would need an institution to facilitate trade and guarantee certificates cannot be ignored. Low liquidity is a problem
delivery if a WT owner is unable to deliver. for efficiency in any market. There are compromise solu-
tions, however. Certificates could be issued to different
Another aspect to consider is whether all renewables technologies in exactly the same proportions, with the
technologies defined in the EU Directive on promotion of less competitive technologies receiving separate subsi-
222
electricity from renewable energy sources should be dies. Or instead of granting direct capital investment sub-
included in a single “umbrella certificate” or whether a sidies, auctions could be held for subsidised contracts, to
encourage competitive bids and provide an incentive to The UK Renewables Obligation
reduce costs, much along the lines of the UK NFFO.
The most ambitious attempt to promote green electricity
Voluntary Demand and TGCs through tradable green certificates is the UK’s RO. This
came into force in April 2002, replacing the much
In the meantime, there is “green marketing” (see section maligned NFFO tendering procedure. Suppliers have to
1.3.1) to consider. There is potential for the introduction demonstrate compliance with the RO through the presen-
of a green pricing market where like-minded citizens and tation of renewables obligations certificates (ROCs) which
companies can opt to pay a premium to receive their elec- are issued in proportion to green electricity production.
tricity from a renewable source. Each ROC represents 1 MWh of renewable electricity from
eligible generators. Ofgem, the regulator, is responsible
Consumers who want to buy renewable energy in excess for the administration of the RO and for the compliance.
of the obligation must have the opportunity of doing so For the period of 2003/2004 suppliers have to meet
without being cheated. So far there is no European struc- 3.4% of total electricity consumption through renewables.
ture in place to allow for this. The obligation runs yearly, rising to 10.4% in 2010 and to
15% in 2015. It is due to finish in 2027. To meet the RO
Discussions to date among electricity sellers have cen- suppliers have three options:
tred on offsetting electricity bought by consumers in • purchase ROCs through the supply of renewable
5
excess of the obligation against the quota requirement. energy purchased from eligible generators;
Should this be allowed, consumers would be fools to buy • buy ROCs from other suppliers or from the NFPA which
excess certificates. Such purchases would not lead to periodically auctions the ROCs it has acquired under
more renewable energy being produced and these green the existing NFFO contracts; or
VOLUME
consumers would be paying for those who declined to • pay the penalty or buy-out price set by Ofgem for not
meet their obligation. A clean environment is considered meeting the quota.
a public good. If a neighbour buys it, they cannot prevent
you from benefiting from their purchase for free. The buy-out price is currently set at £30.51 per MWh. All
payments proceeding from the buy-out price made by sup-
Legislation is necessary if the market is to work properly. pliers for each MWh of shortfall between the amount of

Just as it is on stock and bond markets. Legislation and their obligation and the number of ROCs presented are
rules create a framework within which the market can placed in a central fund. This money is redistributed to
work. Once the framework has been established, inter- suppliers which have met the obligation in proportion to
vention from law-makers should be kept at an absolute the number of ROCs presented. Therefore, the real costs
minimum. Otherwise prices on the market will reflect for a supplier who is not complying with the obligation is
expectations of political action rather than fundamental higher than the buy-out price. That explains why ROCs are
economic relationships. trading at higher prices than the penalty. ROCs are traded
at approximately £40 to £50 while the buy-out price is
Economic theory, however, does not reflect the real world. £30. Figures released by Ofgem in 2003 show that elec-
When politicians get into the game of creating free mar- tricity companies missed the target for renewable elec-
kets, so do numerous opposing views and compromises. tricity by 40% in the first year of the RO. 5.5 TWh of renew-
WT owners want one thing, electricity companies another able electricity was produced, while the obligation was
and consumers (voters) a third. Senseless compromises around 9 TWh. The RO does not differentiate between dif-
can very well create disturbances to the market mecha- ferent renewable technologies so only the cheapest tech-
nism and, in the worst case, render it useless. nologies will be developed unless additional measures
223
are introduced or other markets are developed. For off-
shore wind power, the government supplements the RO
with capital grants to reflect the fact that offshore wind
power is currently more expensive than onshore.
The UK ROC system is providing valuable experiences with

the complicated task of developing TGC systems. So far,
the mechanism has not proved effective in adding renew-
able capacity and the cost of the system (on a cost per kWh
basis) is high. However, it is still early days in the ROC mar-
ket and the UK government seems determined to make the
necessary adjustments and prove sceptics wrong.
Figure 1.3 shows the number of ROCs issued in the UK

between October 2002 and February 2003 by technology
and country.
Figure 1.3: ROCs Issued Between October 2002 and February 2003 by Technology and Country in the UK
224
Source: Ofgem Newsletter (2003).
1.5 Renewables in the New Member ments are currently in the process of debating new
States frameworks for investing in wind power and other renew-
ables. Fixed FITs are currently in place in both countries,
From 1 May 2004, 10 new EU member states will have but it is difficult to obtain long-term PPAs.
to comply with the EU renewables Directive. Electricity
from renewable energy sources met 5.6% of the new Appendix K provides an overview of wind power develop-
member states’ electricity supply in 2000. To meet the ment in various new member states.
indicative targets, that share will need to rise to 11% by
2010 (see Table 1.5).
1.6 The EU Legal and Political
Some of the new member states already have payment Framework
mechanisms in place. Hungary, Latvia and Estonia have
introduced renewable tariffs. But development of wind The Treaty establishing the European Community calls
power faces bigger barriers than simply the payment for: a “balanced and sustainable development of
mechanism. There is a need to develop more accurate economic activities” and “a high level of protection and
resource assessments, wind maps, reliable grid rein- improvement of the quality of the environment” (Article
forcements and grid connection frameworks. In many 2); “the integration of environmental protection require-
countries, distribution and transmission grids are not ments in the implementation of Community policies with
5
robust enough to support large penetration of wind a view to promote sustainable development” (Article 6);
power. In Poland, up to 1,000 MW of wind power is and bases its Community policy on the environment
planned in the north of the country on the Baltic coast. on the principles that preventive action should be taken:
Infrastructure reinforcements will be needed to transport “that environmental damage should, as a priority, be
VOLUME
the electricity generated to the main population areas fur- rectified at source and that the polluter should pay”
ther south. Both the Polish and Czech Republic govern- (Article 174).
Table 1.5: Renewables in the New Member States
1999/2000 2010

Renewable Gross Total Gross Proportion Renew- Renewable Gross Total Gross Proportion Renew-
Consumption (TWh) Consumption (TWh) able Electricity (%) Consumption (TWh) Consumption (TWh) able Electricity (%)
Cyprus 0.00 3.00 0.05 0.26 4.27 6
Czech Republic 2.34 61.70 3.8 5.66 70.7 8
Estonia 0.01 6.75 0.2 0.37 7.3 5.1
Hungary 0.22 28.30 0.7 1.71 47.4 3.6
Latvia 2.76 6.50 42.4 4.09 8.3 49.3
Lithuania 0.33 9.95 3.3 0.80 11.4 7
Malta 0.00 1.80 0 0.10 2 5
Poland 2.35 140.00 1.6 10.50 140 7.5
Slovak Republic 5.09 28.30 17.9 9.24 29.8 31
Slovenia 3.66 12.20 29.9 4.91 14.6 33.6
Total EU 10 16.8 298.5 5.6 37.62 335.77 11.21
Total EU 15 338.41 2435.00 13.9 646.6 3000.26 21.6
Total EU 25 355.2 2733.5 13.03 684.22 3336.03 20.5 225

Source: Directorate-General for Transport and Energy (2003).
The Lisbon strategy aims to make the European econo- But as the contribution of renewable energy sources to
my “the most competitive and dynamic knowledge-based the EU supply mix grows, so too will the distorting effects
economy in the world, capable of sustainable economic of the many different payment mechanisms currently in
growth with more and better jobs and greater social place in the EU. That is not compatible with free trade,
cohesion”. according to the European Commission, and they could
eventually be subject to internal market conditions.
According to the Commission’s Green Paper on security of
energy supply, in two decades Europe will be importing On 26th June 2003 the European Parliament and the
70% of its energy (up from 50% today) unless it changes Council adopted a new Directive concerning common
direction. Wind power can plug the gap in the European rules for the internal market in electricity. The new
energy supply and, at the same time, contribute greatly to Directive seeks to achieve the complete opening of the
the goals set out at Lisbon: economic growth, high quali- EU electricity market by July 2007. It aims to reduce the
ty jobs, technology development, global competitiveness, risk of market dominance and predatory behaviour and
and European industrial and research leadership. ensure non-discriminatory transmission and distribution
Furthermore, wind power and other renewable energy tariffs and network access. Furthermore, it establishes
technologies will have a large impact in meeting the EU’s provisions for the unbundling of transmission and distri-
Kyoto commitments and contributing to sustainable devel- bution operators and establishes labelling requirements
opment. The Green Paper recognises that: for electricity suppliers regarding CO2 emissions and
radioactive waste from electricity production as well as
“Renewable sources of energy have a considerable the contribution of each energy source in the supplier’s
potential for increasing security of supply in Europe. fuel mix.
Developing their use, however, will depend on extreme-
ly substantial political and economic efforts... In the The EU recognises that wind power and other renew-
medium term, renewables are the only source of energy ables remain at a competitive disadvantage to fossil and
in which the European Union has a certain amount of nuclear sources. The support and payment mechanisms
room for manoeuvre aimed at increasing supply in the currently in place in the member states can be regarded
current circumstances. We cannot afford to neglect this as a substitute for a pollution tax on energy. Renewable
form of energy.” electricity technologies do not benefit from decades of
financial support and would be required to compete with

“Effectively, the only way of influencing [European existing nuclear and fossil fuel power stations producing
energy] supply is to make serious efforts with renew- at marginal costs, because interest and depreciation of
able sources.” power plants have already been paid for by electricity
consumers and taxpayers. The medium term solution
1.6.1 THE ELECTRICITY DIRECTIVE could be to create an internal market for renewables and
the European Commission may, if it deems necessary,
In December 1996, joint rules for an internal electricity propose a harmonised framework for renewable electric-
market in the EU were adopted. The overall goal is to ity in 2005.
increase the economic efficiency of the electricity supply
in the EU by introducing competition. EU countries are 1.6.2 THE RENEWABLES DIRECTIVE
therefore in the process of gradually liberalising their
electricity markets and full liberalisation should be Directive 2001/77/EC on the promotion of electricity
achieved by 2007. So far, renewables have avoided from renewable energy sources aims to double the
being included in the liberalisation process because amount of electricity produced by renewable energy by
226
their contribution to total electricity supply is small and 2010. Indicative targets for shares of electricity have
therefore causes little distortions. been set for each member state (see Appendix M). The
Commission is assessing member states’ progress “It is, however, necessary to adapt, after a sufficient
towards meeting the targets. The Directive States: transitional period, support schemes to the developing
(1)” The potential for the exploitation of renewable internal electricity market.”
energy sources is underused in the Community at
present. The Community recognises the need to Furthermore, it is emphasised that the intention is to
promote renewable energy sources as a priority meas- use market forces to make renewable energy sources
ure given that their exploitation contributes to environ- competitive.
mental protection and sustainable development. In
addition this can also create local employment, have a “It is important to utilise the strength of the market
positive impact on social cohesion, contribute to secu- forces and the internal market and make electricity
rity of supply and make it possible to reach Kyoto tar- produced from renewable energy sources competitive
gets more quickly. It is therefore necessary to ensure and attractive to European citizens.”
that this potential is better exploited within the frame-
work of the internal electricity market.” The Directive on electricity from renewable energy pro-
vides the wind power and other renewables industries
(2)” The promotion of electricity produced from renew- with crucial assurance that the EU is determined to fur-
able energy sources is a high Community priority as ther progress the development and integration of renew-
outlined in the White Paper in renewable energy able energy technologies. It sends a powerful signal to
5
sources for reasons of security and diversification of the industry of long-term political commitment at
energy supply, of environmental protection and European level which, in return, should reduce invest-
economic cohesion.” ment risks and thereby the cost to society of developing
and integrating renewable energy sources.
VOLUME
Article 9 stipulates that member states shall bring into
force the laws, regulations and administrative provisions The liberalisation process will not create a perfect mar-
necessary to comply with the Directive not later than ket or a level playing field overnight and the RES-E
October 2003. Directive ensures that short-term distortions of the
European electricity markets do not undermine the pos-
The purpose of the Directive is: sibility of developing those renewable energy technolo-

“To promote an increase in the contribution of renew- gies that will facilitate a future European energy supply
able energy sources to electricity production in the that is cheap, clean and independent of fluctuating oil
internal market for electricity and to create a basis for and gas prices.
a future Community framework thereof.” (Article 1).
For many years, the European Wind Energy Association
The Directive recognises that: (EWEA) has called for a level playing field in the electric-
“It is too early to decide on a Community-wide frame- ity sector, including the internalisation of external costs
work regarding support schemes, in view of the limited in electricity prices and the removal of state subsidies to
experience with national schemes and the current rel- conventional energy sources. The Commission has
atively low share of price supported electricity pro- actively pursued the same agenda, e.g. by proposing
duced from renewable energy sources in the common energy taxes and reductions in the level of
Community.” state aid to conventional energy technologies. However,
these have still to materialise. In the meantime, the
However, the Directive states that mechanisms should renewables Directive serves as a good substitute. It will
eventually be adapted to include renewable energy allow the industry to develop the renewable energy tech-
227
sources in the internal electricity market: nologies that will secure the availability of cheap, clean
energy and a future European indigenous energy supply.
No later than October 2005 the Commission shall present Göteborg. This must include action aimed at decou-
(Article 4): pling environmental degradation and resource use
A well-documented report on experience gained with from economic growth. Despite some progress, the
the application and coexistence of the different [pay- worrying trends observed when the Strategy was
ment] mechanisms. The report shall assess the suc- launched have not been reversed, and a new impe-
cess, including cost-effectiveness, of the support sys- tus must therefore be given.”
tems referred to in paragraph 1 in promoting the con-
sumption of electricity produced from renewable ener- “54. Against this background, the European Council:
gy sources in conformity with the national indicative - invites Member States to accelerate progress
targets referred to in Article 3(2). This report shall, if towards meeting the Kyoto Protocol targets, including
necessary, be accompanied by a proposal for a the reduction of greenhouse gas emissions, the
Community framework with regard to support increase in the share of renewable energy, setting an
schemes for electricity produced from renewable ener- EU-wide indicative target for renewable energy of
gy sources. 12% of primary energy needs and of 22% of electric-
ity needs by 2010 and encouraging national targets;
Any proposal for a framework should: increased energy efficiency, inviting the Environment
(a) contribute to the achievement of the national Council to examine setting indicative targets in a
indicative targets; cost-efficient manner and with minimum distortionary
(b) be compatible with the principles of the internal effects; and achieving a final agreement on the emis-
electricity market; sions trading Directive;”
(c) take into account the characteristics of different
sources of renewable energy, together with the 1.7 Concluding Remarks
different technologies, and geographical diffe-
rences; The analysis in this section has focused specifically on
(d) promote the use of renewable energy sources in national mechanisms. National experiences are obviously
an effective way, be simple and, at the same useful when analysing the effectiveness of various mech-
time, as efficient as possible, particularly in anisms, but the analysis cannot be directly translated into
terms of cost; and a Community-wide system. Introducing cross-border trade
(e) include sufficient transitional periods for natio- in the economic analysis of payment mechanisms raises
nal support systems of at least seven years and a myriad of additional questions related to the functioning
maintain investor confidence.” of the support mechanisms. It is beyond the scope of this
report to analyse the effects of international trade on
Presidency Conclusions - renewable energy sources. The two year RE-Xpansion
Brussels, 20 and 21 March 2003 project funded by the European Commission is currently
looking into the European-wide dimension of payment
At its meeting in March 2003, the Council urged member mechanisms.
states to accelerate progress towards meeting the indica-
tive renewable electricity targets (section 1). Ensuring One issue is crucial, however, regardless of which mech-
delivery on the environmental dimension of sustainable anism is considered at a EU-wide level. A precondition for
development, subsection Reversing unsustainable trends: a well-functioning internal market in renewable energy is
“53. Economic and social development will not be that rules and regulations relating to WT investments are
sustainable in the long run without taking action to harmonised. It therefore makes sense to continue the
curb environmental pressures and preserve natural practice followed by the Community ever since the signing
228
resources within the framework of the comprehen- of the Single European Act in 1986. Without harmonisa-
sive sustainable development strategy launched at tion of rules and regulations, e.g. in relation to grid access
conditions, tax treatment, safety standards, etc., the mar-
ket will be distorted.
It is clear from the industry’s experience of various

mechanisms that, although the price paid for electricity
is vital, of equal importance, particularly if project
finance is needed, is consistency of the market, the
creditworthiness of the offtaker and hence the ability to
make a long-term plan. Finally, planning procedures and
fair grid access at reasonable cost is of equal impor-
tance to the development of wind power and other
renewable energy technologies.
Successful countries are characterised by substantial

inward investment by suppliers who see a future market
in that country, and who see the benefits of local supply.
The most successful in this respect have been Spain,
Denmark and Germany, which now have not only a sub-
5
stantial installed wind capacity, but also a substantial
wind energy industry.
VOLUME
229
2 FUTURE MARKETS
2.1 Introduction grow, with more established conventional energy compa-

nies joining the sector in larger numbers, as wind power
This chapter presents two market scenarios; a conven- emerges as a major energy source, not just an environ-
tional and an advanced scenario to the years 2007 and mental “add-on”.
2012. The advanced scenario is, taken from Wind Force
12 - a scenario for a greatly increased market due to high- 2.2 Conventional Scenario
er political and policy support than is envisaged today. For
the status of European Markets see volume 3, chapter 1. Specific market assessments for the short to medium term
future wind energy markets are made by a number of play-
Future market assessments are essentially scenarios that ers within the sector: wind companies and developers; con-
are shaped by the assumptions and data used, by histori- sultancies; members of the financial services community,
cal trends and their extrapolation, and by individual per- and by external institutions such as the IEA and the EU.
spectives, all interacting with a wide range of external fac- Market analyses that are not published are undertaken by
tors. The assumptions behind any future estimates of what a number of consultancies, investors and institutions. This
is a rapidly evolving and changing market are key to deter- assessment by EWEA for the conventional scenario utilises
mining the status and scope of that prediction. inputs from Garrad Hassan and compares them to other
leading market opinions of DEWI and BTM Consult.
Specific factors that define future wind market prospects
include: The conventional scenario takes the following core assump-
• Future demand for electricity generating capacity. tions which, to some degree, could be classified as a
• Future power plant decommissioning. “favourable business-as-usual scenario”. In this, EWEA esti-
• Government political priorities and policies on energy, mates that the current strong development of the wind
electricity, environment and climate change. power market to date will continue as long as commitment
• Continuous and increasing acknowledgement of the to the sector by a number of governments continues to
environmental benefits of renewable energy production strengthen, and that such support is converted into actual
as well as the external costs to society of conventional deployment. But there is no potent policy intervention on
energy production. the scale of that envisaged in the advanced scenario in
• Prospects for emerging new markets to follow the cur- Wind Force 12, which assumes, for example, that the
rent “big three” (Germany, Spain, Denmark). Production Tax Credit in the USA (a tax incentive for invest-
• The economics of wind power itself and its competitive- ing in renewables) will be renewed. A market assessment
ness with other electricity generating technologies. of around five years is generally regarded as being a more
• Prospects for the large amount of wind power capacity accurate forecast. Beyond that, the predictive ability is
currently awaiting final completion and construction. greatly reduced because the defining factors are difficult to
• Evolving status of the leading commercial wind actors, the foresee with any real accuracy.
industry structure, and new entrants and stakeholders.
• Effectiveness and improvements of wind power technol- What is apparent is that future markets are going to be
ogy through R&D and scale. rather different from historical ones. The annual level of
• Detailed understanding of wind resources and their growth of approximately 35% observed over the last four to
exploitation. five years can only be sustained for a limited period, as
such a growth level is only possible from a low starting
Some of the trends and data on the industry’s development point. Eventually, two obstacles will appear: an inability to
are described earlier in this report. The nature of the major produce additional manufacturing resource at the required
commercial players is changing, and it is predicted that the rate; and the inroads which such rapid development make
230
trend towards larger companies will continue. on the general level of demand for new generating capaci-
Some commentators predict the market will continue to ty, conventional or renewable.
2.2.1 CONVENTIONAL SCENARIO RESULTS
Table 2.1: Market Projection by Region - Annually Installed MW
2002 2003 2004 2005 2006 2007
Europe 5,983 6,050 6,300 6,550 6,750 7,000
(Of which EU-15) (5,871) (5,900) (6,100) (6,300) (6,450) (6,600)
North America 450 1,600 1,200 1,500 2,000 2,000
Central &
South America 10 100 200 300 500 600
Asia 411 600 700 800 850 850
Africa 11 25 50 100 150 150
Australia &
New Zealand 119 150 250 250 300 300
Others 17 75 100 100 100 100
Total 7,001 8,600 8,800 9,600 10,650 11,000
Table 2.2: Market Projection by Region - Cumulative Installed MW
2002 2003 2004 2005 2006 2007
Europe
(Of which EU-15)
North America
23,291
(23,056)
4,923
29,341
(28,956)
6,523
35,641
(35,056)
7,723
42,191
(41,356)
9,223
48,941
(47,806)
11,223
55,941
(54,406)
13,223 5
VOLUME
Central &
South America 144 244 444 744 1,244 1,844
Asia 2,610 3,210 3,910 4,710 5,560 6,410
Africa 148 173 223 323 473 623
Australia &
New Zealand 225 375 625 875 1,175 1,475
Others 59 134 234 334 434 534

Total 31,400 40,000 48,800 58,400 69,050 80,050
Table 2.3: Market Projection for the World - Cumulative Installed MW1
2008 2009 2010 2011 2012
World 92,000 105,800 121,670 139,920 160,900
1
Assuming 15% average annual growth in cumulative capacity between 2007 and 2012.
The average annual growth rate in cumulative capacity leading markets will remain Germany and Spain, although
between 2002 and 2007 is 20.6%. important markets in France, the UK, the Netherlands,
Italy and Sweden will emerge.
The European data corresponds to those outlined in the
targets in Chapter 3. Europe continues to dominate, with The market forecast indicates a slight slowing down of the
increased interest in France and the UK predicted, togeth- onshore European market, but an increase in activity in
er with a gentle decline in German onshore activity fol- countries which have not played a major role to date.
231
lowed by an uptake of the offshore segment. Spanish There will be a time gap before the offshore market takes
activity remains dominant but fairly flat. In Europe, the off to replace it. In the meantime, there will be significant
growth in the US. New markets are starting to develop in 2.4 Wind Force 12 – The Advanced
Australia, Japan and South America. There is relatively lit- Scenario
tle installed capacity in these countries and, hence, the
potential for future growth is large. EWEA has published advanced scenarios for the wind
power sector since 1999 - Wind Force 12 (May 2003) is
Other countries that are considering serious investment in its fourth edition. It is the main long-term scenario
include Canada, Brazil, Tunisia, China, Egypt, Morocco, analysis for the wind power sector worldwide and can be
the Philippines, Turkey and Vietnam. accessed in full at www.ewea.org.
2.3 Other Forecasts 2.4.1 METHODOLOGY
BTM Consult in its World Market Update of 2002, fore- The aim of the Wind Force 12 study is to assess the tech-
casts a global market of 83,000 MW by the end of 2007, nical, economic and resource implications for a penetra-
and 177,000 MW by the end of 2012. tion of wind power into the global electricity system equal
to 12% of total future demand in 2020. The intention has
DEWI’s WindEnergy-Studie 2002, an annual survey of sev- been to work out whether a 12% penetration is possible
eral hundred wind companies, estimates that the market within that timescale in terms of technical feasibility,
will be 80,000 MW installed worldwide by the end of industrial ability and resource availability.
2007, and 120,000 MW by the end of 2010.
The methodology used explores the following questions:
Hamburgische Landesbank’s study in July 2002, Wind • Are the world’s wind resources large enough and appro-
Power: Evaluating International Markets for Wind Power priately distributed geographically to achieve a level of
and Wind Power Generator Manufacturers, estimates 12% penetration?
80,800 MW installed worldwide by the end of 2006 and • What level of electricity output will be required and can
144,000 MW by the end of 2011. this be accommodated in the existing grid system?
• Is wind power technology sufficiently developed to meet
Dresdner Kleinwert Wasserstein’s 2001 report Power this challenge? What is its technical and cost profile?
Generation to the 21st Century predicted the global mar- • With the current status of the wind power industry, is it
ket by the end of 2006 to be 67,000 MW. feasible to satisfy a substantially enlarged demand and
what growth rates will be required?
The EU and other international institutions have also The first Wind Force study was carried out by BTM Consult
made future market assessments (see box). for the Danish Forum for Energy and Development (FED) in
1998. This served as the model for a more detailed analy-
Europe in 2010 sis released in 1999 by FED, Greenpeace and EWEA, enti-
European Commission tled Wind Force 10. An update, Wind Force 12, was pub-
White Paper (1997) 40,000 MW lished in 2003 by EWEA and Greenpeace.
EU Energy Outlook for 2020 (1999) 22,600 MW
EU Energy Trends to 2030 (2003) 69,900 MW The 1998 study examined the potential for 10% wind pen-
etration by working with two different scenarios for world
World in 2010 electricity demand. In the more detailed Wind Force 10
IEA World Energy Outlook reference report (1999) only one parameter of future electricity
scenario (2002) 55,000 MW demand was used - the IEA’s 1998 “World Energy
IEA forecast (2003) Renewables for Outlook”, a projection which assumes “business as
232
Power Generation, Status and Prospects 120,000 MW usual” and in which electricity consumption is predicted to
double by 2020.
2.4.2 REGIONAL UPDATE OF WIND FORCE 12 Table 2.5: Summary of Conventional and Advanced Market
ADVANCED SCENARIO Scenarios (Cumulative Installed MW)
Table 2.4 provides an update from the Wind Force 12 sce- Year 2007 2012
nario, split into regions. In this assessment, a breakdown by

region for the years 2007 and 2012 has been carried out - ADVANCED scenario - Europe 59,000 112,000
taking into account the initial status (2002 figures). It shows Average annual growth rate 20% 14%
that a few regions, namely Europe and the US, are on track ADVANCED scenario - world 106,000 311,000
to achieving their potential, while others have not yet left the Average annual growth rate 27% 24%
starting blocks. To calculate the cumulative figures for each CONVENTIONAL scenario - Europe 55,941 -
continent, a certain average annual growth rate of annual CONVENTIONAL scenario - world 80,050 160,900
installation is chosen and later turned into a growth rate for Average annual growth rate 20.6% 15%
cumulative installation. Typically, this results in a high value

for new and emerging regions and a more modest value for
The average annual growth rates are for the periods 2003 - 2007, and 2008 - 2012.
regions already on track, i.e. OECD Europe.
Table 2.4: Wind Force 12 Scenario Update - Breakdown by Region in 2007 and 2012
Region 20021 Average Growth New Annual Cumulative Average Growth of New Annual Cumulative Distribution of
Installed Capacity of Cumulative
MW
Installation 2003 - 07
%
Capacity
In 2007
MW/yr
Capacity
in 2007
MW
Cumulative
Installation 2008-12
%
Capacity
2012
MW/yr
Capacity
in 2012
MW
the Total World’s
Wind Power Capacity
By 2012
% 5
VOLUME
OECD Europe 23,832* 20% 7,900 59,000 14% 10,750 112,000 36.0 %
US & Canada 4,944 29% 5,200 17,500 30% 14,090 65,000 20.9 %
Latin America 143 74% 1,420 2,300 51% 4,560 18,000 5.8 %
OECD Pacific 730 45% 1,615 4,700 24% 4,630 13,500 4.3 %
East Asia 0 - 410 1,200 38% 2,225 6,000 1.9 %
South Asia 1,714 39% 1,470 9,000 23% 4,560 25,000 8.0 %

China 473 76% 2,270 8,000 37% 7,600 38,000 12.2 %
Middle East 32 90% 270 800 30% 960 3,000 1.0 %
Transition
Economies 22 257% 1,185 2,500 59% 6,490 25,500 8.2 %
Africa 148 47% 305 1,000 38% 1,315 5,000 1.6 %
Total World
Capacity WF12
Scenario
(figures
rounded) 32,038 27% 22,000 106,000 24% 57,000 311,000 100 %
* Source: EWEA (2003c).

1
2002 Figures sourced from BTM Consult (2003).
233
2.5 Overview of Non European Markets that any short to medium term large-scale exploitation is
unlikely. Market commentators do not, therefore, see a
2.5.1 NORTH AMERICA lot of commercial activity in Argentina in the near future.
In the longer term, Argentina does seem to be a good
Canada candidate for substantial investment, however.
Late 2001 and early 2002 saw a major change in the Mexico
Canadian market. Hitherto, apart from one isolated devel-
opment of a 100 MW project in Quebec, the Canadian Mexico has continued to surprise the wind energy indus-
market has been dormant. The Canadian Wind Energy try by its lack of development. A large-scale wind energy
Association has announced a very ambitious plan to resource investigation has been conducted, and there is
achieve 10 GW of installed wind power capacity by the a need for new generating capacity. There have been
year 2010. Commentators consider that this is highly several calls for tender for significant development, but
unlikely under the current legal framework, yet a concrete none have yet been realised. No new capacity was
proposal for a tax credit, rather similar in nature to that installed in Mexico in 2002. However, 2003 has seen a
in the US, has resulted in a high level of interest in sev- significant change in mood, and it is expected that there
eral states. It has also reawakened interest in the wind will be Mexican development in the short term, and
energy business by many Canadian utilities. future development will be significant thereafter. Mexico
has some areas of wind energy resource which are
Space is clearly not a limitation in the Canadian market, exceptional, and these are likely to be the first to be
but the availability of grid connection for large-scale proj- exploited. It now seems that there is interest of suffi-
ects is likely to pose challenges. Wind maps have been cient substance and local influence to allow develop-
produced, or are under production, for several of the ments to take place.
Canadian provinces, and the authorities in Newfoundland,
Saskatchewan, British Columbia and Quebec have Brazil
expressed interest during the 2003. Little development
took place in Canada during 2002, but there are signs In 2001, Brazil suffered a severe water shortage and,
that activity will accelerate. Expressions of interest have since its electricity industry is dominated by hydroelec-
been sent to NRCan for 1,220 MW for commissioning in tric power, this translated directly into a shortage of
2003; 1,500 MW for 2004; and 600 MW for 2005. A sub- electricity. As a result, a Ministry of Crisis was created;
stantial failure rate and delay must be expected, but part of its responsibilities was to investigate new meth-
these are nevertheless sizeable numbers. ods of procuring additional electrical capacity on a short
time-scale. A new law to encourage renewable energy,
2.5.2 CENTRAL AND SOUTH AMERICA and wind power in particular, was enacted: the Pro
Eolica Law called for the immediate construction of
Argentina 1,050 MW of wind energy in Brazil. Substantial interest
was shown, with some 3,000 MW proposed. Ironically,
Argentina has long been considered a promising market for since the level of interest was more substantial than the
wind energy, since it has a prodigious resource in terms Brazilian government had been expecting, and conse-
of both wind speed and space. Particular interest is quently there was no mechanism in place to differentiate
often expressed in the large, open expanses of between the different proposals, the over-subscription
Patagonia. Several European companies have articulat- resulted in delay. At the same time, the level of hydro
ed detailed plans to exploit this resource, and some resources returned to a more normal level. Neither of
234
modest level of installation has taken place. However, these events bodes well for the immediate development
the recent economic difficulties in Argentina suggest of large-scale wind energy in Brazil.
The financial arrangements for payment under the Pro Development of a 24 MW wind farm by the Colombian
Eolica Law are not yet clear, and it is possible that some utility EPM was planned for 2003.
new bidding mechanism or other form of competition
may be invoked in order to determine where the con- 2.5.3 AUSTRALIA AND NEW ZEALAND
tracts will be placed. The Brazilian electricity industry
has recently been privatised, and many of the large ex- Australia
utilities have been purchased by European companies.
As a result, it is possible, even in the absence of a res- The recent review of Mandatory Renewable Energy Target
olution of the Pro Eolica difficulties discussed above, (MRET) which places a legal obligation on wholesale pur-
that some private, but nevertheless large, wind farm chasers of electricity to contribute towards the genera-
developments will proceed. This is possible by virtue of tion of an additional 9,500 GWh of renewable electricity
the use of direct Purchasing Power Agreements (PPAs) annually by 2010 has caused some uncertainty but also
between wind farm developers and utilities who have a sudden and quite substantial increase in activity in
some common ownership. Very recent changes again Australia. To date, the level of installation is modest, but
appear to suggest that large-scale development may be the planned activity is significant – of several thousand
about to start. It is necessary, however, to view these MW. At the time of writing, the obligation was set at 2%
possibilities with a cynical eye since there have been of 1999’s total electricity production to come from renew-
many false starts for wind energy in Brazil. ables or specified waste-product energy sources by
5
2010. Australia’s electricity consumption has risen quite
Costa Rica substantially during the last few years and, hence, this
level now equates more closely to 0.5% of total electrici-
Costa Rica already has a high level of wind power pene- ty production. The current target is estimated to result in
VOLUME
tration into its relatively limited electrical grid. the construction of 900 MW by 2010.
Nevertheless, the Costa Rican state utility has further
plans for additional implementation of wind energy proj- A MRET review report was released in January 2004 and
ects. Continued growth is expected in wind energy instal- will pass through the political process during 2004.
lation in this country. Currently, the dominant players are the incumbent gener-
ators, but many private developers are also present in

Panama the market, and it is particularly interesting to note that
some European developers who are expecting a down-
Panama has recently completed the development of a turn in the onshore German market, have chosen to
wind resource map and, whilst specific plans to develop make investments.
projects are not yet available, it seems reasonable to
assume that some level of modest development will The biggest hindrance to development in Australia is the
take place in the future. lack of substantial grid connections. The electricity grid
follows the centre of populations along the coastal areas
Chile, Columbia, Peru, Venezuela of the South East, with separate grids in the different
states. Western Australia and the Northern Territory are
These countries have all expressed interest in wind completely disconnected from the south. There is no
energy projects, and plans and proposals in various clear mechanism for resolving this issue at present.
forms have been discussed. To date, there has been lit-
tle activity, but it is expected that, as developments Local benefits, in terms of indigenous manufacturing, are
take place in the countries surrounding them in the going to be a key consideration, and several states are
235
region, some modest activities will appear. in the late stages of negotiation with wind turbine manu-
facturers. In order to gain a permit for the construction of
a wind farm in Australia, it will be necessary to demon- for other countries which suffer from power shortages and
strate local benefit. “Local”, meaning the state rather financially weak utilities to use India as a model e.g. some
than the country. of the micro-financing models in place could serve as good
inspiration for developing countries.
Some commentators see Australia as “the new Spain”.
One market commentator considers that the Australian A notable characteristic of the previous Indian develop-
market is going to be substantial within the next few years. ment was the fact that, by the end of the Indian “wind
It has wind, it has space, it has an excellent industrial rush” in the late 1990s, some 70% of the turbines
infrastructure and a clear code of business ethics. It does, installed were manufactured in India. This high level of indi-
however, require some amendments to its electricity and genisation is an important characteristic of the wind indus-
renewable energy legislation. try as a whole, and should be given due weight when con-
sidering new export markets. Most of the turbines manu-
New Zealand factured in this way were made under licence from
European or American vendors. A significant result of this
There has been little activity in New Zealand in recent approach has been the emergence of Suzlon Energy as a
years. Nevertheless, there are some projects at the plan- manufacturer in its own right. Suzlon purchased the
ning stage, and there is significant resource assessment German company Sudwind, and is now developing its own
activity in progress. The early promise of New Zealand as turbine technology and has a research centre in Germany
a new market, indicated by the construction of the 40 MW and a blade factory in the Netherlands. It is presently oper-
Tararua project, has not been realised, due largely to the ating its prototype 1 MW turbine in California and is devel-
continued weakening of the New Zealand dollar, the further oping a 2 MW successor. Suzlon turbines are now being
reorganisation of the electricity industry, and a continued offered for sale outside India and Suzlon itself is operating
drop in the electricity price. The New Zealand market is as a developer in other countries. This step is significant,
expected to continue to be flat for some years. not only for India, but for the industry as a whole: a con-
tinuation of the relatively modest growth over the last two
2.5.4 ASIA years with a slight increase over time is likely.
India Japan
In the mid- to late-1990s there was a great deal of activity Historically, the Japanese market has not been active,
in India, driven by a combination of energy shortages and despite the fact that Japan boasts not only a major manu-
tax credits. Unfortunately, the market mechanisms gave facturer in the form of MHI, but also several of the leading
particular benefit to the installation of wind farms, rather wind farm developers - Tomen Power Corporation (Eurus),
than to their efficient operation. The result was a series of Marubeni and Nichimen. This imbalance has been
irresponsible developments, with poor engineering and redressed in the last two years as a result of some attrac-
inadequate resource assessments in advance of con- tive incentives, both in terms of kWh price and capital
struction. That spate of activity was followed by several grants. The Purchasing Power Agreements (PPAs) have a
years of minimal activity as a result of the general eco- long-term lifespan of 17 years which creates investor con-
nomic crisis in Asia but, recently, the Indian market has fidence. The wind speeds are relatively low, and the cost
restarted in a much more controlled fashion. Market com- of construction in Japan seems high by world standards.
mentators expect this approach to continue and, hence, Therefore, some caution must be exercised in considering
growth in the Indian market to continue at this level. It is tariffs in a more general context. Japan is relatively dense-
considered beneficial to analyse the way in which the ly populated in areas where construction is feasible. It is
236
Indian market has changed and to consider the new mech- considered that some of the Japanese terrain will compli-
anisms which have been put in place. It may be sensible cate large-scale developments. It is, therefore, noted that
the Japanese are also considering the possibility of off- In Tunisia, matters are progressing slowly but steadily
shore development, but there are some limitations and, although there is little activity to date, it is likely that
because of water depth and the occurence of typhoons. the future will see significant installation in Tunisia with-
Commentators expect the level of installation in Japan to in the next two years. The geographical proximity of Italy
increase from that experienced over the last few years, but may, in itself, have an effect, and the presence of the
not to rise dramatically due to space limitations and a Vestas manufacturing plant in Taranto will be useful.
reluctance on the part of the Japanese utilities to accept
wind energy projects under current conditions. Libya has recently called for assessments to be made of its
wind resource and for the establishment of a technology
China centre for the development of wind energy. No concrete plans
are known to date but it is thought that some will appear in
The development of the Chinese market has been some- a foreseeable future. The political situation, however, makes
what disappointing. Several years ago there was a lot of Libya a difficult market, at least in the short to medium term.
activity aimed at developing both a domestic Chinese
manufacturing industry, and various aid-assisted projects Egypt has benefited for a long period from various over-
which were intended to lead to a substantial uptake of seas aid-related projects from different European coun-
wind energy in various of the Chinese states. Many of the tries and, hence, does have a base of installed capacity.
aid-assisted projects now appear to have foundered and, Together with the World Bank, these aid agencies are con-
5
hence, the rate of uptake expected is rather slower than tinuing to show an interest in Egypt. The wind resource
was previously anticipated. The construction of conven- along the Nile is substantial and is now reasonably well
tional power stations in China, using predominantly brown documented, but there is still no sign of any real com-
coal, has slowed slightly. This is also likely to have the mercial development of this market.
VOLUME
effect of reducing the uptake of wind. It is considered that
the level of introduction of wind energy in China will depend Mild growth is expected in this block of countries over
very strongly on the level of licence agreements or indige- the next five years.
nous manufacture which can be achieved. Several Chinese
companies are developing indigenous technology, to date The Rest of Africa
with limited success.

The vast majority of the African continent does not have
2.5.5 AFRICA sufficient wind to merit any serious investigation for
large-scale installations. The situation may be rather dif-
North Africa ferent for local, small-scale generation. There should,
nevertheless, be a promising market in South Africa,
All of the countries along the southern coast of the since it is well away from the equatorial belt, where
Mediterranean, with the exception of Algeria, are express- winds are limited. Interest has been shown in South
ing a keen interest in wind energy. Plans have developed Africa but, until the South African utility, ESKOM,
in Morocco, Tunisia, Libya and Egypt. There has been becomes convinced of wind power’s efficacy, and a polit-
much discussion in Morocco over the last two to three ical decision is taken to develop it, it seems unlikely that
years of developments of the order of 100 MW. The ten- there will be substantial development there. Wind maps
dering process has been very protracted, however, and is have been produced and detailed discussions have
still not resolved. It is, nevertheless, acknowledged that taken place, but there is still little real sign of activity at
there is a very substantial resource available, with excep- a larger scale. Notwithstanding the comments made
tional wind speeds at some sites. However, the incumbent above, recent expressions of interest have been shown
237
utility’s lack of willingness to develop these projects is a from Kenya and, perhaps in the longer term, this coun-
major cause for delay. try may see some modest developments.
2.5.6 THE EAST
Turkey
Turkey is considered to be a very substantial long-term

market. It has a major shortage of energy, substantial
space, a reasonably good electrical infrastructure and a
very good wind resource. Turkish industrial companies
with substantial financial muscle have invested in a
large-scale resource assessment, and some 2,000 MW
of potential development exists. A major hurdle has
been the country risk, in particular political and currency
instability. During 2001, this resulted in the virtual ces-
sation of Turkish wind development activities. However,
new legislation is being considered and, it is expected
that projects will eventually be completed and, hence, a
market is anticipated, although not for several years.
The Middle East
Syria, Jordan and Iran have all shown interest in wind

energy. The promise of some of the early development,
particularly in Iran, has not been followed through with
further projects and, given the political situation in these
countries, a substantial long-term market is not envis-
aged at present.
The Far East

Interest in wind energy is beginning to emerge in

Malaysia, Indonesia, the Philippines, Vietnam and South
Korea. There are some concrete projects being planned
in all these countries, with some modest activity expect-
ed. Many of these countries suffer from a combination
of relatively low mean wind speeds and very high
extreme wind speeds which, therefore, require a major
capital investment for a relatively limited return in ener-
gy. Consequently, these countries are not expected to
have a substantial market in the near future, although
Korea, for example, has repeatedly expressed an inter-
est in establishing an indigenous industry and this
appears to be under way.
238
3 NEW TARGETS
3.1 EWEA Targets - Onshore and Figure 3.2: EWEA Targets for 2010 (MW Installed in EU)
Offshore
EWEA targets for wind energy installed capacity in the

EU-15 are as follows:
• In 2010 - 65,000 MW onshore, and 10,000 MW
offshore: 75,000 MW in total
• In 2020 - 110,000 MW onshore, and 70,000 MW
offshore: 180,000 MW in total
In 1997, EWEA adopted the target set out in the European

Commission’s White Paper on Renewable Sources of
Energy of 40,000 MW by 2010 (Figure 3.2). Three years
later, EWEA revised its target to 60,000 MW by 2010
including 5,000 MW offshore; and 150,000 MW by 2020
5
including 50,000 MW offshore (see figures 3.2 and 3.3).
These figures were then revised to the above in 2003.
Figure 3.1: EWEA Targets for 2000 (MW Installed in EU) Figure 3.3: EWEA Targets for 2020 (MW Installed in EU)
VOLUME
239
In 2003, EWEA published Wind Force 12, which demon- 3.1.2 EUROPEAN COMMISSION – NEW
strates that wind power is capable of supplying 12% of TARGETS
the world's electricity within two decades, even if the
overall electricity demand increases by two-thirds in that The European Commission has dramatically increased its
period, given increased political will. The study is not a projections for wind power installed by 2010. In its recent-
long-term forecast or a prediction, but rather a feasibili- ly published report European Energy and Transport, the
ty study for future scenarios taking into account the Commission stretched its previous forecast by more than
physical limitations for large-scale development of wind 200%.
power. It assesses and compares actual industrial
growth patterns seen in the wind power sector so far Its earlier prediction (based on the energy model PRIMES)
with hydro and nuclear power development. for installed wind power capacity in the EU-15 was 22.6
GW by 2010, a level already reached in 2002. The
In this 12% global scenario, Europe (defined more broad- Commission now predicts a total of 69.9 GW of wind
ly as OECD Europe, under IEA classifications) will install power capacity to be installed in the EU by 2010. Thus,
100 GW by 2010 and 230 GW by 2020, producing 564 the new Commission estimate is more in line with the
TWh/year and saving 338 million tonnes of CO2 emis- EWEA target of 75 GW by 2010.
sions per year. Equally important, it is envisaged that
around 1,000 GW of wind power will be installed in non- However, the Commission’s energy model is exceptional-
European countries during the same period. ly pessimistic regarding the long-term forecast for wind
power beyond 2010. For the two decades from 2010 to
BTM Consult (2003) in its recent World Market Update 2030 PRIMES predicts a combined net increase in capac-
2002 predicts that the cumulative capacity in Europe will ity that will be less than the net increase in the current
reach 58 GW by the end of 2007 and 108 GW by 2012. decade (see Table 3.1).
3.1.1 EUROPEAN COMMISSION - HISTORICAL Table 3.1: Primes - Installed Generation Capacity by Plant
TARGETS Type in EU (GW)
1995 2000 2010 2020 2030

In 1997, the European Commission’s White Paper on
Nuclear 126.2 131 121.9 100.1 105

Renewable Sources of Energy set the goal of doubling
Large Hydro 85.1 87.7 88.9 88.9 89.2
the share of renewable energy in the EU from 6% to 12%
Small Hydro 2 2.1 8.1 12.2 14.5
by 2010. One of the White Paper targets was to
Wind 2.5 12.8 69.9 94.8 120.2
increase EU renewable electricity production from 337
Other RES 0 0.2 0.5 0.6 14
TWh in 1995 to 675 TWh in 2010. Within this target,
Thermal Plant 322.9 344.8 399.5 516.1 608.1
the goal for wind power was for 40,000 MW (40 GW) of
installed capacity in 2010, which could produce 80 TWh Total 539 579 689 813 951
of electricity.
One of the shortcomings of the model is that it does not
The subsequent Directive (2001/77/EC) on the promo- consider technological change. The new Commission esti-
tion of electricity from renewable energy sources sets mates for wind power in 2010 have been dramatically
national indicative targets for the contribution of electrici- revised to better reflect current reality and future industry
ty from renewable energy sources as a percentage of projections. However, the projections for the period 2010
gross electricity consumption. The overall Community goal to 2030 remain unrealistically low. As wind power
is to increase renewables’ share of electricity from 14% in becomes cheaper, electricity demand is likely to increase
240
1997 to 22% in 2010. by some 1.5% per year in the coming decades, and many
conventional power plants will be decommissioned in the
coming 25 years. Furthermore, wind power will become In order to estimate the electricity production from wind
increasingly competitive. power, and the corresponding CO2 emission reductions,
during the period 2001 - 2010, a yearly projection devel-
3.2 Increasing Wind Power Targets opment has been performed. The average capacity factor
for Europe (see glossary) of WTs is assumed to increase from 0.25
in 2001 to 0.28 in 2010. Over the past two decades,
Figures 3.4 and 3.5 detail the new industry targets which capacity factors have improved as a result of both better
show that wind installation will continue to increase, but initial design and better siting.
at a lower rate. The high growth of the last few years has
been based mainly on the German and Spanish markets. The major contributions to improved capacity factors have
Market forecasts for coming years indicate that annual been the increased hub height above ground level of the
installations will stabilise in Spain and will decrease in larger turbines and technological improvements from R&D
Germany. A 3.5% increase in annual installations is activities. It is worth noting that, for a technology that
assumed for the year 2003, decreasing gradually to 1.4% utilises a free resource, a high capacity factor is not a
in 2010. This corresponds to a 25.7% increase in total goal in itself. Improving the capacity factor of WTs pres-
installed capacity in 2003 gradually decreasing to 10.3% ents no technical problems, but it does affect grid inte-
in the year 2010. gration, modelling and generation costs.
5
Figure 3.4: Projections of Annual Installations (2003 – 2010) in the EU-15
VOLUME
241
Figure 3.5: Wind Power Target Projections up to 2010 (MW) in the EU-15
242
3.3 Targets for the EU-15 in 2010
Table 3.2: Installed Wind Power Capacities by Member State (MW)
1996 1997 1998 1999 2000 2001 2002 2010
Austria 10 20 30 34 77 94 140 500
Belgium 4 4 6 6 13 31 35 250
Denmark 842 1,129 1,443 1,771 2,417 2,489 2,889 5,000
Finland 7 12 17 39 39 39 43 500
France 6 10 19 25 66 78 148 6,000
Germany 1,552 2,081 2,875 4,442 6,113 8,754 11,994 28,000
Greece 29 29 39 112 189 272 297 2,000
Ireland 11 53 73 74 118 125 137 1,500
Italy 70 103 180 277 427 697 788 3,700
Luxembourg 2 2 9 10 10 15 17 50
Netherlands 299 319 361 433 446 493 693 2,500
Portugal 19 38 60 61 100 125 195 1,500
5
Spain 249 512 834 1,812 2,235 3,337 4,825 15,000
Sweden 103 122 174 220 231 290 345 2,500
United Kingdom 273 319 333 362 406 474 552 6,000
VOLUME
European Union 3,476 4,753 6,453 9,688 12,886 17,313 23,098 75,000
Given the current distribution of wind power in EU-15

countries, historical rates of growth, the wind potential of
each country, and the current status of the wind-related

policies and targets of each, a possible distribution of the
total installed capacity for each member state in 2010 is
shown in Table 3.2.
In this table, the projected capacities in 2010 are shown,

together with the installed capacities of the last seven
years for each EU member state. Over the period 2001 -
2010, Germany, Spain, France and the UK comprise 74%
of the expected total capacity installed. The increase for
a certain number of countries, like the Netherlands,
Denmark and the UK is based on the foreseen rapid
development of offshore wind during the second half of
the decade. Some, like the UK, France, Ireland and
Greece have the potential to increase their projected
installed capacity substantially if the framework condi-
243
tions become more favourable for renewables and if sev-
eral existing barriers are removed.
3.3.1 HOW MUCH ELECTRICITY WILL THIS household electricity consumption will increase by 16%
PROVIDE? by 2010 and by 30% by 2020.
The electricity output from the EWEA targets can be The number of people per household will decline by 2020.
expressed in terms of the equivalent amount of house- Therefore, the total amount of households or individuals
hold electricity consumed by the average individual or whose average electricity use is provided by the wind
household in Europe, as in Table 3.3. The calculations power targets will become progressively less as some of
are based on data and forecasts from Eurostat and the additional wind power targets output is used for the
Eurelectric, and the European Commission’s Energy additional consumption patterns. In 2020, for example,
Outlook to 2020 report. These forecasts assume that wind power will generate 425 TWh; this is 50% of the fore-
whilst population and number of households will cast EU household electricity consumption in 2020, but
increase by only a small amount, average 66% of that consumed in 2001.
Table 3.3: Household Electricity Consumption and Supply Forecast
Country Unit 2001 2005 2010 2020
Austria TWh 15.00 16.10 17.50 18.70

Belgium TWh 17.10 18.10 19.50 21.90
Denmark TWh 9.60 9.50 9.90 10.40
Finland TWh 19.40 21.50 22.50 23.70
France TWh 129.74 143.42 157.10 178.20
Germany TWh 131.00 132.00 132.00 132.00
Greece TWh 14.50 17.10 20.80 28.40
Ireland TWh 7.40 8.40 10.00 12.60
Italy TWh 61.60 65.80 74.20 87.70
Luxembourg TWh 0.80 0.80 0.90 1.00
Netherlands TWh 22.90 25.30 28.90 34.50
Portugal TWh 10.60 11.70 13.60 16.90
Spain TWh 50.60 60.10 72.60 94.90
Sweden TWh 43.10 42.40 43.70 44.20

United Kingdom TWh 115.30 122.50 127.20 136.80
Total EU-15 Consumption TWh 648 694 750 841

Total EU-15 Consumption Million KWh 648,640 694,720 750,400 841,900
Total EU-15 Population Million 378 381 384 386

Average Household Size 2.50 2.50 2.50 2.30
EU-15 Number of Households Million 151 152 153 167
Average Household Electricity Consumption KWh 4,284 4,558 4,885 5,016

Average household Electricity Consumption KWh 1,713 1,823 1,954 2,181
per person
Wind Power Production TWh 32.40 86.50 167.40 425.00
Households Supplied by Wind Power Million 7.56 18.98 34.27 84.72
People Supplied by Wind Power Million 18.91 47.44 85.66 194.86
Households/People Supplied by Wind Power % 5.00 12.45 22.31 50.48
244
Figure 3.6: Equivalent Electricity Needs Met by Wind Power 2001 - 2020 (Million People)
3.3.2 WHAT PROPORTION OF TOTAL EU Assuming that wind power is substituting intermediate
5
VOLUME
ELECTRICITY FROM WIND? loads covered by fossil fuels (gas, oil and coal) with the
average efficiency foreseen by the IEA study, the total
According to the IEA’s World Energy Investment Outlook installed generation capacity of 445 GW will be increased
2002, consumption of electricity is expected to increase by by 63.7 GW in the period 2001 - 2020 due to the lower
1.6% per year over the period 2001 - 2020 (International capacity factor of wind.
Energy Agency, 2003). With this assumption, total electrici-

ty demand in the EU will increase from 2,572 TWh in 2000 In 2000, wind power represented 2.1% of the total EU gen-
to 3,064 TWh in 2010 and to 3,511 TWh in 2020. erating capacity. This share will increase to 10.6% in 2010
and 21% in 2020.
The total share of the EU’s electricity consumption that is
generated by wind power will be 5.5% in 2010 and 12.1% 3.3.3 WHAT SHARE WILL WIND HAVE OF
in 2020. TOTAL NEW CAPACITY INSTALLED?
The IEA study estimates that the installed power capacity The leading role that wind power will play in the power gen-
requirements are expected to increase by some 210 GW erating system of the EU in the coming two decades is even
during this period and, additionally, that approximately 235 more evident when considering its share of new generating
GW of new capacity will be required to replace decommis- capacity expected to be installed in Europe in the first two
sioned plants. Thus, the EU is projected to build approxi- decades of this century.
mately 445 GW of new plants over the 2001 - 2020 period.
Wind power can cover a substantial part of this new capac- In the period 1995 - 2000, wind power accounted for 23.4%
ity. As mentioned above, the contribution of wind power is of net increase in generating capacity across the EU. During
245
underestimated in the IEA scenario. If we assume that the the period 2001 - 2010, wind power will account for 50% of
wind industry targets will be met, wind will then substitute net increase installed generating capacity and 70.3% for
other conventional energies foreseen in the IEA scenario. the period 2011 - 2020 (Figure 3.8).
Figure 3.7: Contribution of Wind Power (GW) to EU Electricity Generation Capacity 1995 – 2020 (%)
1995 2000 2010 2020
wind 0.46% wind 2.1% wind 10.6% wind 21%
1995 2000 2010 2020
Total Generation
Capacity (GW) 538.8 580.7 706 855.4
Total Wind Generation

Capacity (GW) 2.5 12.3 75 180
Share of Wind (%) 0.46% 2.1% 10.6% 21%
Figure 3.8: Contribution of Wind Power to New EU Generation Capacity (GW)
1995 - 2000 2001 - 2010 2011 - 2020
wind 13.6% wind 27.8% wind 36.9%

1995 - 2000 2001 - 2010 2011 - 2020
Additional Generation
Capacity (GW) 41.9 125.3 149.4
New Wind
Capacity (GW) 9.8 62.7 105
Share of Wind (%) 23.4% 50% 70.3%
Capacity Retirement 30 100 135

(GW)
Total New Capacity 71.9 225.3 284
Construction (GW)
Share of Wind (%) 13.6% 27.8% 36.9%

246
Of total new construction of electricity generating capaci- 3.4.1 OECD ALTERNATIVE POLICY SCENARIO
ty, including capacity replacement of decommissioned
plants, wind power will represent 27.8% during the period It is important to point out that the alternative IEA sce-
2001 - 2010 and 36.9% for the period 2011 - 2020. nario includes increased policy support for renewables.
The need for such is voiced in the EWEA feasibility study
3.4 International Energy Agency Wind Force 12, which states that the wind industry is suf-
Scenarios ficiently advanced to be able to supply 12% of the world’s
electricity by 2020, but increased political will and policy
The IEA estimations in the “reference scenario” present- support will be required if it is to do so.
ed in the recent World Energy Investment Outlook 2003
report, are very conservative and do not reflect current While the reference scenario only includes policies
trends in the market: 33 GW in 2010, 57 GW in 2020 and already in place by mid 2002, the IEA’s alternative policy
71 GW in 2030. Given the current rate of installation in scenario visualises electricity consumption and use under
the EU (almost 6 GW in 2002) and the actual growth rates the influence of more aggressive policy measures, princi-
during the previous years, the IEA’s reference scenario pally aimed at CO2 abatement through increased use of
estimations would mean a complete reversal of this trend renewables, among other measures. Under the alterna-
in the next few years resulting in continuously decreasing tive policy scenario, emissions of CO2 would fall to 2000
rates of installation. Even the predictions in the “alterna- levels by 2030.
5
tive policy scenario” are very conservative (see below).
The policy measures in the alternative scenario directly
In World Energy Investment Outlook 2003 the IEA visualis- related to renewables include the EU’s renewable energy
es an increase in electricity demand of around 50% up to Directive, the renewable portfolio standard in the US and
VOLUME
2030. This would require an additional 650 GW of capac- Canada, and renewable energy targets in Japan, Australia
ity, and the replacement of about 330 GW of existing and New Zealand. Under the influence of these supportive
capacity. The agency projects that more than half of the measures, renewable energy technologies feature much
new capacity installation will take the form of gas fired more prominently.
plant, and 20% in the form of renewable energy technolo-
gies, excluding hydro, with the emphasis on wind and bio- The policies under consideration in the alternative sce-

mass. Renewables, according to the reference scenario of nario are projected to achieve a 25% share across the
the study, will capture approximately one-third of invest- OECD of renewable generation by 2030 compared to
ment in new power plants in OECD countries. 17% under the reference scenario above.
Table 3.4: Reference Scenario for EU-15 The support mechanisms in use in this scenario do not
constitute an exhaustive list of the measures available to
Generating Capacity (GW)
policy-makers: for more information, please consult chap-
2000 2010 2020 2030
ter 1 in this volume, on policy support mechanisms.
Coal 146 134 122 136
Oil 78 77 55 33
Gas 98 176 310 372
Hydrogen fuel cell 0 0 1 30
Nuclear 124 118 88 76
Hydro 118 124 129 134 Endnotes

Other RES (inc. wind) 19 50 87 120 1
Directive 2003/96/EC of 27 October 2003.
2
Source: European Commission, European Energy & Transport Trends to 2030.
Total 584 679 792 901
3
Wiser, Bolinger and Holt, Customer Choice and Green Power Marketing: A Critical
247
Review and Analysis of Experience to Date (University of California: Lawrence
Berkeley National Laboratory).
Source: IEA (2003).
APPENDICES
APPENDICES - TABLE OF CONTENTS
A ONSHORE WIND MAPS 253
B OFFSHORE WIND MAPS 262
C WORKED EXAMPLE FOR CUILLIAGH MOUNTAIN WIND FARM, IRELAND 268
D DETAILED DESCRIPTION OF CORRELATION TECHNIQUES 281
E CONNECTION CONSIDERATIONS 283
F POSSIBLE GRID CODE REQUIREMENTS 287
G CALCULATION OF SPECIFIC EMISSIONS OF STANDARD AIR POLLUTANTS FROM

FOSSIL FUEL ELECTRICITY GENERATION (METHODOLOGY) 290
H ASSUMPTIONS FOR THE CALCULATIONS OF EXTERNAL COSTS WITH ECOSENSE 292
I MARGINAL COST OF CO2 EMISSIONS 293
J CALCULATED EMPLOYMENT IN EU LEVEL SECTORS 294
K EASTERN PROMISE, WIND DIRECTIONS, EWEA, MARCH 2003 295
L EWEA MEMBERS 301
M DIRECTIVE 2001/77/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

OF 27 SEPTEMBER 2001 ON THE PROMOTION OF ELECTRICITY PRODUCED FROM
RENEWABLE ENERGY SOURCES IN THE INTERNAL ELECTRICITY MARKET 305
APPENDICES
251
APPENDIX A: ONSHORE WIND MAPS
Figure A.1: European Wind Atlas, Onshore (EU-12). Source: Risø National Laboratory.
APPENDIX
253
Figure A.2: Denmark Wind Atlas. Source: Risø (1999).
254
Figure A.3: German Wind Atlas. Source: Deutscher Wetterdienst.
Figure A.4: Finland Wind Atlas. Source: FMI/Energy Group (1991).
APPENDIX
Sheltered Terrain Open Plain At a Sea Coast Open Sea Hills and Ridges
m/s W/m2 m/s W/m2 m/s W/m2 m/s W/m2 m/s W/m2
> 6.0 > 250 > 7.5 > 500 > 6.0 > 700 > 9.0 > 800 > 11,5 > 1800
5.0-6.0 150-250 6.5-7.5 300-500 7.0-8.5 400-700 8.0-9.0 600-800 10.0-11.5 1200-1800
4.5-5.0 100-150 5.5-6.5 200-300 6.0-7.0 250-400 7.0-8.0 400-600 8.5-10.0 700-1200
3.5-4.5 50-100 4.5-5.5 100-200 5.0-6.0 150-250 5.5-7.0 200-400 7.0-8.5 400-700 255
< 3.5 < 50 < 4.5 < 100 < 5.0 < 150 < 5.5 < 200 < 7.0 < 400
Wind resources at 50 meters above ground level for five different topographic conditions
Figure A.5: Greece Wind Atlas. Source: CRES (2001).
256
Figure A.6: Ireland Wind Atlas. Source: True Wind Solutions (2003).
APPENDIX
257
Figure A.7: Sweden Wind Atlas. Source: SMHI, Vindatlas för Sverige (1992).
258
Figure A.8: UK Wind Atlas. Source: Burch & Ravenscroft (1992).
APPENDIX
259
Figure A.9: Central European Wind Atlas. Source: Dobesch & Kury (1997).
Czech Republic
Slovakia
Austria
Slovenia Hungary
Croatia
260
Figure A.10: Armenia Wind Atlas. Source: Elliott et al., NREL (2003).
APPENDIX
261
APPENDIX B: OFFSHORE WIND MAPS
Figure B.1: European Wind Atlas, Offshore. Source: Risø National Laboratory.
262
OFFSHORE WIND MAPS MODELLED IN “STUDY OF OFFSHORE WIND ENERGY IN THE EU”
(GARRAD HASSAN ET AL., 1995)
Figure B.2: Denmark - Germany
Figure B.3: France - Atlantic
APPENDIX
263
Figure B.4: France - Mediterranean
Figure B.5: Great Britain - North

264
Figure B.6: Great Britain - South
Figure B.7: Greece
APPENDIX
265
Figure B.8: Ireland
Figure B.9: Italy

266
Figure B.10: Netherlands - Belgium
Figure B.11: Spain - Portugal
APPENDIX
267
APPENDIX C: WORKED EXAMPLE FOR CUILLIAGH MOUNTAIN
WIND FARM, IRELAND
C.1 Introduction C.3 Malin Head Meteorological Station
The main text has provided a general discussion of the The assessment of the wind climate at the site uses data
assessment of the wind resource and energy production. recorded at a nearby meteorological station, Malin Head,
This Appendix is included in order to provide a “worked which is situated on the coast approximately 65 km north-
example”. It demonstrates all the different aspects of the northeast of the Cuilliagh site. From discussions with Met
process outlined in the main text. The project considered Éireann (the Irish Meteorological Service) staff and con-
is the Cuilliagh Mountain wind farm in Ireland, which consideration of other meteorological stations in the region,
sists of 18 Vestas V47 WTs and was constructed in it was concluded that Malin Head was the most appropri-
2000. The following specific analyses are presented: ate reference meteorological station for this analysis.
1 The results of the pre-construction projection of the Data from 1979 to 2000 have been used in the analysis
expected energy production of the wind farm, reported here. Discussions with Met Éireann staff indicate
including uncertainty analysis. that there has been no change during this period which
2 The review of the actual production of the wind farm will have a significant effect on the consistency of the
over a 17-month period. measurements. This is important since the analysis
3 The results of a “wind in–energy out” validation test method used here relies on long-term consistency of the
of the predictive methodologies employed in 1) above. measurements at the meteorological station.
Airtricity, a leading international wind farm developer,

owns the Cuilliagh Mountain Wind Farm and thanks are to C.4 Wind Data
be extended to Airtricity for allowing their proprietary data
to be used for this example case. The data sets from Malin Head and the Cuilliagh site, as
used in the analyses described in the following sections,
are summarised in Table C.1:
C.2 Description of the Site and
Monitoring Equipment Table C.1: Data available from Cuilliagh and from Malin Head
The site lies in central County Donegal approximately 14 Cuilliagh Mountain Hourly mean wind 05 July 1997 - 24 Jan
km southwest of Letterkenny. The wind farm site lies on Mast 05 NRG speed, standard devi- 1999
(206940, 402500) ation, gust and direc-
Cuilliagh Mountain with maximum elevation of approxi- tion at 30 m
mately 360 m.
Hourly mean wind
speed, standard
The site at Cuilliagh Mountain has had one 30 m and two deviation, and
10 m temporary meteorological masts installed in the direction at 10 m
period since mid-1997. The 10 m data are not considered Malin Head Hourly record of 10- 05 July 1997 -
further within this report. Meteorological minute mean wind 24 January 1999
Station NRG speed and direction
(241950, 458550) (time series data).
The wind data from the 30 m site mast have been record-
ed using NRG sensors with a Maximum 40 anemometer Hourly record of 10- 1979 - 1998
minute mean wind
and wind vane at 10 m and 30 m. A NRG 9210 logger was speed and direction
programmed to record hourly mean wind speed, wind (frequency table).
speed standard deviation, three-second gust and direction.

268
C.5 Description of the Proposed Wind From data recorded at local meteorological stations and
Farm with standard lapse rate assumptions, the Cuilliagh
Mountain site is predicted to have an air density of 1.205
The wind turbine model selected for the Cuilliagh kg/m3. Since the predicted mean air density at the site
Mountain Wind Farm was the Vestas V47 660 kW model differs from the air density for which the power curves
with a hub height of 45 m. The basic parameters of the were supplied, a small air density adjustment following
turbine are presented in Table C.2 below. the IEC 61400-12: 1998 standard was made to the power
curves used in the analysis.
Table C.2: Main Parameters of the Vestas V47 660 kW Wind
Turbine The power curve for the Vestas V47 660 kW turbine has
been compared to a reference curve from an independent
Diameter 47.0 m test of the performance of the turbine. It was found that
Hub height 45 m the reference curve out-performed the supplied curve by
Rotor speed 28.5 rpm 2% for the wind regime at the Cuilliagh site. This result
No. of blades 3 indicates that the supplied curve is broadly in line with the
Nominal rated power 660 kW performance that might be expected.
The Cuilliagh Mountain Wind Farm has a total nameplate

The power curve used in the analysis has been supplied capacity of just under 12 MW. It is located approximately
for an air density of 1.225 kg/m3 and is presented in 1.5 km south of the Cark wind farm. The effect of these
Table C.3. turbines on the predicted energy production of the Cuilliagh
development was estimated.
Table C.3: Performance Data for the Vestas V47 660 kW
Wind Turbine
Wind Speed V47 Power Output

[m/s at hub height]
4
5
[kW]
2.9
43.8
C
6 96.7
APPENDIX
7 166
8 252
9 350
10 450
11 538
12 600
13 635
14 651
15 657
16 659
17 660
18 660
19 660
20 660
21 660
22 660
23 660
24 660 269
25 660
Figure C.1: Layout of the Cuilliagh Mountain Wind Farm C.7 Correlation of Wind Regime at
Cuilliagh Mountain and Malin Head
The measured wind direction at Cuilliagh Mast 05 at 30

m is compared to the concurrent wind direction measured
at Malin Head in Figure C.2. The directions recorded
between the two locations show some scatter but are gen-
erally well correlated for the most frequent sectors.
The monitored wind speeds at 30 m height in each of 12

30˚ direction sectors are compared to the concurrent wind
speed at Malin Head in Figure C.3. The quality of the cor-
relation is considered to be reasonable for all direction
sectors. The wind speed ratios for each direction sector
are presented in Table C.4.
Table C.4: Wind Speed Ratios between Cuilliagh Mast 05 at

30 m and Malin Head
Direction Sector Number of Wind Speed Ratio

Hours Analysed
345-15 278 0.701
15-45 194 0.767
45-75 229 0.800
75-105 461 0.718

C.6 Results of the Analysis 105-135 795 0.957
135-165 1098 0.976

The analysis to determine the wind regime and expected
165-195 1622 0.879
energy production of the proposed Cuilliagh Mountain
195-225 1208 0.897
wind farm involved several steps:
225-255 1210 0.894
• The directional correlations between wind speeds
255-285 1230 0.868
recorded at Cuilliagh Mast 05 at 30 m and at Malin
285-315 708 0.834
Head were established.
315-345 421 0.819
• The correlation relationships were applied to historical

All 9454 0.861
wind data recorded at Malin Head to produce a descrip-
tion of the long-term wind regime at Cuilliagh mast 05.
• Wind flow modelling was carried out to determine the
hub height wind speed variations over the site relative to
the 30 m anemometry mast.
• The energy production of the wind farm was calculated,
taking account of array losses and topographic effects.
• The seasonal variation in the energy production of the
wind farm was calculated.
• Sources of uncertainty in the wind speed and energy pro-
270
duction estimates were identified and quantified.
Figure C.2: Correlation of Wind Direction at Malin Head and at Cuilliagh Mast 05 at 10 m
APPENDIX
271
272
Figure C.3a: Correlation of Wind Speed at Malin Head and at Cuilliagh Mast 05 at 30 m – Continued
Figure C.3b: Correlation of Wind Speed at Malin Head and at Cuilliagh Mast 05 at 30 m – Continued
273
APPENDIX
C
274
Figure C.3c: Correlation of Wind Speed at Malin Head and at Cuilliagh Mast 05 at 30 m – Concluded
C.8 Long-term Mean Wind Speed at Table C.5 shows the predicted long-term mean wind
Cuilliagh Mountain speed at each wind turbine location at hub height. The
average long-term mean wind speed at a hub height of 45
The wind speed ratios listed in Table C.4 were used to fac- m for the whole wind farm was found to be 8.1 m/s.
tor the long-term wind speeds at Malin Head for the peri-
od 1979 to 1998. By this method, the long-term mean Table C.5: Mean Wind Speed and Projected Energy Output of
wind speed at Cuilliagh Mast 05 at 30 m was calculated Individual Wind Turbines
to be 7.2 m/s.
Turbine Number Mean Hub Height Wind Energy Output2
Speed1 [m/s] [GWh/Annum]
The corresponding joint wind speed and direction fre-
1 7.7 2.2
quency distribution for Cuilliagh Mast 05 over the histori-
2 7.8 2.1
cal period 1979 to 1998 is presented in Figure C.4 in the
3 7.8 2.1
form of a wind rose.
4 7.6 1.9
5 7.4 2.0
Figure C.4: Annual Wind Rose for Cuilliagh Mast 05 at 30 m.
6 7.8 2.0
7 8.0 2.1
8 8.1 2.3
9 8.4 2.5
10 8.0 2.2
11 8.2 2.3
12 7.6 2.0
13 8.6 2.4
14 8.2 2.3
15 8.2 2.3
16
17
18
8.8
8.5
8.3
2.5
2.4
2.3
C
APPENDIX
Overall 8.1
1
Wind speed at location of turbines at 45 m height, not including wake effects.
2
Individual turbine output includes topographic and array effects only.
C.9 Site Wind Speed Variations at

Cuilliagh Mountain
The variation in wind speed over the Cuilliagh Mountain site

has been predicted using the WAsP computational flow
model. WAsP was used to model the wind flow over the site,
being initiated from the long-term wind speed and direction
275
frequency distribution derived for Mast 05 at 30 m.
C.10 Projected Energy Production Table C.7: Monthly Variation of the Projected Energy Output1
of the Wind Farm
The predicted energy production for the wind farm is
detailed in Table C.6 below. The energy capture of indi-
January 4.27
vidual turbines is given in Table C.5.
February 3.87
March 3.84
Table C.6: Predicted Energy Production of Cuilliagh Mountain
April 2.53
Wind Farm
May 2.16
June 1.86
Ideal energy production 40.2 GWh/annum
July 2.05
Topographic effect 107.0 % Calculated
August 2.21
Array effect 92.7 % Calculated
September 2.85
Electrical transmission efficiency 97.0 % Estimate
October 3.60
Availability 97.0 % Estimate
November 3.67
Icing and Blade fouling 99.0 % Estimate
December 3.99
High wind hysteresis 99.6 % Estimate
Substation maintenance 100.0 % Not considered

1
Energy output includes all losses.
Utility downtime 100.0 % Not considered
Power curve adjustment 100.0 % Not considered
Columnar control losses 100.0 % Not considered C.12 Uncertainty Analysis

Wake effect of existing wind farms 99.8 % Estimate
Net energy production 36.9 GWh/annum The main sources of deviation from the central estimate
have been quantified and are shown in Tables C.8a and
C.8b which consider future periods of 10 years and one
The energy production predictions include calculation of year respectively.
the array and topographic effects, an estimate of avail-
ability and electrical loss and factors to account for WT Table C.8a: Uncertainty in Projected Energy Output1 of the
icing, high wind hysteresis and the wake effect of existing Proposed Wind Farm – 10 Year Future Period
turbines. Other potential sources of energy loss are also
listed. It is recommended that the client carefully consid- Source of Uncertainty Wind Speed Energy Output1
[%] [m/s] [%] GWh/
er these issues since at the time of this energy assess- Annum
ment there was insufficient information to estimate the Anemometer accuracy 2.0 0.14
effect on the predicted energy production. Correlation accuracy 0.19
Period representative
of long-term 1.3 0.10
Total wind 0.26 2.22

C.11 Seasonal Variations
Wake and topographic
calculation - - 3.0 1.11
The monthly energy production of the wind farm is pre-
Wind variability (10 years) 1.9 0.14 1.19
sented in Table C.7. There is a large seasonal variation of
Overall (10 years) 2.75
the predicted long-term monthly energy production, with
winter and summer months producing approximately
140% and 60%, respectively of the long-term mean month-
1
Sensitivity of net production to wind speed is calculated to be 8.68
GWh/annum/(m/s)
ly energy production.
276
Table C.8b: Uncertainty in Projected Energy Output1 of the 3 There is an uncertainty associated with the assumption
Proposed Wind Farm – One Year Future Period made here that the historical period at the meteoro-
logical site is representative of the climate over longer
Source of Uncertainty Wind Speed Energy Output1 periods. A study of historical wind records from a num-
[%] [m/s] [%] Wh/
Annum ber of reference stations indicates an average variabil-
Anemometer accuracy 2.0 0.14 ity of 6% in the annual mean wind speed. This figure is
Correlation accuracy 0.19 used to define the uncertainty in assuming the long-
Period representative term mean wind speed over a 20-year period.
of long-term 1.3 0.10 4 For a finite number of future years, the mean wind speed
Total wind 0.26 2.22 may differ from the long-term mean due to the natural
Wake and topographic variability of a random process. Account is taken of the
calculation - - 3.0 1.11
future variability of wind speed in the energy confidence
Wind variability (1 year) 6.0 0.43 3.75
analysis but not the wind speed confidence analysis.
Overall (1 year) 4.49
1
Sensitivity of net production to wind speed is calculated to be 8.68 It is assumed that the time series of wind speed is ran-
GWh/annum/(m/s)
dom with no systematic trends. Care was taken to ensure
that consistency of the Malin Head measurement system
The figures in these tables, when added as independent and exposure has been maintained over the historical
errors, give the following uncertainties in net energy pro- period and no allowance is made for uncertainties arising
duction of 4.5 GWh/annum for a future one year period due to changes in either.
and 2.7 GWh/annum for a future 10 year period. The
detailed derivation of the above uncertainties is present- Uncertainties type 1, 2 and 3 from above are added as
ed below: independent errors on a root-sum-square basis to give the
total uncertainty in the site wind speed prediction for the
There are four main categories of uncertainty associated historical period considered.
with the site wind speed prediction at Cuilliagh Mountain:
1 There is an uncertainty associated with the measure-
ment accuracy of the site anemometers. The instru-
There are four categories of uncertainty in the energy out-
put projection:
C
APPENDIX
ments used on this site have not been individually cal- 1 Long-term mean wind speed dependent uncertainty is
ibrated to MEASNET standards and a consensus cali- derived from the total wind speed uncertainty (types 1,
bration has been applied. Batch calibration of NRG 2 and 3 above) using a factor for the sensitivity of the
Maximum 40 anemometers have shown them to con- annual energy output to changes in annual mean wind
form to the consensus calibration to within 1.5%. speed. This sensitivity is derived by a perturbation
Therefore, a figure of 2% is assumed here so as to analysis about the central estimate.
account for other second order effects such as over- 2 Wake and topographic modelling uncertainties.
speeding, degradation, air density variations and sen- Validation tests of the methods used here, based on
sor mounting. No allowance has been made for uncer- full-scale wind farm measurements made at small wind
tainty in the Malin Head anemometer as consistency farms have shown that the methods are accurate to 2%
and not absolute accuracy is important. in most cases. For this development, an uncertainty in
2 An error analysis was carried out on the correlation for the wake and topographic modelling of 3% is assumed.
each direction sector and from this the standard error 3 Future wind speed-dependent uncertainties described
for the long-term mean wind speed was determined. in 4 above have been derived using the factor for the
This was carried out for the correlation between Malin sensitivity of the annual energy output to changes in
277
Head and Cuilliagh Mountain. annual mean wind speed.
4 Turbine uncertainties are generally the subject of con- These predictions of net energy include topographic
tract between the developer and turbine supplier and effects, array losses, availability, electrical transmis-
therefore no allowance has been made for them in this sion losses, air density adjustments and factors to
work. account for turbine icing, high wind hysteresis and the
wake effect of existing turbines. Other potential
Again, those uncertainties which are considered are added sources of energy loss are listed section C.9.
as independent errors on a root-sum-square basis to give
the total uncertainty in the projected energy output. The net energy predictions presented above represent
the long-term mean, 50% exceedence levels, for the
annual energy production of the wind farm. These val-
C.13 Summary of the Results of the ues are the best estimate of the long-term mean value
Analysis to be expected from the project. There is therefore a
50% chance that, even when taken over very long
Wind data were recorded at the Cuilliagh Mountain site for periods, mean energy production will be less than the
a period of 18 months. Analysis of this data, in combina- value given in the table. Estimates of long-term mean
tion with concurrent data and historical wind data record- values with different levels of exceedence are set out
ed at Malin Head Meteorological Station, results in the fol- below.
lowing conclusions with regard to the wind regime at the
Cuilliagh Mountain site: 5 The standard error associated with the prediction of
1 The long-term mean wind speed is estimated to be energy capture has been calculated and the confidence
7.2 m/s at a height of 30 m above ground level. limits for the prediction are given in Table C.10.
2 The standard error associated with the predicted Table C.10 Confidence Limits - Energy
long-term mean wind speed at 30 m is 0.26 m/s. If a
Probability of Net Energy Output Net Energy Output
normal distribution is assumed, the confidence limits Exceedence [GWh/Annum] [GWh/Annum]
for the prediction are as given in Table C.9: [%] 1 Year Average 10 Year Average
90 31.1 33.4
Table C.9: Confidence Limits – Wind Speed 75 33.9 35.1
50 36.9 36.9
Probability of Exceedence Long-term Mean
75 39.9 38.7
[%] Wind Speed at 30 m [m/s]
90 42.7 40.4
90 6.9
75 7.0
50 7.2
Site wind flow and array loss calculations have been car-
ried out, from which the following conclusions are drawn:
3 The long-term mean wind speed averaged over all tur-
bine locations at 45 m is estimated to be 8.1 m/s.
4 The projected net energy capture of the proposed

Cuilliagh Mountain wind farm is predicted to be 36.9
278
GWh/annum.
C.14 Actual Production of the Wind Farm detailed assessment of the availability of the wind farm
over the above operational period has not been undertak-
Commissioning of the Cuilliagh Mountain wind farm took en, but it is understood that high availability levels have
place in late 2000; by November 2000 the wind farm was been achieved.
in full commercial operation. A review of its performance
was undertaken early in 2002. The data recorded at Malin Head indicates that the windi-
ness of the period from November 2000 to March 2002
Table C.11: Expected and Actual Production of Cuilliagh was some 4.9% down on long-term expectations, making
Mountain Wind Farm suitable assumptions about the seasonal variation of wind
speed. This implies that over the longer term it is likely
Month Year Expected Actual Production that the energy production of the wind farm will, in fact,
Production (GWh) (GWh)
exceed the central estimate value of 36.9 GWh/annum
Nov 2000 3.670 3.703
and may settle at a level which is close to the 25% excee-
Dec 2000 3.990 3.530
dence level presented above. A more detailed assessment
Jan 2001 4.270 3.546
which includes issues such as wind direction, air density
Feb 2001 3.870 2.876
and availability would be required to provide a revised cen-
Mar 2001 3.840 3.410
tral estimate of wind farm production.
Apr 2001 2.530 2.850
May 2001 2.160 1.699

A separate validation of the accuracy of the modelling
Jun 2001 1.860 2.608
techniques employed to predict the long-term energy pro-
Jul 2001 2.050 1.813
duction of the Cuilliagh Mountain Wind Farm was under-
Aug 2001 2.210 1.538
taken. A comparison was made between the expected
Sep 2001 2.850 2.941
energy production of the wind farm, based on the actual
Oct 2001 3.600 4.369
mean wind speed recorded at Malin Head Meteorological
Nov 2001 3.670 3.645
Station and the actual wind farm energy production. This
C
Dec 2001 3.990 3.679
was undertaken on an hourly basis. Thus, the accuracy of
Jan 2002 4.270 4.801
the correlation relationships between Malin Head and the
Feb 2002 3.870 4.604
site, and of the site flow model and turbine wake models
APPENDIX
Mar 2002 3.840 4.037
was assessed using a “wind in–energy out” test. Suitable
Total 56.540 55.649
adjustments were made to reflect the actual air density at
the site. The comparison was undertaken for the opera-
tional period described above and data were only com-
Table C.11 presents the expected long-term monthly ener- pared where all turbines were available and when wind
gy production of the wind farm, along with the actual ener- farm SCADA data and data from Malin Head
gy production over the period November 2000 to March Meteorological Station were also available. Using these
2002. It can be seen that individual months can deviate criteria, a comparison was made over a total of approxi-
substantially from long-term expectations; for example mately 8,300 hours. The results of the comparison of the
February 2001 experienced production which was only expected and actual energy production of the wind farm
74% of the long-term expectations for this month while in are presented in Figure C.5 as a cumulative plot. Over the
June 2001 140% of the long-term expectations for energy full period considered, the actual production was 99.7% of
production in this month was produced. Over the 17- that expected, which provides confidence in the accuracy
month period for which data are available the actual proof the methods employed. It is noted that for individual
duction of the wind farm was 1.6% below long-term expec- months and for individual turbines larger discrepancies
279
tations. This figure is well within the 75% and 90% excee- between the expected and actual energy production is
dence levels for the prediction presented above. A observed.
Figure C.5: Cumulative Plot Showing Measured Energy against Concurrent Expected Energy for the Operating Period
C.15 Concluding Remarks
This Appendix has shown that the techniques outlined in

the main text can be used to predict the behaviour of a
wind farm with a good level of agreement. It has also
demonstrated that the methods can be used to determine
both mean values and associated uncertainties. It is
hoped that it has proved a useful illustration of the tech-

niques which are presently used by the industry.
280
APPENDIX D: DETAILED DESCRIPTION OF CORRELATION TECHNIQUES
Over the past decade, there has been an ongoing indus- with different methods. In such circumstances, careful
try debate over which correlation methodologies provide checks are required to ensure that the correlation is suf-
the best prediction of the long-term mean wind speed at ficiently good to justify the use of the reference meteoro-
a site. All correlation methods have common features in logical station. Due consideration also needs to be given
that they (i) establish a relationship between the concur- to the interpretation of the uncertainty associated with a
rent data recorded at the site and reference station, and specific correlation methodology.
(ii) apply the relationship to the historic data recorded at
the reference station to predict the long-term wind regime The methods based on 10-minute data or hourly data typ-
at the site. Such methodologies are commonly called ically use the long-term wind rose recorded at the refer-
measure correlate predict (MCP) analyses. Variables in ence meteorological station. Those based on daily or
such correlation analyses mooted over the past decade monthly correlations are dependent on the site wind rose.
include those defined in Tables D.1 and D.2 below. In practice, it is often observed that where hourly or 10-
minute correlations between a site and reference station
Table D.1: Prediction Methodologies Based on 10-minute or are poor, a reasonable correlation is observed over longer
Hourly Data data collection periods, such as monthly.
Technique Option 1 Option 2 Others … Detailed Description of a MCP Analysis

Directional bin size 30˚ Other
Regression analysis Principal Least A detailed description of the steps within a MCP analysis
technique component squares
analysis fit is described below, based on hourly data from the site
Fitting method One Two Non- and reference station. As indicated in the previous sec-
parameter parameter linear
tion, different approaches may be used. In the following
fit fit
discussion, the proposed wind farm site is referred to as
Low wind speed cut off Exclude Include
lowest wind lowest wind the “target site” and the meteorological station is referred
speed data speed data
to as the “reference site”.
Table D.2: Prediction Methodologies Based on Longer Term Data

The first stage is to measure, over a period of about a
year, concurrent wind data from both the target site and
D
APPENDIX
the nearby reference site for which well-established long-
Technique Option 1 Option 2 Others … term wind records are available. The short-term measured
Averaging period Monthly Daily wind data are then used to establish the correlation
Fitting method One Two Non- between the winds at the two locations. Finally, the corre-
parameter parameter linear
fit fit lation is used to adjust the long-term historical data

Threshold for data coverage Varies recorded at the reference site to calculate the long-term
mean wind speed at the target site.
The above tables present a bewildering array of options. The concurrent data are correlated by comparing wind
While the technical merit of some methods over others speeds at the two locations for each of 12 30˚ direction
can be argued, experience has shown that where the wind sectors, based on the wind direction recorded at the ref-
regimes at the site and reference meteorological station erence site. This correlation involves two steps:
are well correlated, the results obtained tend to be rela- • Wind directions recorded at the two locations are com-
tively insensitive to the specific correlation methodology pared to determine whether there are any local features
adopted. For cases where the correlation between the site influencing the directional results. Only those records
281
and reference station is less good, then significant diver- with speeds in excess of, say, 5 m/s at both locations
gence is sometimes seen between the results obtained are used.
• Wind speed ratios are determined for each of the direc-
tion sectors using a “principal component analysis”.
In order to minimise the influence of localised winds on

the wind speed ratio, the data are screened to reject
records where the speed recorded at the reference site
falls below 3 m/s (or a slightly different level) at the tar-
get site. The average wind speed ratio is used to adjust
the 3 m/s wind speed level for the reference site to obtain
the different level for the target site, so ensuring an unbi-
ased exclusion of data. The wind speed at which this level
is set is a balance between excluding low winds from the
analysis and still having sufficient data to carry out the
analysis. The level used only excludes wind speeds below
the cut-in wind speed of a WT, which do not contribute to
the energy production.
The result of the analysis described above is a table of

wind speed ratios, each corresponding to one of 12 direc-
tion sectors. These ratios are used to factor the wind data
measured at the reference site over the historical refer-
ence period, to obtain the long-term mean wind speed at
the target site. This estimate therefore includes the fol-
lowing influences:
• “Speed-up” between the target site and the reference
site on a directional basis. This can be a very important
characteristic; sometimes speed-ups differ by a factor
of as much as 2.
• The wind patterns at the reference site have been
translated through the correlation process so that the
long-term pattern at the target site has also been
established.
282
APPENDIX E: CONNECTION CONSIDERATIONS
E.1 Timing Constraints ciently high to enable the addition of new wind generation
capacity, then the options are to connect to another
In many cases, the time to construct the network connec- stronger network (which may be further away), or to rein-
tion and any additional network reinforcement can be force the network.
longer than the time it takes to design, finance and build
the wind farm itself. This is a new situation which applies
to all embedded generation and also to some new forms E.3 Voltage Range
of large-scale generation. For traditional conventional gen-
eration projects, the network connection is not on the For distribution networks, the rise in voltage adjacent to
“critical path”. the wind farm is often the limiting factor for wind farm
size. The power and reactive power produced or con-
Timing issues are particularly significant on the transmis- sumed by the wind farm causes the voltage levels within
sion system, as it can take several years to obtain permis- the network to change. The exact effect is complex and
sion and construct new or reinforced transmission lines. depends on other power and reactive power flows, as well
as the voltage control equipment which already exists in
the network. If the voltage level at a customer is estimat-
E.2 Network “Strength” ed to go outside the statutory limits, then something must
be done. For example:
An important consideration is the strength of the network • improve voltage control equipment;
at the proposed point of connection, varying from “strong” • install power factor correction equipment;;
to “weak”. Embedded generation and large consumers, • control the reactive power output of the wind farm;
where either output or demand can change significantly • limit the wind farm size; or
over a short time-frame, can cause relatively large • reinforce the network.
changes in the network voltage on a weak electrical sys-
tem. A strong network, on the other hand, will be relative- Sometimes, the point on the network which is most affect-
ly unaffected by changes in generation and demand.
A weak electrical system will have a low “fault level” or

ed by such changes is the wind farm itself. In this case, it
is possible to agree a connection where the voltage range is
expected to be greater than the statutory range. However,
E
APPENDIX
“short circuit level”, which is usually measured in MVA. the voltage range at the WT terminals must still be within
The strength of a point on the network is determined by the acceptable range specified by the turbine supplier.
impedance between that point and the main generators
on the system. Put another way, a weak point on the elec- If this issue cannot be addressed through the above
trical system is one which is further away from large means then it may be necessary to curtail the output of
amounts of generation than a strong point on the system. the wind farm when voltage levels may go outside the
The strength of a network also determines the current acceptable range. It may be worth accepting this occa-
that will flow in the event of a fault. sional loss of production for rare combinations of circum-
stances. Equally, it may be possible to adopt a different
Wind farms in windy rural areas often find that the network turbine choice that will be better equipped for dealing with
is weak, primarily because these are often sparsely pop- the characteristics of the network.
ulated areas of low demand, with long distances from the
main users and generators. Although on its own the
strength of a network does not indicate the maximum E.4 Thermal Rating Limits
wind capacity that can be connected, it is a good indica-
283
tor of the kind of issues that might emerge. If a study Thermal ratings of cables, overhead lines and transform-
shows that the strength of a chosen network is not suffi- ers may also be limiting factors. In this case, network rein-
forcement is often the only solution. It may be possible to The standard also provides formulae by which the charac-
negotiate some automatic or manual means to reduce teristic parameters of a WT can be used in conjunction
wind farm output when thermal limits are approached. with project dependent parameters, in order to estimate
This may only occur rarely, e.g. when one line of two is out what the power quality effects of a proposed wind farm
of service, and the loss of production may therefore be will be. The results can be compared with the network
acceptable. operator’s requirements to decide if the wind farm will be
acceptable.
E.5 Fault Current Ratings E.6.1 VOLTAGE STEP CHANGE
The calculated ”fault current levels” (the current that flows Currents flowing in the electricity network affect the volt-
in the event of defined faults) on a system may be close to age seen by other customers. A sudden increase or
the ratings of the switchgear which will have to interrupt decrease in current will cause a step change in voltage
that fault current. This is particularly true in urban and on which may be perceptible to other customers. Such
higher voltage networks. New generation on the system changes may also occur too rapidly for the voltage control
may therefore force replacement of the switchgear with new systems operated by the network operator (principally
equipment with a higher rating, which can be expensive. adjustment of the ratio of main transformers), causing the
voltage at some point in the network to go outside the
The fault current issue may also be more significant at statutory limits. Typically, the maximum voltage step is
transmission level, as the high cost of uprated switchgear limited to between 2% and 5%.
means that transmission networks are often already oper-
ated close to switchgear ratings. Voltage step changes can be caused by WTs starting up,
particularly fixed speed wind turbines with induction gen-
erators. It is especially an issue for fixed speed stall-reg-
E.6 Power Quality Issues ulated wind turbines because they have no control over
the rate at which the rotor accelerates during start-up, and
Power quality does not often limit wind farm size, but it is have to energise the generator just as its rotational speed
of concern to network operators and must always be con- matches synchronous speed. There is often some speed
sidered. Network operators need to provide a certain qual- mismatch at this point and so there is an “inrush current”
ity of power to their customers. WTs can affect networks to accelerate or decelerate the rotor to match synchro-
such that power quality seen by other customers is affect- nous speed, as well as the normal inrush current to mag-
ed. Therefore, network operators demand that these netise the generator.
effects are quantified and, if necessary, limited. There are

four main issues: The same effect occurs when fixed speed WTs with two
• Voltage step changes. speeds (two generators) change from one speed to the
• Flicker. other. A similar effect is seen when the WT stops, espe-
• Harmonic distortion. cially when shutting down from full power due to the upper
• Voltage imbalance. wind speed limit being exceeded.
IEC 61400-21 is an international standard which provides Pitch-regulated WTs can control rotor speed during start-
means to calculate, from measurements, parameters that up, so this effect is reduced. The effect is even less for
characterise the power quality of a WT design. WT manu- variable speed turbines. Fixed speed machines often use
facturers can get their machine tested by a third party and “soft start” power electronic devices to reduce the voltage
284
produce test certificates in a similar way to a power curve step change to insignificant levels.
or noise emission characteristic.
As it is accepted that WTs in a wind farm will not start monics at frequencies (approximately 2 kHz) above the
simultaneously, this is only a problem for single turbines range usually of concern to network operators.
or small clusters, perhaps two or three, on weak networks.
In summary, harmonic emissions should not be a problem
E.6.2 VOLTAGE FLICKER unless there are specific features of the network in question.
Flicker is a concern for many system operators. In reality, E.6.4 VOLTAGE IMBALANCE
it is rarely a real problem, although it must always be
checked. As a general rule, if the effect of a wind farm in Network operators try to keep the currents in all three
terms of voltage rise and other basic technical issues is phases of the network similar so that the voltages are
satisfactory, then flicker is probably satisfactory too. also similar. Unbalanced currents can cause the voltages
to differ, which can damage customer equipment.
Flicker is caused by small voltage changes occurring rap-
idly and sequentially, which causes lighting to flicker and Voltage imbalance is often included within power quality
hence customer annoyance. Flicker therefore tends to be by network operators. However, for WTs it is a different
more of an issue with one or a few WTs on lower voltage kind of issue and is not therefore dealt with by the IEC
distribution networks with low fault levels. For large wind 61400-21 standard.
farms, the smoothing effect of the power fluctuations
from large numbers of turbines means that the flicker As three-phase rotating machines, WTs make no real con-
effect is greatly reduced. tribution to system voltage imbalance. In fact, as induc-
tion machines they tend to reduce the imbalance. The
The most important flicker-producing events are turbine downside of this is that large “negative phase sequence”
starts and stops, and switching between generators (for currents can flow within the generator and cause exces-
two-speed WTs). Fortunately, switching operations can be sive heating.
limited in frequency of occurrence through appropriate
programming of the turbine controller. Again, flicker is not
normally a problem for variable speed turbines.
Network operators often do not know the existing levels of
voltage imbalance at a specific point on their system. It
can be measured but, as it is a function of load currents,
E
APPENDIX
E.6.3 HARMONIC DISTORTION it varies during the day, week and, possibly, over the year.
Network operators may say they aim to keep it below
Variable speed WTs have power electronic converters agreed limits, but it could, in practice, be significantly
which can emit currents at frequencies above the funda- higher at some locations. If this problem is identified,
mental frequency (50 or 60 Hz). These harmonic currents network operators will attempt to cure it by re-allocating
can cause annoyance to customers and can even damage single-phase customers across the three phases. This
equipment. This is only an issue for variable speed wind can take some time and may not be a complete solution.
turbines. There is a tendency for the more modern WTs to A complete solution may require network reinforcement
use power electronic converters and so harmonics do which implies cost and delay.
need to be addressed. Harmonic emissions are, however,
well understood by WT manufacturers. Generally, they are In some rural networks these risks may justify making
also quite low and rarely considered an issue in practice. measurements of voltage imbalance before choosing a
point of connection.
To keep harmonic emissions within the required levels,
WTs often use frequency converters with “pulse width Voltage imbalance is usually only an issue on weak lower-
285
modulation” (PWM). These produce very low levels of har- voltage networks.
E.7 Typical Upper Limits offshore development, rather than the wind farm con-
structing a line to an existing utility network. As with all
Table E.1 lists some “rule of thumb” limits for the amount wind farm developments, the location of the connection
of embedded generation that may be connected within a into the utility system is likely to bear heavily on the costs
network. (Note that in many systems, 100 kV or above of the development and so will be subject to case-by-case
would be considered as part of the transmission system.) discussions between the parties involved on how these
costs are to be met.
These figures are upper limits. In any one case, there
could be many factors which would limit the maximum
capacity below these levels.
Table E.1: Rule of Thumb Limits for Embedded Capacity
Connection Point Typical Maximum Generation

Capacity Which May be Connected
Low voltage A few kW of embedded generation

capacity
Lower levels of the distribution Up to 2 MW, or possibly more
system (typically 10 or 11 kV) than 2 MW close to the trans-
former feeding the network
Upper levels of the distribution Can take 10 to 15 MW
system (20 – 35 kV):
existing overhead line or cable
Upper levels of the distribution Is likely to accept up to the rating

system (20 – 35 kV): of the transformers, which could
existing busbar in a substation be 60 MW or more.
‘Subtransmission’ system A typical limit is 100 MW
(70 – 150 kV):
existing overhead line or cable
‘Subtransmission’ system A typical limit would be several

(70 – 150 kV): hundred MW
existing busbar in a substation
Transmission systems Generalisations not possible
E.8 Connections for Offshore Wind Farms

Offshore wind farms must connect to land-based networks

in the same manner as for onshore wind farms. They are
therefore subject to the same network considerations.
However, as economic reasons favour most offshore wind
developments to be large in comparison to onshore devel-
opments, it is more common for an offshore wind farm to
connect directly into a transmission system rather than a
distribution system. Issues such as power quality are
therefore unlikely to be important.
It should be noted that, although offshore developments

286
must connect to land-based networks, there may be
cases where a utility will extend its own system out to an
APPENDIX F: POSSIBLE GRID CODE REQUIREMENTS
Grid code documents set the requirements for users of in advance in order to keep the ramp rate at less than the
the transmission or distribution system. These have set limit.
evolved to suit conventional generation, and substantial
modifications are required to apply them to new forms of As for the power cap function, ramp rate limits will waste
generation, particularly wind. Such modifications have free energy and should only be used when essential.
been produced, or are in the process of being developed,
in many European countries. The following sections sum-
marise common grid code requirements for wind, drawn F.3 Voltage and Frequency Operating
from published or draft documents. Range
Historically, when a disturbance was seen on the system

F.1 Power Cap (manifested as voltage or frequency going out of the
acceptable range), wind farms were required to discon-
A power cap is an adjustable maximum limit on the output nect as soon as possible. However, when wind capacity
of a wind farm. It is relatively easy to implement and is a penetration is high enough for the system not to be able
function that is, or will shortly be, available from most of to withstand this sudden loss of generation, this principle
the major WT manufacturers. The power cap allows the has to be changed. Instead, the wind generation must
system operator to limit the maximum output of wind continue to operate over a wider range of voltage and fre-
farms in its area for short periods. This is particularly use- quency. This is perfectly feasible for most WTs, although
ful for critical periods such as: not all existing designs will be able to meet the new cri-
• times of very low demand; teria.
• times of great volatility in demand; and
• times when great volatility in the output of the wind
generation is expected, for example, due to the F.4 Frequency Regulation
passage of a storm.
Clearly, it is economically advantageous to use all other

available options before applying a power cap, as this
Conventional generation provides frequency regulation by
automatically changing the output of a few of the major
generators in response to changes in system frequency
F
APPENDIX
wastes “free” energy and increases emissions from con- which, in turn, are due to changes in customer demand.
ventional forms of generation. This is relatively easy to do.
Pitch-regulated WTs can also provide adjustment of output

F.2 Ramp Rate power in response to changes in system frequency but

only by “spilling” available wind. This is, therefore, an
A ramp rate is a limit on the maximum rate of change of expensive way to provide this function.
the output from a wind farm. A positive ramp rate is easy
to implement using WT and wind farm SCADA systems. Stall-regulated turbines will have much greater difficulty in
Rates of around 10% of wind farm capacity per minute providing a similar effect.
have been specified by network operators.
An alternative solution is to set up a market for frequency
A negative ramp rate is much harder to achieve. A crude regulation to which generators can bid. Those generators
partial solution is to limit the rate at which turbines can (such as major conventional plant) which can provide this
shut down. To fully meet the intention of this requirement, service at relatively low cost will therefore be chosen to do
287
forecasting of wind farm output is necessary so that if a so. Generation which finds this difficult (such as wind) will
sudden drop in output is foreseen, output can be reduced not need to do so.
F.5 Reserve
Reserve is required by system operators in order to cope Wind generation cannot provide a contribution to reserve
with the sudden loss of a major generator either due to a requirements except at high cost, i.e. by always operating
failure of that generator or to a failure of the transmission at a point below the power that could be produced in the
system connected to that generator. Reserve is conven- given wind conditions. Wind is, therefore, a very expensive
tionally provided by the stored energy within conventional way of meeting reserve requirements and, as for frequen-
generation including that stored within the steam systems cy regulation, there is an argument that a market for
of thermal plant. For the first few seconds after the loss reserve may be more appropriate than enforcement of
of a major generator, the system frequency will drop and reserve requirements on generators.
energy is rapidly extracted from all the remaining genera-
tors by opening their steam governors (or equivalent) in There is a possible argument that variable speed WTs
response to the drop in frequency. This process takes may be able to contribute to reserve requirements for the
only a few seconds; then, increased fuel flow into the con- first few seconds by extracting energy from the spinning
ventional generators acts to restore system frequency. If inertia. Although in principle simple to achieve, in practice
necessary, other generators may be started up on the implementation and the requirement to be able to
timescales of minutes. demonstrate this capability to system operators mean
that, in all likelihood, WT manufacturers will not develop
Figure F.1 demonstrates such a drop in frequency, and this facility unless a market or some other reward for pro-
the rapid increase in production from the remaining viding this benefit exists.
generation within a 30-second timeframe.
Figure F.1: System Reserve

288
F.6 Reactive Power and Voltage Control
Conventional generation can be controlled to produce or

consume reactive power almost at will and at little cost.
This feature is used to control voltages at points on the
system. If conventional generation is displaced, wind
must fulfil the same function. This is not currently possi-
ble with most WTs, but some manufacturers offer such a
facility and more are expected to follow. The costs for this
function are expected to be minor for variable speed WTs
and more expensive for fixed speed turbines.
It is not clear whether electricity systems actually require

the large reactive power production or consumption abili-
ty of synchronous generators (power factor 0.85 or less).
Therefore, it is not clear if this wide range should be
required of wind generation.
F.7 Transient Stability

(“Fault Ride-through”)
Wind farms can no longer expect to be disconnected in

the event of system transient disturbances. As wind pen-
etration increases, it is increasingly important that wind
generation continues to operate during transient system
disturbances. It is usually necessary to demonstrate that
this is possible to the network operator before any such
F
APPENDIX
event occurs. For conventional power generation this is
done using simulation models. Such models are being
developed for WTs, but there is considerable difficulty in
understanding, developing and validating these models.
This is currently a research area.
289
APPENDIX G: CALCULATION OF SPECIFIC EMISSIONS OF STANDARD
AIR POLLUTANTS FROM FOSSIL FUEL ELECTRICITY
GENERATION (METHODOLOGY)
All calculations are based on the only available data, that
for electricity generation per fuel (TWh/a) and for total
emissions from electricity generation (kt/a). The latter are
divided according to the shares of electricity generated
from the different fossil fuels. It is assumed that different
types of power plants have different specific emissions.
Calculations are made assuming three different reference
emission level scenarios (best case - very good scrubbers;
intermediate case - good scrubbers; worst case - no scrub-
bers at all and worst fuel quality), shown in Table G.1.
Table G.1: Assumed Emission Levels of Power Plants
Fuel Type Best Case Intermediate Case Worst Case
SO2 (mg/kWh)
hard coal 350 700 15000
lignite 350 700 90000
fuel oil 350 700 25000
natural gas, derived gas 0 0 100
mixed firing, not specified Average emissions
NOX (mg/kWh)
hard coal 350 700 3000
lignite 350 700 3000
fuel oil 350 700 3000
natural gas, derived gas 150 300 1500
mixed firing, not specified Average emissions
It is assumed that the specific emissions for all types of

power plants and fuels are at the same technical level for
each country. The graphs in Figure G.2 show a fixed rela-
tionship between emissions and different fuels.
The mathematical model used for the calculation of spe-

cific SO2 emissions, the inputs to the EcoSense model,
plus the software and all other relevant calculations are
documented on the CD attached to this report.
290
Figure G.1: Fixed Relations of Emission Levels of SO2 for Different Fuels in a Country
Source: own calculations.
Figure G.2: Fixed Relations of Emission Levels of NOX for Different Fuels in a Country
APPENDIX
291
Source: own calculations.
APPENDIX H: ASSUMPTIONS FOR THE CALCULATIONS
OF EXTERNAL COSTS WITH ECOSENSE
Calculations of Specific External Costs Assume the Following Geographical Locations of the Power Plants in each Country
Country Northern Latitude Eastern Longitude
AT Austria 48 16
BE Belgium 51 4
DK Denmark 56 12
FI Finland 60 24
FR France 49 2
DE Germany 51 7
GR Greece 40 22
IE Ireland 53 -9
IT Italy 45 9
LU Luxembourg 50 6
NL Netherlands 52 5
PT Portugal 39 -9
ES Spain 43 -6
SE Sweden 56 14
GB UK 53 -1
CY Cyprus 35.25 29*
CZ Czech Republic 50 13
EE Estonia 59 27
HU Hungary 48 20
LV Latvia 57 24
LT Lithuania 55 25
MT Malta 36 14
PL Poland 50 19
SK Slovakia 49 22
SI Slovenia 46 15
BG Bulgaria 42 26
RO Romania 45 25
TR Turkey 40 29*
included area from 35.25 -10
to 73.86 29
* In the case of Turkey and Cyprus, the generation facilities are assumed to be at the eastern border of the area covered by the model because
of the limited area covered by EcoSense.
292
APPENDIX I: MARGINAL COST OF CO2 EMISSIONS
The values of the marginal cost of CO2 in the following

Table are in US$/tonne of carbon (1996 dollars). The
lower value in each box corresponds to a time horizon of
300 years, the upper value to 1,000 years. In the first
row, the distribution of income is not taken into account.
In the second row, this aspect is included on the assump-
tion that it remains constant over time (Azar and Sterner,
1996, p. 181).
The Pure Rate of Time Preference
0 %/year 0.1%/year 1%/year 3%/year
The marginal cost of CO2 emissions, MC1 85-200 75-140 32-33 13-13
The marginal cost of CO2 emissions, MC2 260-590 230-410 95-98 39-39
APPENDIX
293
APPENDIX J: CALCULATED EMPLOYMENT IN EU LEVEL SECTORS
Employment per Final Demand Unit (mio ECU) 1995 EU Direct Indirect Total EU
1 Agriculture, forestry and fishery products 31.66 5.78 37.44
2 Fuel and power products 3.29 19.49 22.79
3 Ferrous and non-ferrous ores and metals 4.93 8.51 13.44
4 Non-metallic mineral products 10.13 4.73 14.86
5 Chemical products 6.12 13.83 19.95
6 Metal products except machinery 13.11 6.73 19.84
7 Agricultural and industrial machinery 9.75 3.85 13.60
8 Office and data processing machines 9.87 0.85 10.72
9 Electrical goods 9.56 4.66 14.22
10 Transport equipment 7.25 2.90 10.14
11 Food, beverages, tobacco 6.37 10.69 17.05
12 Textiles and clothing, leather and footwear 17.89 1.89 19.78
13 Paper and printing products 10.67 8.58 19.25
14 Rubber and plastic products 9.11 5.16 14.27
15 Other manufacturing products 15.12 1.91 17.03
16 Building and construction 13.33 4.81 18.14
17 Recovery, repair services, wholesale, retail 20.62 19.19 39.81
18 Lodging and catering services 19.04 2.37 21.41
19 Inland transport services 14.89 6.84 21.73
20 Maritime and air transport services 6.58 1.25 7.83
21 Auxiliary transport services 6.89 5.11 12.01
22 Communication services 13.60 4.52 18.12
23 Services of credit and insurance institutions 12.02 19.34 31.36
24 Other market services 12.89 44.72 57.61
25 Non-market services 21.30 3.33 24.62
Simple Average 12.24 8.44 20.68

294
APPENDIX K: EASTERN PROMISE, WIND DIRECTIONS, EWEA, MARCH 2003
New EU member states and other Central and Eastern 1990s. But one development of 14 larger turbines is
European countries expected to start construction this spring, with more to
follow. German developer UTEC-Thomsen, active across a
BULGARIA number of East European states, has plans for up to
250 MW.
Population: 8.3 million
Capital: Sofia HUNGARY
Generation capacity: Fossil fuels (54%), nuclear (31%),
hydro (15%) Population: 10 million
Although a countrywide wind atlas has identified areas with Capital: Budapest
average wind speeds in excess of 9m/s, there are no tur- Generation capacity: Fossil fuels (77%), nuclear (22%),
bines operating yet and no mechanism in place to encour- hydro (0.6%)
age them. Nonetheless, a project with 19 Nordex 1.3 MW Although average wind speeds are relatively low, Hungary
turbines has been proposed by a private developer at Peak has the potential to join the wind power leaders among
Murgash, north of Sofia, where wind speeds are said to the new East European accession states, especially with
average 10 m/s. Construction is scheduled for 2005. successful implementation of government support pro-
grammes aimed at reducing CO2 levels. One projection
A study by the National Institute of Meteorology in Sofia from the Horvath Engineering consultancy is that 800 MW
has estimated the total wind power potential in Bulgaria could be installed over the next 10 years.
at 2,200 to 3,400 MW. The most promising sites are on
the Black Sea coast, in the central mountain range and in Hungary’s electricity supply presently comes from a mix-
the Rhodop Mountains in the southwest. Meanwhile, four ture of nuclear and fossil fuel plants, increasingly using
out of six reactors at the Kozloduy nuclear plant are natural gas. Under a new law, however, every licensed
scheduled to close as part of the conditions for Bulgaria’s electricity company has to accept renewable power gener-
membership of the EU, expected to take place in 2007. ation, as long as it complies with certain technical
CZECH REPUBLIC
requirements. The price paid is guaranteed within a range
varying from 6.5 c€/kWh up to 9.5 c€/kWh, depending on
the time of day the power is delivered.
AK
APPENDIX
Population: 10.3 million
Capital: Prague The Hungarian government has also pledged to meet 6%
Generation capacity: Coal (54%), nuclear (32%), hydro of the country’s electricity needs from renewables by
(14%) 2010, double the present 3%. However, the powerful
Wind power in the Czech Republic was given a boost at Ministry of Economic Affairs, sees the main renewables
the end of 2001 with the introduction of a new feed-in tar- contribution coming from biomass, geothermal and solar
iff set at 9.5 c€/kWh. This is the minimum amount to be PV, and is rather less than enthusiastic about wind.
paid to wind producers by the distribution companies, who
have an obligation to accept new supply. A number of Supply competition: Steady privatisation has moved
areas around the country have wind speeds in the range through Hungary’s electricity sector, with a 2001
of 8-9 m/s, so the prospects are good. The Czech Wind Electricity Act introduced specifically to bring the country
Power Association puts the total potential at 600-700 into line with EU Directives on third party access, subsidy
MW, half of which could be installed over the next five elimination and segmenting the electricity market into
years if long-term contracts were available. generation, distribution and power trading companies.
Supply competition is also being introduced in a rolling
295
Operating wind capacity at the end of 2002 was 7 MW, programme. A handful of WTs have gone up so far, all indi-
mainly consisting of small turbines erected during the vidual Nordex and Enercon machines. Three Enercon E-40
600 kW turbines have been developed with backing from price to be raised to 6.2 c€/kWh, a level at which (com-
E.ON Hungaria, a major power supplier and developer. The bined with other incentives) developers would be prepared
most prominent developer active in Hungary is German to invest. Working together with environmental organisa-
company UTEC-Thomsen, which is progressing plans for a tions and the Ministry of the Environment, the EWPA
number of large wind farms using a mixture of Nordex and would also like to see the price unbundled from the Narva
NEG Micon turbines. The first of these, 40 x 1.5 MW tur- plant and for contracts under the purchase obligation to
bines near Tesz, should start building during 2003. last at least 10 years.
Leading manufacturers Bonus, Vestas and GE Wind have
also all shown interest in the market potential. Weak grid: Estonia has the potential for at least 560 MW
of wind capacity, generating roughly 1.28 TWh, according
Consultant Dr Gabor Horvath says that the most important to the EWPA. But that would require major improvements
technical criteria for wind farms in Hungary are a minimum to the relatively weak grid in many parts of the country.
average wind speed of 5-6 m/s at a height of 40-50 m, a The best sites are to be found round the long Baltic coast-
nearby grid connection, accessibility to a main road for line and on the large islands of Hiiumaa and Saaremaa.
heavy vehicle traffic and a satisfactory environmental In the short term, the EWPA expects about 100 MW to be
impact assessment. Wind resource and environmental built. Only one wind farm is already operating – a 1.8 MW
assessments are being carried out at sites with the poten- development with three Enercon E-40 turbines commis-
tial for several hundred MW of capacity, he adds. sioned last October at Virtsu on the Baltic coast with
German government assistance. This was inaugurated by
ESTONIA the Estonian President, Arnold Rüütel. But many more are
in the pipeline, the first of which should be a project with
Population: 1.4 million up to eight Nordex 2.5 MW turbines on the Paldiski penin-
Capital: Tallinn sula, just along the coast from the capital Tallinn. German
Generation capacity: Oil shale (94%), gas (6%) project developer Ostwind is involved in a joint venture
The coastal regions of Estonia have a good potential for with an Estonian partner at Tamba, on the Baltic coast
wind energy, and a number of national policy decisions near the town of Pärnu. Good wind conditions and a sat-
have already been taken to encourage its exploitation. isfactory grid connection have already been established,
Following a 1998 long-term development plan for the ener- ready for the installation of four Südwind 1.5 MW tur-
gy sector, which envisaged a strategic increase in the con- bines. Ostwind is looking for a total of 20-30 MW in
tribution from both wind and hydro power, an energy law Estonia, but “this requires partners willing to share the
passed the same year placed an obligation on the nation- risk,” says the company’s Christoph Markl-Meider. “This
al distribution company to purchase renewable energy. is linked to the political and legal conditions, which do not
Most recently, new legislation will link the tariff for renew- yet comply with EU structures, but which are a prerequi-
able power to the price for output from the large oil shale site for successful development in the long run.” German
power stations at Narva on the Russian border, which gen- developer UTEC-Thomsen also has plans for 150 MW of
erates most of the country’s electricity. Renewables, capacity, using NEG Micon or Vestas turbines.
including wind, will receive 1.8 times the Narva price, at
present 2.8 c€/kWh, bringing the wind price up to 5.1 Tax concessions: Apart from the payment tariff still pend-
c€/kWh. ing in the current legislation, wind developers can benefit
from tax concessions under Estonian law. No VAT is
Payments under the new law, expected to start operating payable on wind or hydro power until summer 2004, when
from this summer, would last for two years, but with a cut- a rate of between 5% and 10% will be introduced.
off date at the end of 2015. This is not a strong enough Although import duty is payable on equipment such as tur-
296
incentive, according to the Estonian Wind Power bines brought in from overseas, this is recoverable in full
Association (EWPA), which has been calling for the tariff at the end of the financial year. Tax on business turnover
of 18% is also recoverable if it is used for reinvestment in state’s current dependence on the Ignalina nuclear plant
the business itself. for 73% of its power. However, phased closure of Ignalina,
starting with one unit in 2005, is a condition of accession
The EWPA adds that a soft loan may be available through to the EU. According to the country’s National Solar
the government agency KIK’s environmental arm Programme, which runs until 2005, the aim is to introduce
(www.kik.es) and equity involvement through the Baltcap a range of renewable technologies, including solar PV,
fund (www.baltcap.com). geothermal, wind and small hydro, but with the emphasis
on biomass. The programme also calls for a lessening in
LATVIA the influence of state energy monopolies and the estab-
lishment of a guaranteed purchase price for renewable
Population: 2.4 million electricity.
Capital: Riga
Generation capacity: Hydro (74%), fossil fuel (26%) The only WTs installed so far in Lithuania have been in the
Latvia has 23.8 MW of installed wind capacity. Most of 60 kW range, mostly designed and built locally during the
this is accounted for by the 19.8 MW Veja wind farm at 1990s. Subsequent technical problems with these
Liepaja on the Baltic coast, with 33 Enercon E-40 turbines machines are explained by a mixture of lack of expertise
installed in April 2002. An indication of the problems and the need for more thorough resource assessment.
encountered by such projects in a country with relatively
poor infrastructure is that the crane used to install these A 4 MW demonstration wind farm is planned at Butinge on
turbines had to be imported from Finland. The Veja wind the Baltic coast, but has yet to secure financing.
park was commissioned under a now abandoned system Meanwhile, the Lithuanian government now says it wants
based on a payment of twice the household tariff. The gov- to replace Ignalina with a new nuclear plant, a decision
ernment has since passed a new law providing a guaran- which, apart from the safety implications, would commit
teed tariff for the first eight years of a turbine’s operation. the country to a major capital outlay. It is still unclear how
This is set lower, at twice the average electricity selling much of the promised EU compensation payment for the
price, about 5 c€/kWh. After eight years, however, the
price falls to the average selling price, which is currently
just 2.5 c€/kWh. A further drawback is that the system is
closure of Ignalina will go on decommissioning costs, and
how much to support alternatives like wind. K
APPENDIX
based on competitive tenders for a fixed amount of capac- POLAND
ity. A wind atlas of the country shows that there are sev-
eral areas with wind speeds in excess of 6m/s at a height Population: 38.6 million
of 30 m. The best sites are located along the Baltic coast Capital: Warsaw
and around the Gulf of Riga. According to a study by the Generation capacity: Fossil fuels (94%), hydro (6%)
European Bank of Reconstruction and Development Poland encourages renewables through a quota system
(EBRD), there is potential for 550 MW of wind capacity. under which power utilities are expected to source an
increasing proportion of their supply from renewables. This
LITHUANIA is supposed to rise from 2.4% in 2001 to 7.5% in 2010.
The system does not contain any real penalty for failure to
Population: 3.7 million comply, however, and potential investors are now waiting
Capital: Vilnius for a new energy law, scheduled to come into force July
Electricity production: Nuclear (73%), fossil-fuelled CHP 2003, to strengthen the purchase obligation. Wind power
(21%), hydro (6%) producers are currently obtaining a price of about 6-6.5
Lithuania could successfully accommodate 500 MW of c€/kWh for their output from the distribution companies,
297
wind capacity, according to a study carried out for the although the power regulatory body URE has been reluctant
EBRD. This would be an important shift from the Baltic to accept any “green premium” element in payments.
A demand by the national grid operator that generators Electricity production: Fossil fuels (62%), hydro (28%),
provide details of their output 48 hours in advance has nuclear (10%)
also disadvantaged wind producers. Although low interest Although Romania has eagerly signed up to the principal
loans and financial support are available from the international agreements on environmental protection and
EcoFund and National Fund for Environment Protection, climate change mitigation, including the Kyoto Protocol,
the biggest problem for investors is the lack of long-term any follow-up measures, such as a clear environmentally
PPAs lasting at least 10 years. The country presently has based energy policy, have been hesitant and slow.
52 MW of modern WTs, all installed in the last three Nuclear power is still viewed favourably by the govern-
years. The largest developments, both with 2 MW Vestas ment, whilst large hydro projects are seen as satisfying
machines, are 18 MW at Cisowo near Darlowo, developed the need for renewable energy.
by Energia-Eco, and 30 MW at Zagorze near Wolin on the
Baltic coast, developed by EPA and commissioned at the A broadly based national wind map was drawn up in the
beginning of this year. The best sites are in the southern 1990s showing that the most favourable areas are along
mountainous region and close to the Baltic. the Black Sea coast, with average wind speeds up to 7.1
m/s, and in the mountains at above 1,500 m. The same
New developments: The Zagorze wind farm, Poland’s study assessed the overall potential at 2,000 MW, half of
largest to date, is located in an ideal site close to the that offshore in the Black Sea itself. Work has also been
Zalew Szczecinski seawater lake. Construction work was carried out under the EU’s OPET programme to map the
overseen by Wolin North Spolka, a subsidiary of the wind potential along the Black Sea coast more closely.
Danish utility Elsam, which owns the project. Polish com- Only a few demonstration turbines currently operate in
panies were involved in supplying the towers, foundations Romania, however, mainly because of the lack of any reg-
and electrical connections. Annual output is expected to ulatory framework or incentives to encourage new renew-
be between 63 and 70 million kWh. Many more projects ables.
are in the pipeline, a number involving developers from
across the German border. UTEC-Thomsen, for example, The most ambitious plan for a larger scale project has
says it wants to build up to 650 MW of capacity, using a been a 24.5 MW wind farm proposed along two dikes
mixture of Vestas and NEG Micon turbines, over the peri- totalling 14 km in length at the port of Constanta on the
od up to 2007. German utility MVV, working with turbine Black Sea. Despite these reservations, a recent assess-
supplier DeWind, is looking to install 10 wind parks, each ment by the EBRD placed Romania as the “top candidate
with 50 MW capacity, along the Baltic coast. Dutch utility for wind energy development” among East European
NUON has plans for 60 MW near Wolin, using the Kyoto states. The assessment concluded: “Well documented
Joint Implementation mechanism. The biggest Polish resources, a broad range of applications, from small
development company is EPA from Szczecin, which has autonomous units for rural areas to large offshore poten-
been involved in all the recent projects. tial, and the government’s will to comply with EU regula-
tions, all indicate fertile ground worth tilling.”
Vis Venti, the Polish wind energy lobbying group, expects
50 MW to be installed during 2003 and 100 MW during RUSSIA
2004. Up to 1,600 MW could be built over the next five
years, it anticipates. There are also major plans to devel- Population: 147 million
op Poland’s offshore potential. Capital: Moscow
Electricity production: Fossil fuels (66%), hydro (19%),
ROMANIA nuclear (15%)
Less than 10 MW of wind capacity is currently operating
298
Population: 22 million in Russia, but an enormous potential is waiting to be
Capital: Bucharest tapped. One estimate is that the technically exploitable
wind resource in just the European part of the country, SLOVAKIA
where most of the population lives, amounts to 2,308
TWh/year. This is as much as the whole of the EU’s elec- Population: 5.4 million
tricity consumption in 1995. The main problem is that Capital: Bratislava
some of the windiest sites, for instance in the north of the Generation capacity: Nuclear (36%), fossil fuels (33%),
country, are both distant from population centres and hydro (31%),
often not easily accessible by the grid. There are no large-scale WTs operating yet in Slovakia,
although a 2.4 MW project (4 x 600 kW) is being planned
Most of the WTs erected over the past 10 years have at Malé Kaparty, north of Bratislava, part funded under
been small, locally produced models. But the first sub- the EU’s PHARE programme. A mixture of government
stantial wind farm went up during 2002 at Kulikovo in the inaction and low electricity prices hampers further
Kaliningrad region as a result of a cooperation agreement progress. Under proposed legislation, however, there
between the Russian Ministry of Energy and the Danish would be a duty on power companies to “purchase and
government. This 4.5 MW project consists of 20 Vestas transmit” electricity from renewable sources for “econom-
V27 225 kW turbines originally installed in Denmark, and ically acceptable prices”. A report to the EBRD has point-
relocated to Kaliningrad by Danish company SEAS Energy ed out that increasing the renewables contribution would
Service following a repowering project. In the Arctic region, not only improve the environment, but would also create
Russian company VetroEnergo has already installed an up to 5,000 jobs. German developer UTEC-Thomsen has
initial turbine near the port of Murmansk, and now has plans for 35 MW of capacity, to be built during 2004-5
plans for a 3-5 MW wind park at Teriberka, 100 km east using a mixture of Vestas and NEG Micon turbines.
of Murmansk. A study of the Kola Peninsula, including
Murmansk, shows that the region has the potential for up SLOVENIA
to 800 MW of wind capacity.
Population: 2 million
Two years ago, the Russian government approved a nation- Capital: Ljubljana
al energy plan which included installation of up to 232 MW
in 28 regions. This was estimated to cost 10.6 billion rou-
bles, 7% of which would come from the government, the
Electricity production: Fossil fuels (43%), hydro (31%),
nuclear (26%)
Assessing Slovenia’s wind power potential is limited by a
K
APPENDIX
rest from local budgets and private sources. The same shortage of suitable meteorological data, although wind
plan envisaged domestic manufacture of turbines, with an speeds are generally low (under 5m/s on average). The
investment of 1.3 billion roubles, 17% from the govern- government provides only fiscal incentives for investment
ment. Specific future plans include the country’s first off- in renewables. Its favoured option is expansion of the cur-
shore wind farm - a 50 MW development with 25 x 2 MW rent hydro capacity - by upgrading existing plant and build-
turbines set close to the Baltic coast near the town of ing new ones.
Baltiysk. A €1 million feasibility study is due to be carried
out by SEAS, and the project would be implemented by a UKRAINE
joint Russian-Danish company. Like other East European
countries, Russia is going through a process of electricity Population: 49 million
market liberalisation, with a timescale running through to Capital: Kiev
2009. This includes breaking up large state companies Electricity production: Fossil fuels (48%), nuclear (45%),
into competing units, establishing separate generation and hydro (7%)
distribution entities, and creating a wholesale market Although nearly all small turbines by European standards,
which, among other things, will need to see electricity Ukraine now has 44 MW of wind capacity. How fast this
299
prices rise in order to reflect the real cost of delivering expands, however, may depend on establishing closer
power. This should help the economics of wind. ties with the mainstream European industry.
The most important initiative has been the Complex Wind New partnerships: Most turbines currently operating in
Farms Construction Programme, introduced by the gov- Ukraine are USW 56-100s with a capacity of 107.5 kW.
ernment in 1996. This set a target for 1,990 MW of These are based on a design by the American company
capacity to be installed by 2010. By 2030 it is projected US Windpower, which successfully installed many thou-
that 20%-30% of the country’s electricity production could sands of similar machines in California during the 1980s
be satisfied by wind power, saving the annual equivalent and early 1990s. Altogether, over 490 of these turbines
of 18 million tonnes of oil. Financial support for the pro- had been produced by autumn 2002 employing a cluster
gramme comes through an additional payment for the out- of Ukrainian factories, and they are still being installed.
put from officially accepted projects. This currently
increases the tariff to between 3 and 3.5 $c/kWh. If Ukraine is to seriously expand its wind capacity, howev-
However, during 2003 it is planned to introduce a new er, it will almost certainly have to develop partnerships
variable tariff based on a turbine’s productivity, with a with established European manufacturers. This is already
lower rate for the most productive. beginning to happen. German manufacturer Fuhrländer
says it is working on a licensing agreement to produce its
Guaranteed payback: For potential overseas investors, FL1000 1 MW turbine, but is still investigating suitable
the uncertainty of return on their investment is the main partners to produce the towers, nacelles and rotor blades,
stumbling block to involvement in the Ukrainian wind and to carry out a qualified service. Two Turbowinds 600
market. A number of companies have expressed inter- kW machines are also expected to be erected at existing
est over the past decade, including Enercon, Nordex wind farms during the first half of 2003. The Ukrainian
and the US developer SeaWest, but the absence of a government’s main demand is that 100% of the compo-
fixed tariff over a guaranteed period has eventually nents are manufactured inside the country, allowing for-
proved the major obstacle. The government is now work- mer military production plants, for example, to be trans-
ing out a new appendix to the law on electricity supply formed into wind power factories. The plus side of this is
which would stipulate a tariff guaranteeing a payback that a Ukrainian produced 600 kW machine could cost as
period of seven years. Quite apart from these funda- little as US$380,000, compared to more than
mental economic questions, a number of technical US$650,000 if it was imported. The minus side is that
issues must be addressed if wind power is to progress some plants are just not capable of coping with more
in Ukraine. Most important of these is the effective complicated production processes.
identification and selection of suitable sites, especially
in a country with relatively low average wind speeds.
The most promising areas are in the Crimean peninsu-
la, along the coastline and shallow shores of the Azov
and Black Seas, and in the Carpathian Mountains.

Effective use of site selection programs still depends on
reliable long-term data covering wind speed and direc-
tion. This is not generally available in Ukraine. Although
the Inter-branch Scientific and Technical Centre has
developed an original technique of numerically design-
ing meteorological parameters where no wind data is
available, which has proved its effectiveness in prac-
tice, this offers only a partial solution. Until historic
wind monitoring data is available for at least a year at a
given site there will be no guarantee against the sort of
300
mistakes which have already been made at some oper-
ating wind farms.
APPENDIX L: EWEA MEMBERS
Full details of the EWEA membership can be found at www.ewea.org
3E Belgium
A2SEA A/S Denmark
ABB Motors OY Finland
ABB TRANSMIT OY Finland
ACB Engineering France
ADEME France
AEDIE (Associacion para la investigation
y Diagnosis de la Energia) Spain
Airtricity Ireland
ANZ Investment Bank United Kingdom
APER Italy
APPA Spain
APREN Energias Renováveis Portugal
ARMINES - Ecole des Mines de Paris France
Australian Wind Energy Association Australia
Austrian Wind Energy Association Austria
Ballast Nedam Offshore Energy The Netherlands
Baltimore Technologies Spain
Barclays Bank N. Ireland
Bayerische Hypo und Vereinsbank Germany
Black Emerald USA
Bonus Energy A/S Denmark
British Wind Energy Association
Brown Rudnick
Bulgarian Wind Energy Association
UK
United Kingdom
Bulgaria
L
APPENDIX
Bundesverband Windenergie (BWE) Germany
Casco A/S Denmark
Catamount Energy USA
CDE France
CENER, Centro Nacional de Energías Renovables Spain

Ciemat Spain
Circe Foundation Spain
Clipper Windpower Inc USA
CORUS BI-STEEL UK
C-Power Belgium
Czech Society for Wind Energy Czech Republic
Danish Turbine Owners Association Denmark
Danish Wind Industry Association Denmark
De Brandt N.V. Belgium
Densit A/S Denmark
301
Det Norske Veritas Denmark
Deutsche Structured Finance GmbH Germany
DeWind GmbH Germany
DHD France France
DUWIND The Netherlands
ECN Solar and Wind Energy The Netherlands
Eco Wind Power Ltd Ireland
Ecofys BV The Netherlands
Ecotecnia SCCL Spain
Electricité de France France
EMD (Energi og Miljodata) Denmark
Endesa Cogeneration and Renewables (ECYR) Spain
Enercon GmbH Germany
Energia Hidroelectrica de Navarra (EHN) Spain
Enis Renewable Energy Systems LLC USA
Envimac Technology and Consultants Corporation Taiwan
Eole RES France
EPA Poland
ERGA Italy
Ernst & Young UK
Escuela Universitaria Politecnica (University L.P.G.C.) Spain
Espace Eolien Developpement France
Estonian Wind Power Association Estonia
Eurowind AB Sweden
Feria de Zaragoza Spain
FINE - Faroe Island New Energy Faroe Islands
Finnish Wind Power Association Finland
FME-Groep Windenergie The Netherlands
FOI - Aeronautics FFA Sweden
Fördergesellschaft Windenergie e.V (FGW) Germany
Forgital S.p.a. Italy
France Energie Eolienne France
Frisa Forjados SA de CV Mexico
Gamesa Energia S.A. Spain
Garrad Hassan and Partners Ltd UK

GE Wind Energy GmbH Germany
Germanischer Lloyd WindEnergie GmbH Germany
Gothaer Allgemeine Versicherung AG Germany
Hamburg Messe Germany
Hammonds UK
Hansen Transmissions Int. NV Belgium
Hellenic Aeolian Parks Greece
Hellenic Association of Wind Energy Investors Greece
Hellenic Wind Energy Association Greece
Hempel Paints Denmark
302
Hrvoje Pozar Energy Institute Croatia
Hungarian Wind Energy Association Hungary
Hungarian Wind Energy Scientific Association Hungary
Hydratight Sweeney Ltd United Kingdom
IED Innovation Energie Développement France
Indian Wind Energy Association India
INEGI Portugal
Institutt for energiteknikk Norway
Irish Wind Energy Association Ireland
IRO Offshore Wind Energy Group The Netherlands
ISES Italia Italy
ISET Germany
Israeli Ministry of National Infrastructures Israel
IVPC Srl Italy
Japan Wind Energy Association Japan
Japan Wind Power Association Japan
KBC Finance Ireland Ireland
KK Electronic A/S Denmark
Korean Wind Energy Research Group Korea
La Compagnie du Vent France
La Française d'Eoliennes France
Latvian Wind Energy Association Latvia
LM Glasfiber A/S Denmark
Madesta Trade Ltd Ukraine
Mammoet Van Oord BV The Netherlands
Messe Husum Germany
Metso Drives OY Finland
Moteurs Leory-Somer
National Technical University Athens
NEG Micon A/S
France
Greece
Denmark
L
APPENDIX
Netherlands Wind Energy Association (NEWIN) The Netherlands
New Zealand Wind Energy Association New Zealand
Nigerian Wind Energy Association Nigeria
Nippon Chemi-Con Corp. Germany
NMH Search UK
Nordex Energy GmbH Germany
Norwegian University of Science and Technology (NTNU) Norway
NRG Systems Inc USA
Nuon Renewable Energy Projects The Netherlands
Observ'ER France
Orix Corporate Finance Ltd United Kingdom
Orrick, Herrington & Sutcliffe United Kingdom
Owens Corning Composites Belgium
P&T Technology AG Germany
Pauwels International NV Belgium
303
PB Power UK
Plataforma Empresarial Eolica Spain
Polski Rejestr Statkow SA Poland
Power@Sea NV Belgium
Pricewaterhouse Coopers LLP UK
Promau Italy
QinetiQ United Kingdom
Ramboll Denmark
REM Chemicals, Inc. USA
Renewable Energy Systems Ltd UK
RISOE National Laboratory Denmark
Romanian Wind Energy Association Romania
Russian Association WindPower Industry (RAWI) Russia
Saint-Gobain Advanced Ceramics USA
Scanvib Denmark
SER (Syndicat des Energies Renouvelables) France
Shell WindEnergy BV The Netherlands
SIIF ENERGIES France
Slovak Association for Wind Energy Slovakia
South African Wind Energy Association South Africa
Ssesco, Inc USA
Suisse-Eole Switzerland
Swedish Wind Energy Technology Group - SWIND Sweden
TBS Shipping Services Europe GmbH Germany
Technical University of Denmark Denmark
Tech-Wise A/S Denmark
Tripod Consult APS Denmark
TTZ - Bremerhaven Germany
Turkish Wind Energy Association Turkey
Ukranian Wind Energy Association (UANE) Ukraine
United Utilities Green Energy Ltd United Kingdom
Urenco Power Technologies UK
VDMA Germany
Vergnet France
Verlinde SA France
Vestas Wind Systems A/S Denmark
Vindkraftforeningen i Finland Finland
Vindkraftsleverantörerna i Sverige Sweden
Vis Venti Poland
Volker Stevin Marine Contracting The Netherlands
WindLab Systems Australia
Windpro UK
WINDTEST Kaiser-Wilhelm-Kooj GmbH Germany
WIP Germany
Wirtshaftsverband Windkraftwerke e.V. Germany
304
WKN WINDKRAFT NORD Germany
APPENDIX M: DIRECTIVE 2001/77/EC
27.10.2001 EN Official Journal of the European Communities L 283/33
DIRECTIVE 2001/77/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

of 27 September 2001
on the promotion of electricity produced from renewable energy sources in the internal electricity
market
THE EUROPEAN PARLIAMENT AND THE COUNCIL OF THE on Climate Change, and of any policy package to meet
EUROPEAN UNION, further commitments.
Having regard to the Treaty establishing the European

Community, and in particular Article 175(1) thereof, (4) The Council in its conclusions of 11 May 1999 and the
European Parliament in its resolution of 17 June 1998
on electricity from renewable energy sources (7) have
Having regard to the proposal from the Commission (1),
invited the Commission to submit a concrete proposal
for a Community framework on access for electricity
Having regard to the opinion of the Economic and Social produced from renewable energy sources to the internal
Committee (2), market. Furthermore, the European Parliament in its
resolution of 30 March 2000 on electricity from renew-
Having regard to the opinion of the Committee of the able energy sources and the internal electricity market (8)
Regions (3), underlined that binding and ambitious renewable energy
targets at the national level are essential for obtaining
Acting in accordance with the procedure laid down in Article results and achieving the Community targets.
251 of the Treaty (4),
(5) To ensure increased market penetration of electricity
Whereas: produced from renewable energy sources in the medium
term, all Member States should be required to set
(1) The potential for the exploitation of renewable energy national indicative targets for the consumption of elec-
sources is underused in the Community at present. The tricity produced from renewable sources.
Community recognises the need to promote renewable
energy sources as a priority measure given that their
exploitation contributes to environmental protection (6) These national indicative targets should be consistent
and sustainable development. In addition this can also with any national commitment made as part of the
create local employment, have a positive impact on climate change commitments accepted by the
social cohesion, contribute to security of supply and
make it possible to meet Kyoto targets more quickly. It
is therefore necessary to ensure that this potential is
better exploited within the framework of the internal (7)
Community under the Kyoto Protocol.
The Commission should assess to what extent Member

M
APPENDIX
electricity market. States have made progress towards achieving their
national indicative targets, and to what extent the
(2) The promotion of electricity produced from renewable national indicative targets are consistent with the global
energy sources is a high Community priority as outlined indicative target of 12 % of gross domestic energy
in the White Paper on Renewable Energy Sources (here- consumption by 2010, considering that the White
inafter referred to as ‘the White Paper’) for reasons of Paper's indicative target of 12 % for the Community as a
security and diversification of energy supply, of environ- whole by 2010 provides useful guidance for increased
mental protection and of social and economic cohesion. efforts at Community level as well as in Member States,
That was endorsed by the Council in its resolution of 8 bearing in mind the need to reflect differing national
June 1998 on renewable sources of energy (5), and by circumstances. If necessary for the achievement of the
the European Parliament in its resolution on the White targets, the Commission should submit proposals to the
Paper. (6) European Parliament and the Council which may include
mandatory targets.
(3) The increased use of electricity produced from renew-
able energy sources constitutes an important part of the (8) Where they use waste as an energy source, Member
package of measures needed to comply with the Kyoto States must comply with current Community legislation
Protocol to the United Nations Framework Convention on waste management. The application of this Directive
is without prejudice to the definitions set out in Annex
(1) OJ C 311 E, 31.10.2000, p. 320 and OJ C 154 E, 29.5.2001, p. 2a and 2b to Council Directive 75/442/EEC of 15 July
89.
(2) OJ C 367, 20.12.2000, p. 5. 1975 on waste (9). Support for renewable energy sources
( ) OJ C 22, 24.1.2001, p. 27.
3 should be consistent with other Community objectives,
(4) Opinion of the European Parliament of 16 November 2000 (OJ C in particular respect for the waste treatment hierarchy.
223, 8.8.2001, p. 294), Council Common Position of 23 March
2001 (OJ C 142, 15.5.2001, p. 5) and Decision of the European 305
Parliament of 4 July 2001 (not yet published in the Official (7) OJ C 210, 6.7.1998, p. 143.
Journal). Council Decision of 7 September 2001. (8) OJ C 378, 29.12.2000, p. 89.
(5) OJ C 198, 24.6.1998, p. 1. (9) OJ L 194, 25.7.1975, p. 39. Directive as last amended by Commis-
( ) OJ C 210, 6.7.1998, p. 215.
6
sion Decision 96/350/EC (OJ L 135, 6.6.1996, p. 32).
L 283/34 EN Official Journal of the European Communities 27.10.2001
Therefore, the incineration of non-separated municipal report on experience gained with the application of
waste should not be promoted under a future support national schemes. If necessary, the Commission should,
system for renewable energy sources, if such promotion in the light of the conclusions of this report, make a
were to undermine the hierarchy. proposal for a Community framework with regard to
support schemes for electricity produced from renew-
able energy sources. That proposal should contribute to
(9) The definition of biomass used in this Directive does not the achievement of the national indicative targets, be
prejudge the use of a different definition in national compatible with the principles of the internal electricity
legislation, for purposes other than those set out in this market and take into account the characteristics of the
Directive. different sources of renewable energy, together with the
different technologies and geographical differences. It
(10) This Directive does not require Member States to recog- should also promote the use of renewable energy
nise the purchase of a guarantee of origin from other sources in an effective way, and be simple and at the
Member States or the corresponding purchase of elec- same time as efficient as possible, particularly in terms
tricity as a contribution to the fulfilment of a national of cost, and include sufficient transitional periods of at
quota obligation. However, to facilitate trade in elec- least seven years, maintain investors' confidence and
tricity produced from renewable energy sources and to avoid stranded costs. This framework would enable elec-
increase transparency for the consumer's choice between tricity from renewable energy sources to compete with
electricity produced from non-renewable and electricity electricity produced from non-renewable energy sources
produced from renewable energy sources, the guarantee and limit the cost to the consumer, while, in the
of origin of such electricity is necessary. Schemes for the medium term, reduce the need for public support.
guarantee of origin do not by themselves imply a right
to benefit from national support mechanisms established
in different Member States. It is important that all forms (17) Increased market penetration of electricity produced
of electricity produced from renewable energy sources from renewable energy sources will allow for economies
are covered by such guarantees of origin. of scale, thereby reducing costs.
(11) It is important to distinguish guarantees of origin clearly

from exchangeable green certificates.
(18) It is important to utilise the strength of the market
forces and the internal market and make electricity
(12) The need for public support in favour of renewable produced from renewable energy sources competitive
energy sources is recognised in the Community guide- and attractive to European citizens.
lines for State aid for environmental protection (1),
which, amongst other options, take account of the need
to internalise external costs of electricity generation.
However, the rules of the Treaty, and in particular (19) When favouring the development of a market for renew-
Articles 87 and 88 thereof, will continue to apply to able energy sources, it is necessary to take into account
such public support. the positive impact on regional and local development
opportunities, export prospects, social cohesion and
employment opportunities, especially as concerns small
(13) A legislative framework for the market in renewable and medium-sized undertakings as well as independent
energy sources needs to be established. power producers.
(14) Member States operate different mechanisms of support

for renewable energy sources at the national level, (20) The specific structure of the renewable energy sources
including green certificates, investment aid, tax exemp- sector should be taken into account, especially when
tions or reductions, tax refunds and direct price support reviewing the administrative procedures for obtaining
schemes. One important means to achieve the aim of permission to construct plants producing electricity
this Directive is to guarantee the proper functioning of from renewable energy sources.
these mechanisms, until a Community framework is put
into operation, in order to maintain investor confidence.
(21) In certain circumstances it is not possible to ensure fully
(15) It is too early to decide on a Community-wide frame- transmission and distribution of electricity produced
work regarding support schemes, in view of the limited from renewable energy sources without affecting the
experience with national schemes and the current rela- reliability and safety of the grid system and guarantees in
tively low share of price supported electricity produced this context may therefore include financial compensa-
from renewable energy sources in the Community. tion.
(16) It is, however necessary to adapt, after a sufficient trans-

itional period, support schemes to the developing (22) The costs of connecting new producers of electricity
306 internal electricity market. It is therefore appropriate that from renewable energy sources should be objective,
the Commission monitor the situation and present a transparent and non-discriminatory and due account
should be taken of the benefit embedded generators
(1) OJ C 37, 3.2.2001, p. 3. bring to the grid.
(23) Since the general objectives of the proposed action Article 3

cannot be sufficiently achieved by the Member States
and can therefore, by reason of the scale or effects of the
action, be better achieved at Community level, the National indicative targets
Community may adopt measures, in accordance with
the principle of subsidiarity as set out in Article 5 of the
Treaty. Their detailed implementation should, however, 1. Member States shall take appropriate steps to encourage
be left to the Member States, thus allowing each Member greater consumption of electricity produced from renewable
State to choose the regime which corresponds best to its energy sources in conformity with the national indicative
particular situation. In accordance with the principle of targets referred to in paragraph 2. These steps must be in
proportionality, as set out in that Article, this Directive proportion to the objective to be attained.
does not go beyond what is necessary in order to
achieve those objectives,
2. Not later than 27 October 2002 and every five years
thereafter, Member States shall adopt and publish a report
setting national indicative targets for future consumption of
HAVE ADOPTED THIS DIRECTIVE:
electricity produced from renewable energy sources in terms of
a percentage of electricity consumption for the next 10 years.
The report shall also outline the measures taken or planned, at
national level, to achieve these national indicative targets. To
Article 1 set these targets until the year 2010, the Member States shall:
— take account of the reference values in the Annex,
Purpose
— ensure that the targets are compatible with any national
The purpose of this Directive is to promote an increase in the commitments accepted in the context of the climate change
contribution of renewable energy sources to electricity produc- commitments accepted by the Community pursuant to the
tion in the internal market for electricity and to create a basis Kyoto Protocol to the United Nations Framework Conven-
for a future Community framework thereof. tion on Climate Change.
Article 2 3. Member States shall publish, for the first time not later
than 27 October 2003 and thereafter every two years, a report
which includes an analysis of success in meeting the national
Definitions indicative targets taking account, in particular, of climatic
factors likely to affect the achievement of those targets and
For the purposes of this Directive, the following definitions
shall apply:
(a) ‘renewable energy sources’ shall mean renewable non-fossil

which indicates to what extent the measures taken are consis-
tent with the national climate change commitment.
M
energy sources (wind, solar, geothermal, wave, tidal, hydro- 4. On the basis of the Member States' reports referred to in
APPENDIX
power, biomass, landfill gas, sewage treatment plant gas paragraphs 2 and 3, the Commission shall assess to what
and biogases); extent:
(b) ‘biomass’ shall mean the biodegradable fraction of products, — Member States have made progress towards achieving their
waste and residues from agriculture (including vegetal and national indicative targets,
animal substances), forestry and related industries, as well
as the biodegradable fraction of industrial and municipal — the national indicative targets are consistent with the global
waste; indicative target of 12 % of gross national energy consump-

tion by 2010 and in particular with the 22,1 % indicative
(c) ‘electricity produced from renewable energy sources’ shall mean share of electricity produced from renewable energy
electricity produced by plants using only renewable energy sources in total Community electricity consumption by
sources, as well as the proportion of electricity produced 2010.
from renewable energy sources in hybrid plants also using
conventional energy sources and including renewable elec-
tricity used for filling storage systems, and excluding elec- The Commission shall publish its conclusions in a report, for
tricity produced as a result of storage systems; the first time not later than 27 October 2004 and thereafter
every two years. This report shall be accompanied, as appro-
(d) ‘consumption of electricity’ shall mean national electricity priate, by proposals to the European Parliament and to the
production, including autoproduction, plus imports, minus Council.
exports (gross national electricity consumption).
In addition, the definitions in Directive 96/92/EC of the Euro- If the report referred to in the second subparagraph concludes
pean Parliament and of the Council of 19 December 1996 that the national indicative targets are likely to be inconsistent,
concerning common rules for the internal market of elec- for reasons that are unjustified and/or do not relate to new 307
tricity (1) shall apply. scientific evidence, with the global indicative target, these
proposals shall address national targets, including possible
(1) OJ L 27, 30.1.1997, p. 20. mandatory targets, in the appropriate form.
Article 4 — serve to enable producers of electricity from renewable

energy sources to demonstrate that the electricity they sell
Support schemes is produced from renewable energy sources within the
meaning of this Directive.
1. Without prejudice to Articles 87 and 88 of the Treaty,
the Commission shall evaluate the application of mechanisms
used in Member States according to which a producer of 4. Such guarantees of origin, issued according to paragraph
electricity, on the basis of regulations issued by the public 2, should be mutually recognised by the Member States, exclu-
authorities, receives direct or indirect support, and which could sively as proof of the elements referred to in paragraph 3. Any
have the effect of restricting trade, on the basis that these refusal to recognise a guarantee of origin as such proof, in
contribute to the objectives set out in Articles 6 and 174 of the particular for reasons relating to the prevention of fraud, must
Treaty. be based on objective, transparent and non-discriminatory
criteria. In the event of refusal to recognise a guarantee of
origin, the Commission may compel the refusing party to
2. The Commission shall, not later than 27 October 2005,
recognise it, particularly with regard to objective, transparent
present a well-documented report on experience gained with
and non-discriminatory criteria on which such recognition is
the application and coexistence of the different mechanisms
based.
referred to in paragraph 1. The report shall assess the success,
including cost-effectiveness, of the support systems referred to
in paragraph 1 in promoting the consumption of electricity 5. Member States or the competent bodies shall put in place
produced from renewable energy sources in conformity with appropriate mechanisms to ensure that guarantees of origin are
the national indicative targets referred to in Article 3(2). This both accurate and reliable and they shall outline in the report
report shall, if necessary, be accompanied by a proposal for a referred to in Article 3(3) the measures taken to ensure the
Community framework with regard to support schemes for reliability of the guarantee system.
electricity produced from renewable energy sources.
Any proposal for a framework should: 6. After having consulted the Member States, the Commis-
sion shall, in the report referred to in Article 8, consider the
(a) contribute to the achievement of the national indicative form and methods that Member States could follow in order to
targets; guarantee the origin of electricity produced from renewable
energy sources. If necessary, the Commission shall propose to
(b) be compatible with the principles of the internal electricity the European Parliament and the Council the adoption of
market; common rules in this respect.
(c) take into account the characteristics of different sources of
renewable energy, together with the different technologies,
and geographical differences;
Article 6
(d) promote the use of renewable energy sources in an effec-
tive way, and be simple and, at the same time, as efficient
as possible, particularly in terms of cost; Administrative procedures
(e) include sufficient transitional periods for national support
systems of at least seven years and maintain investor confi- 1. Member States or the competent bodies appointed by the
dence. Member States shall evaluate the existing legislative and regula-
tory framework with regard to authorisation procedures or the
other procedures laid down in Article 4 of Directive 96/92/EC,
Article 5 which are applicable to production plants for electricity
produced from renewable energy sources, with a view to:
Guarantee of origin of electricity produced from renew- — reducing the regulatory and non-regulatory barriers to the
able energy sources increase in electricity production from renewable energy
sources,
1. Member States shall, not later than 27 October 2003,
ensure that the origin of electricity produced from renewable — streamlining and expediting procedures at the appropriate
energy sources can be guaranteed as such within the meaning administrative level, and
of this Directive according to objective, transparent and non- — ensuring that the rules are objective, transparent and non-
discriminatory criteria laid down by each Member State. They discriminatory, and take fully into account the particulari-
shall ensure that a guarantee of origin is issued to this effect in ties of the various renewable energy source technologies.
response to a request.
2. Member States may designate one or more competent 2. Member States shall publish, not later than 27 October
bodies, independent of generation and distribution activities, to 2003, a report on the evaluation referred to in paragraph 1,
supervise the issue of such guarantees of origin. indicating, where appropriate, the actions taken. The purpose
of this report is to provide, where this is appropriate in the
3. A guarantee of origin shall: context of national legislation, an indication of the stage
reached specifically in:
308 — specify the energy source from which the electricity was
produced, specifying the dates and places of production, — coordination between the different administrative bodies as
and in the case of hydroelectric installations, indicate the regards deadlines, reception and treatment of applications
capacity; for authorisations,
— drawing up possible guidelines for the activities referred to 5. Member States shall put into place a legal framework or
in paragraph 1, and the feasibility of a fast-track planning require transmission system operators and distribution system
procedure for producers of electricity from renewable operators to set up and publish their standard rules relating to
energy sources, and the sharing of costs of system installations, such as grid
connections and reinforcements, between all producers bene-
— the designation of authorities to act as mediators in fiting from them.
disputes between authorities responsible for issuing author-
isations and applicants for authorisations.
The sharing shall be enforced by a mechanism based on objec-
tive, transparent and non-discriminatory criteria taking into
3. The Commission shall, in the report referred to in Article account the benefits which initially and subsequently connected
8 and on the basis of the Member States' reports referred to in producers as well as transmission system operators and
paragraph 2 of this Article, assess best practices with a view to distribution system operators derive from the connections.
achieving the objectives referred to in paragraph 1.
6. Member States shall ensure that the charging of transmis-
sion and distribution fees does not discriminate against elec-
tricity from renewable energy sources, including in particular
electricity from renewable energy sources produced in periph-
Article 7 eral regions, such as island regions and regions of low popula-
tion density.
Grid system issues

Where appropriate, Member States shall put in place a legal
framework or require transmission system operators and
distribution system operators to ensure that fees charged for
1. Without prejudice to the maintenance of the reliability the transmission and distribution of electricity from plants
and safety of the grid, Member States shall take the necessary using renewable energy sources reflect realisable cost benefits
measures to ensure that transmission system operators and resulting from the plant's connection to the network. Such cost
distribution system operators in their territory guarantee the benefits could arise from the direct use of the low-voltage grid.
transmission and distribution of electricity produced from
renewable energy sources. They may also provide for priority
access to the grid system of electricity produced from renew- 7. Member States shall, in the report referred to in Article
able energy sources. When dispatching generating installations, 6(2), also consider the measures to be taken to facilitate access
transmission system operators shall give priority to generating to the grid system of electricity produced from renewable
installations using renewable energy sources insofar as the energy sources. That report shall examine, inter alia, the feasi-
operation of the national electricity system permits.
M
bility of introducing two-way metering.
2. Member States shall put into place a legal framework or

require transmission system operators and distribution system
APPENDIX
Article 8
operators to set up and publish their standard rules relating to
the bearing of costs of technical adaptations, such as grid
connections and grid reinforcements, which are necessary in Summary report
order to integrate new producers feeding electricity produced
from renewable energy sources into the interconnected grid.
On the basis of the reports by Member States pursuant to
Article 3(3) and Article 6(2), the Commission shall present to
These rules shall be based on objective, transparent and non- the European Parliament and the Council, no later than 31
discriminatory criteria taking particular account of all the costs December 2005 and thereafter every five years, a summary
and benefits associated with the connection of these producers report on the implementation of this Directive.
to the grid. The rules may provide for different types of
connection.
This report shall:
— consider the progress made in reflecting the external costs
3. Where appropriate, Member States may require transmis- of electricity produced from non-renewable energy sources
sion system operators and distribution system operators to and the impact of public support granted to electricity
bear, in full or in part, the costs referred to in paragraph 2. production,
— take into account the possibility for Member States to meet
the national indicative targets established in Article 3(2), the
4. Transmission system operators and distribution system global indicative target referred to in Article 3(4) and the
operators shall be required to provide any new producer existence of discrimination between different energy
wishing to be connected with a comprehensive and detailed sources.
estimate of the costs associated with the connection. Member 309
States may allow producers of electricity from renewable
energy sources wishing to be connected to the grid to issue a If appropriate, the Commission shall submit with the report
call for tender for the connection work. further proposals to the European Parliament and the Council.
Article 9
Transposition
Member States shall bring into force the laws, regulations and administrative provisions necessary to
comply with this Directive not later than 27 October 2003. They shall forthwith inform the Commission
thereof.
When Member States adopt these measures, they shall contain a reference to this Directive or shall be
accompanied by such a reference on the occasion of their official publication. The methods of making such
reference shall be laid down by the Member States.
Article 10
Entry into force
This Directive shall enter into force on the day of its publication in the Official Journal of the European
Communities.
Article 11
Addressees
This Directive is addressed to the Member States.
Done at Brussels, 27 September 2001.
For the European Parliament For the Council

The President The President
N. FONTAINE C. PICQUÉ
310
ANNEX
Reference values for Member States' national indicative targets for the contribution of electricity produced from
renewable energy sources to gross electricity consumption by 2010 (*)
This Annex gives reference values for the fixing of national indicative targets for electricity produced from renewable
energy sources (‘RES-E’), as referred to in Article 3(2):
RES-E TWh 1997 (**) RES-E % 1997 (***) RES-E % 2010 (***)
Belgium 0,86 1,1 6,0
Denmark 3,21 8,7 29,0
Germany 24,91 4,5 12,5
Greece 3,94 8,6 20,1
Spain 37,15 19,9 29,4
France 66,00 15,0 21,0
Ireland 0,84 3,6 13,2
Italy 46,46 16,0 25,0 (1)
Luxembourg 0,14 2,1 5,7 (2)
Netherlands 3,45 3,5 9,0
Austria 39,05 70,0 78,1 (3) M
APPENDIX
Portugal 14,30 38,5 39,0 (4)
Finland 19,03 24,7 31,5 (5)
Sweden 72,03 49,1 60,0 (6)

United Kingdom 7,04 1,7 10,0
Community 338,41 13,9 % 22 % (****)
(*) In taking into account the reference values set out in this Annex, Member States make the necessary assumption that the State aid
guidelines for environmental protection allow for the existence of national support schemes for the promotion of electricity produced
from renewable energy sources.
(**) Data refer to the national production of RES-E in 1997.
(***) The percentage contributions of RES-E in 1997 and 2010 are based on the national production of RES-E divided by the gross national
electricity consumption. In the case of internal trade of RES-E (with recognised certification or origin registered) the calculation of
these percentages will influence 2010 figures by Member State but not the Community total.
(****) Rounded figure resulting from the reference values above.
( ) Italy states that 22 % would be a realistic figure, on the assumption that in 2010 gross national electricity consumption will be 340 TWh.
1
When taking into account the reference values set out in this Annex, Italy has assumed that gross national electricity production from
renewable energy sources will attain up to 76 TWh in 2010. This figure includes the contribution of the non-biodegradable fraction of
municipal and industrial waste used in compliance with Community legislation on waste management.
In this respect, the capability to reach the indicative target as referred to in this Annex, is contingent, inter alia, upon the effective level of
the national demand for electric energy in 2010.
( 2) Taking into account the indicative reference values set out in this Annex, Luxembourg takes the view that the objective set for 2010 can be
311
achieved only if:
— total electricity consumption in 2010 does not exceed that of 1997,
— wind-generated electricity can be multiplied by a factor of 15,
— biogas-generated electricity can be multiplied by a factor of 208,

— electricity produced from the only municipal waste incinerator in Luxembourg, which in 1997 accounted for half the electricity
produced from renewable energy sources, can be taken into account in its entirety,
— photovoltaically generated electricity can be raised to 80 GWh, and
in so far as the above points can be achieved from the technical standpoint in the time allowed.
In the absence of natural resources, an additional increase in electricity generated by hydroelectric power stations is ruled out.
(3) Austria states that 78,1 % would be a realistic figure, on the assumption that in 2010 gross national electricity consumption will be
56,1 TWh. Due to the fact that the production of electricity from renewable sources is highly dependent on hydropower and therefore
on the annual rainfall, the figures for 1997 and 2010 should be calculated on a long-range model based on hydrologic and climatic
conditions.
(4) Portugal, when taking into account the reference values, set out in this Annex, states that to maintain the 1997 share of electricity
produced from renewable sources as an indicative target for 2010 it was assumed that:
— it will be possible to continue the national electricity plan building new hydro capacity higher than 10 MW,
— other renewable capacity, only possible with financial state aid, will increase at an annual rate eight times higher than has occurred
recently.
These assumptions imply that new capacity for producing electricity from renewable sources, excluding large hydro, will increase at a rate
twice as high as the rate of increase of gross national electricity consumption.
(5) In the Finnish action plan for renewable energy sources, objectives are set for the volume of renewable energy sources used in 2010.
These objectives have been set on the basis of extensive background studies. The action plan was approved within the Government in
October 1999.
According to the Finnish action plan, the share of electricity produced from renewable energy sources by 2010 would be 31 %. This
indicative target is very ambitious and its realisation would require extensive promotion measures in Finland.
(6) When taking into account the reference values set out in this Annex, Sweden notes that the possibility of reaching the target is highly
dependent upon climatic factors heavily affecting the level of hydropower production, in particular variations in pluviometry, timing of
rainfall during the year and inflow. The electricity produced from hydropower can vary substantially. During extremely dry years
production may amount to 51 TWh, whereas in wet years it could amount to 78 TWh. The figure for 1997 should thus be calculated
with a long-range model based on scientific facts on hydrology and climatic change.
It is a generally applied method in countries with important shares of hydropower production to use water inflow statistics covering a
time span of 30 to 60 years. Thus, according to the Swedish methodology and based on conditions during the period 1950-1999,
correcting for differences in total hydropower production capacity and inflow over the years, average hydropower production amounts to
64 TWh which corresponds to a figure for 1997 of 46 %, and in this context Sweden considers 52 % to be a more realistic figure for
2010.
Furthermore, the ability of Sweden to achieve the target is limited by the fact that the remaining unexploited rivers are protected by law.
Moreover, the ability of Sweden to reach the target is heavily contingent upon:
— the expansion of combined heat and power (CHP) depending on population density, demand for heat and technology development,
in particular for black liquor gasification, and
— authorisation for wind power plants in accordance with national laws, public acceptance, technology development and expansion of
grids.
312
GLOSSARY
A Boundary Layer The layer of the atmosphere in which

interaction with the Earth’s surface influences air flow pat-
Active Stall Another term for pitch-to-stall WT power con- terns. The thickness of the boundary layer varies between
trol (see Pitch Regulation). about 100 m on clear nights with low wind speeds to 2
Air Density The power output of a WT at a given wind km on fine summer’s days.
speed increases with increasing air density, ie. the mass
of the air per cubic metre of space at the WT location. C
Anemometer An instrument for measuring wind speed.
Ultrasonic, laser, hot wire and cup varieties are available: Carbon Dioxide (CO2) CO2 is “a naturally occurring gas,
cup anemometers (consisting of a number of cups and also a by-product of burning fossil fuels and biomass,
attached to an axle) are almost invariably used throughout as well as land-use changes and other industrial process-
the wind industry. Good quality anemometers are vital for es. It is the principal anthropogenic greenhouse gas that
accurate wind resource assessment. affects the Earth’s radiative balance” (IPCC, 2001; p.70).
Angle of Attack The angle between the oncoming wind Capacity Credit A WT can only produce when the wind
and the blade chord. blows and, therefore, it is not directly comparable to a
Aspect Ratio The ratio of the characteristic chord of the conventional power plant. The capacity credit is the per-
blade to its length. centage of conventional capacity that a given turbine can
Auxiliary Costs Costs other than those of the turbine itself, replace. A typical value of the capacity credit is 25% (see
ie. foundation, grid-connection, electrical installation, road Capacity Factor), ie. 100 MW of wind power is assumed
construction, consultancy, financial charges, etc. to replace 25 MW of conventional fossil fuelled capacity.
Availability The availability of a WT describes the amount of Capacity Factor The amount of energy a WT actually pro-
the time that it is actually functional, not out of order or duces in a year, divided by the amount of energy if could
being serviced. Clearly, the higher the availability the better. theoretically produce if it were to run at its rated power 24
hours a day, 365 days a year. The capacity factor is
B expressed as a percentage; for a typical WT installation it
GLOSSARY
is around 25%-30% (although it can get up to 50%).
Betz Limit The theoretical limiting efficiency of a WT Capital Costs The total investment cost of the turbine,
(typically about 59%). including auxiliary costs.
Blade The blades of a WT are the wing-like structures CENELEC The European Committee for Electrotechnical
attached to the hub. A blade is very similar in look and func- Standardisation, a non-profit technical organisation com-
tion to an aeroplane wing - air flow past the blade generates posed of the National Electrotechnical Committees of 23
aerodynamic lift which, in the case of a WT, causes the rotor European countries. CENELEC’s mission is to prepare vol-
to spin and makes power generation possible. Most modern untary electrotechnical standards to help develop the
electricity-generating WTs have three blades (the so-called Single European Market/European Economic Area for
“Danish concept”). However, many two-bladed examples electrical and electronic goods and services, so removing
also exist, and even one-bladed designs have been known. barriers to trade, creating new markets and cutting com-
Rotors for water pumping and battery charging may have pliance costs.
many more blades to allow them to function in low winds. Certification Authority An organisation which checks the
Blade Element-Momentum Theory (BEM) An aerodynamic designs of WT manufacturers and issues an independent
theory linking the drag and lift forces experienced by each assessment, often called a certificate. The most promi-
section of a WT blade to the change in momentum of air nent are Det Norsk Veritas and Germanischer Lloyd.
passing through the rotor disc. Climate Change Defined as “a change of climate which
Blade Passing Frequency The frequency at which the is attributed directly or indirectly to human activity that
313
blades of a WT pass the tower. For a three-bladed WT, this alters the composition of the global atmosphere and
will be three times the rotational frequency. which is in addition to natural climate variability
observed over comparable time periods” (IPCC, 2001; ments (eg. air temperature and pressure gauges) at a
p. 71). monitoring station. Some data loggers store data on site,
Control Strategy Modern WTs feature computer control others are able to relay it to distant computer systems.
(the controller) whereby the computer will execute a given Dinotail An “add-on” device sometimes attached to a
control strategy, telling the WT how to behave under differ- stall-regulated turbine blade which enhances the lift on a
ent conditions. Examples of components of a control strat- section. It extends backwards from the trailing edge and
egy would be the cut-in and cut-out wind speeds, when the looks rather like a dinosaur tail.
WT should cut out in low, falling wind speeds and when it Direct-Drive A new generation of WTs has recently emerged
should cut back in at high, falling wind speeds when it has where the rotor is connected directly on a single shaft to a
already cut out (see Hysteresis). Depending on the type of special high-torque, low speed generator without the use of
WT (fixed/variable speed, pitch/stall-regulated), there will a gear box. Such WTs are generally variable speed and fea-
be other more complicated logic in the control strategy. ture power electronic converters to convert the frequency of
Controller The computer equipment that monitors the tur- the generated power to the grid frequency. Direct-drive WTs
bine and controls its operation. offer higher efficiency and lower noise levels due to the
Coriolis Force A virtual force experienced by a body rela- absence of a gear box, but the complex generator and
tive to a rotating body as it moves outward from the cen- power electronics may make them more expensive.
tre of rotation. For example, the Coriolis force causes air Direct Employment Direct employment is the total num-
flowing southwards in the northern hemisphere to veer ber of people (skilled, unskilled and self-employed)
westwards relative to the rotating globe below; hence it employed in companies belonging to a specific sector, ie.
has a huge influence on the behaviour of the atmosphere. WT manufacturing.
It is a consequence of the principle of conservation of Discount Rate The interest rate used to calculate the
angular momentum (but should be distinguished from present-day cost of turbine installations.
gyroscopic effects). Diurnal A term describing a daily frequency, ie. a period of
Coriolis Parameter A parameter describing the strength of 24 hours. Some weather effects occur on a diurnal basis,
the Coriolis force at any point on the globe, featuring in associated with the daily heating and cooling of the
the log law. Earth’s surface.
Costs of Generated Wind Power See Levillised Costs. Doubly Fed Induction Generator (DFIG) The DFIG pro-
Cut-In and Cut-Out Wind Speeds A WT is usually provides the operational advantages of an induction and a
grammed only to start operating above a particular wind synchronous machine. It offers variable speed operation
speed (the cut-in wind speed) and to stop operating when and the advantage of reactive power control. This is
the wind speed exceeds another particular wind speed achieved by injecting appropriate currents into the rotor
(the cut-out- wind speed). There is very little energy avail- circuit of a wound rotor induction generator. Control of the
able below the cut-in wind speed, making operation of the frequency of the injected currents provides variable speed
WT unviable or impossible. Above the cut-out wind speed control of the machine, whereas manipulation of the mag-
(typically 25 m/s) the loads on the WT could cause dam- nitude and phase of the rotor currents provide power fac-
age if it did not shut down. Though such high-speed winds tor control.
are clearly very energetic, they are experienced so rarely Dynamic Stall As WT blades rotate, they experience con-
at most locations that very little energy is wasted by cut- stantly changing angles of attack due to wind shear and
ting out at these speeds. yaw. Under these conditions, a phenomenon known as
dynamic stall (as opposed to static stall) occurs. The
angle of attack will increase beyond the point at which the
D blade would normally have stalled in a quasi-static situa-
tion in a wind tunnel, but the delayed stall is very sudden
314
Data Logger The electronic equipment used to record the and hard when it occurs at some large angle of attack,
output of anemometers, wind vanes and other instru- causing large loads in the blades and significant fatigue.
E a negative angle of attack which will rapidly slow the rotor
to a standstill. This effect can be used to assist WT cut-
Edgewise Used to define the direction in an axis set in the out in very high winds and as an emergency brake. In very
rotating blade. Edgewise motion is motion in the plane of high winds, if the WT can be parked with the blades feath-
rotation and is perpendicular to flatwise motion. ered, this will reduce the loads on the whole WT.
Efficiency In general of any component, this describes the Fixed Speed A fixed speed WT will, once started, always
amount of energy coming out of the component as a per- rotate at the same speed, regardless of the wind speed.
centage of the energy put into it. For a WT, it describes As power must be injected at a constant frequency onto
the amount of active electrical power generated as a per- the grid, operating at a fixed speed simplifies the genera-
centage of the wind power incident on the rotor area (see tor and power electronic requirements for the WT consid-
also Betz Limit). erably, making it cheaper. However, as the WT will only be
Emissions Defined by the UN (2002) as “the discharge of able to operate at maximum efficiency at one particular
pollutants into the atmosphere from stationary sources wind speed, fixed speed WTs are around 10% less effi-
such as smokestacks, other vents, surface areas of com- cient than variable ones. A compromise is to use a WT
mercial or industrial facilities and mobile sources, for that can operate at two different fixed speeds.
example, motor vehicles, locomotives and aircraft”. With Flatwise Used to define the direction in an axis set in the
respect to climate change, emissions refer to “the rotating blade. Flatwise motion is motion perpendicular to
release of greenhouse gases into the atmosphere over a both the plane of rotation and to edgewise motion.
specified area and period of time” (IPCC, 2001; p. 72). Forced Yaw See Motor Yaw.
Energy Pay-Back Period The amount of time it takes for a Free Stream Used to describe wind conditions at a loca-
WT to generate as much energy is required to make the tion in the absence of the WT.
WT in the first place, install it, maintain it throughout its Free Yaw Rotating WT rotors will tend to align themselves
lifetime and, finally, scrap it(typically two to three months to face the oncoming wind (even upwind rotors will do
at a site with reasonable wind exposure). this). A WT which allows this natural tendency to align the
Experience Curve The curve relates the cumulative quan- rotor is known as a free yaw WT. This mode of operation
GLOSSARY
titative development of a product with the development of is not popular in large WTs due to the possibility of dam-
the specific costs. Thus, the more that is produced of a age from gyroscopic forces. Also, rotors will not recover
product, the more efficient the production process and from positions of large yaw misalignment, for example
the cheaper it becomes. after periods of calm or sudden large changes in wind
External Costs Those costs incurred in activities which direction.
may “cause damage to a wide range of receptors, includ- Friction Velocity A parameter featuring in the log law.
ing human health, natural ecosystems and the built envi- Fuel Cycle The impacts of power production are not exclu-
ronment” (European Community, 1994), and yet are not sively generated during the operation of the power plant,
reflected in the price paid by consumers. but also in the entire chain of activities needed for the
electricity production and distribution, such as fuel extrac-
F tion, processing and transformation, construction and
installation of the equipment as well as the disposal of
Fatigue The phenomenon by which a repeated loading waste. These stages, which constitute the chain of elec-
and unloading of a structure causes its various compo- tricity production and distribution, are known as the fuel
nents to gradually weaken and eventually fail. Owing to cycle.
the highly repetitive nature of WT operation, fatigue is a Full Load Hours The turbine’s average annual production
serious issue. divided by its rated power. The higher the number of full
Feather Blade feathering is possible on WTs with load hours, the higher the turbine’s production at the cho-
315
adjustable pitch. The blades are rotated so the chord is sen site.
pointing upwind which, when the blade is in motion, gives
G H
Gearbox The gearbox in a WT is used to convert the low Harmonic Ideally, the alternating electrical signal output
speed, high torque rotation of the rotor to a high speed, by any device onto the electrical grid should be perfectly
low torque rotation suitable for driving the generator to sinusoidal in form. A regular, non-sinusoidal signal can be
produce alternating current at the correct grid frequency. made up of sinusoidal components at integer multiples of
Not all modern WTs have gearboxes (see Direct Drive). the so-called base frequency. These higher frequency
Generator (Synchronous, Induction) The generator is the components are known as harmonics, and whilst they
piece of electrical machinery that actually converts the make music sound nice, they cause problems in electrical
rotating motion of the rotor into alternating current elec- networks (see Power Quality). The sinusoidal output from
trical power. Most fixed (or partially variable) speed WTs a power electronic converter (PEC) is digitally synthesised
have induction generators, whereas variable speed WTs and hence is not a perfect sinusoid. Older PECs would
have synchronous generators. The principle components inject significant harmonics on to the electrical grid; more
of a generator are the rotor and the stator. modern ones do not.
Geostrophic Wind Winds higher in the atmosphere driven High-Speed Shaft In WTs featuring gearboxes, the high-
purely by temperature and pressure differences within the speed shaft is the shaft connecting the gearbox to the
atmosphere, and unaffected by the surface of the Earth. generator. This shaft also features lower torques.
Glass-Reinforced Plastic (GRP) The material from which High Voltage Direct Current (HVDC) It is proposed that in
most WT blades are made. GRP consists of a web of glass linking offshore wind farms to land by buried cable, HVDC
fibres set into solid plastic (polyester is often used). links will be used instead of alternating current power
Greenhouse Gas “Greenhouse gases are those gaseous transmission.
constituents of the atmosphere, both natural and anthro- Hill Effect Wind speeds up as it passes over the top of
pogenic, that absorb and emit radiation at specific wave- hills, making hilltops good sites for WTs. Note, however,
lengths within the spectrum of infrared radiation emitted by that in extreme cases negative wind shear may occur (ie.
the earth’s surface, the atmosphere, and clouds. Entirely the wind closer to the ground on the hilltop is faster than
human made greenhouse gases in the atmosphere such as that higher up) and complex hilly terrain may cause addi-
halocarbons and other chlorine- and bromine-containing sub- tional turbulence and flow separation.
stances are dealt under the Montreal Protocol. Beside car- Horizontal Axis Wind Turbine (HAWT) A generic descrip-
bon dioxide, nitrous oxide and methane, the Kyoto protocol tion of the propeller-type WTs seen throughout the world
deals with the greenhouse gases sulphur hexafluoride, hydro- (see also Vertical Axis Wind Turbine).
fluorocarbons, and perfluorocarbons” (IPCC, 2001; p. 74). Hub The hub of a WT is the rotating component to which
Grid-Connected A WT is said to be grid-connected when the blades are fixed.
its output is channelled directly into a national electric Hub Height The height of the rotor axis above the ground.
grid or the like (see also Stand-Alone). Hysteresis In general, this term describes a process that
Grid Reinforcement A weak grid can be reinforced by does not proceed in the same way when run in different
uprating its connection to the rest of the grid. The cost of directions; a simple example would be that you do not get
doing this may fall to the wind farm developer. the same amount of energy out of an elastic band when
Gurney Flaps Aerodynamic devices which are used to you allow it to relax as when you stretch it (some energy
enhance lift on a stall-regulated blade. goes to heat the elastic band). Hysteresis is often used in
Gyroscopic Effects A rotating WT rotor will experience wind energy to describe the way in which, whilst a WT may
gyroscopic forces if direction of the axis of rotation is cut out when the wind speed reaches 25m/s, it will not
shifted (stemming from the principle of conservation of cut back in again until the wind speed drops below (for
angular momentum). If the rotor is allowed to yaw in an example) 23m/s. There is also a hysteresis loop involved
316
uncontrolled manner, these forces can cause large loads in blade behaviour in dynamic situations (see Dynamic
in the WT, potentially causing failures. Stall).
I Islanding The phenomenon by which a piece of the elec-
trical network which has become disconnected from the
Impact Pathway Approach The impact pathway approach rest of the network (for instance due to a lightning strike
developed by ExternE establishes the effects and spatial on an overhead pylon) continues to function, as the power
distribution of the burdens from the fuel cycle to find out sources on that section of the grid (for example a wind
their final impact on health and the environment. farm) continue to supply the loads. Islanding is a serious
Subsequently, the economic valuation assigns the respec- problem that can result in danger to personnel and dam-
tive costs of the damages induced by a given activity. age to equipment; a number of safeguards are used to try
Indirect Employment Refers to those employed in sectors to prevent it.
and activities supplying intermediate products/components
to, for example, WT manufacturers. Indirect employment L
includes employment throughout the production chain.
Induced Velocity The change in the velocity of the wind Laminar Laminar air flow is where layers of air moving at
caused by the presence of a WT blade. different speeds slide smoothly across one another with-
Input-Output The national accounts of a country’s or out mixing, for example around the leading edge of an
region’s economic transactions keep track of all the unstalled WT blade.
inputs and outputs between economic sectors. Leading Edge The blunter edge of a WT blade; this edge
Insulated Gate Bipolar Transistor (IGBT) The IGBT has the is the one that moves forwards through the oncoming
output switching and conduction characteristics of a bipo- wind.
lar transistor but is voltage-controlled like a MOSFET. Learning Rate A learning curve parameter. The learning
IGBTs are used in power switching applications where rate is estimated on available data for WTs; it tells you the
high current handling and ease of control are desired. achieved reduction in specific product costs. Thus, if the
Intermedial Load Intermedial load refers to those elec- learning rate is 15%, then costs are assumed to be
tricity generation technologies contributing to satisfy the reduced by this percentage when total installations of
demand in a range between base load and peak load of WTs are doubled.
GLOSSARY
the electricity system. A generating unit that normally Levellised Costs The present-day average cost per kWh
operates at a constant output (amount of electricity deliv- produced by the turbine over its entire lifetime, including
ered) for several hours, eg. during a year, take all or part all costs (investments, reinvestments and operation and
of the base load of a system. In contrast, a peak load unit maintenance costs). The levellised costs are calculated
is only used to reach specific peak periods of a few hours using the discount rate and the turbine lifetime.
when the demand is high. Lift Defined as the force experienced by a body in an air-
International Standards Organisation (ISO) ISO is a fed- flow perpendicular to the direction of the airflow.
eration of national standards bodies from around 150 dif- Log Law The log law is a parameterised mathematical
ferent countries. Established in 1947, it is non-govern- equation used to describe the increase in wind speed with
mental organisation with a mission broadly to promote the height above ground in the surface layer (see also Wind
development of standardisation with the objective of facil- Shear). The equation includes the surface roughness of
itating international trade. ISO standards are prefixed with the ground and the Coriolis parameter.
ISO (eg. ISO 9000). Logger See DataLlogger.
Inverter A power electronic device used to convert direct Low Speed Shaft This connects the rotor to the gearbox
current to alternating current. An inverter will often form in WTs featuring gearboxes. It is a high-torque shaft.
part of the power electronics used with variable speed
WTs, along with a rectifier. M
Investment Cost of Turbine The cost of the turbine itself,
317
including transport from the factory to the place where the Measure, Correlate and Predict (MCP) To produce a con-
turbine is erected. fident prediction of how much energy a proposed wind
farm will produce, many years of wind data from the site N
would be necessary. As there is usually only a year or
maybe two of data from the site available, the MCP Nacelle The nacelle of a WT is the enclosed volume
process is adopted instead. The data measured at the mounted on top of the tower, containing the gearbox,
site itself is correlated with data for the same period of generator and yaw drive, among other things.
time from a nearby meteorological station, and this corre- Nitrogen Oxide (NOx) According to the United Nations
lation is then applied to the remainder of the data from (2002), NOx is a “product of combustion from transporta-
the station to produce a synthetic long-term set of wind tion and stationary sources. It is a major contributor to
data for the wind farm site. This long-term data is then acid depositions and the formation of ground-level ozone
taken to be indicative of the likely future wind climate at in the troposphere. It is formed by combustion under high
the site and used to predict the future energy output of pressure and high temperature in an internal combustion
the wind farm. engine. It changes into nitrogen dioxide in the ambient air
Meteorological Mast The tall, thin, guyed pole upon and contributes to photochemical smog”.
which instruments such as anemometers and wind vanes
are mounted when conducting wind resource measure- O
ments.
Meteorological Station A long-term installation for meas- Offshore Wind speeds are typically higher and turbulence
urement of various properties of the atmosphere, wind lower offshore. Combined with reduced visual impact to
speed, direction, air temperature and pressure, precipita- sensitive landscapes this has meant that interest in build-
tion, insolation. etc. It is often possible to buy many ing offshore wind farms has increased in recent years.
decades worth of hourly data from these installations, and Operation and Maintenance Costs (O&M) The cost of
this can be used in combination with considerably shorter repairs to, and servicing of, WTs throughout their lives.
periods of more detailed 10-minute measurements made Overspeed Protection Devices fitted to the WT to prevent
at the proposed site of a new wind farm to produce good the rotor from accelerating to dangerous speeds should a
estimates of how much energy a new wind farm could pro- failure (generator, gearbox) occur in the WT. Such devices
duce in the long term. This is done by the measure, cor- include tip brakes and shaft brakes.
relate predict process.
Monopile A means of securing offshore WTs by boring a P
deep hole in the seabed and hammering a large pile into
the hole, to which the WT is attached. Park Effect The effect whereby WTs positioned together in
Motor Yaw The opposite of free yaw. In a motor yaw, WT large wind parks each produce less energy than they
electrical motors are used to precisely control the orien- would if in the same position on their own, due to the wind
tation of the nacelle and rotor and to prevent free yawing shadows of the other WTs in the park.
motion. The WT control system will decide in which direc- Pitch To control the power output and to assist in starting
tion to point the nacelle based on wind direction data and stopping, the blades of some WTs can be rotated
averaged over a few minutes. Also known as forced yaw. about their longitudinal axes by hydraulic activators. This
Multiplier/Multiplicator An employment multiplicator, for motion of the blade is known as pitching. The angle that
example, measures the direct and indirect employment the blade chord makes with the rotor disc is the pitch.
effect of producing €1 million worth of output from the WT Pitch Regulation Once the rated wind speed has been
manufacturing sector. Basically, this assumes that it is reached, a WT will be unable to make use of all of the
valid to multiply total WT manufacturing in euros with a wind energy incident on the rotor plane and must shed
factor giving the necessary employment to produce this some. A pitch-regulated WT will do this either by pitching
output. Series of multipliers for historic national account the blades to reduce the angle of attack and thus the
318
statistics exist. torque and power captured (in pitch-to-feather machines)
or by pitching the blades to increase the angle of attack
to cause sections of the blade to stall (in pitch-to-stall- Profile (Blade) The blade profile describes the shape you
machines). would see if you were to take a slice through the blade. It
Pole Switching A means by which a generator can be has a profound effect on the behaviour of the blade. There
made to operate at two different rotational frequencies. are a large number of standard profiles in use, described
Pollutant Defined as a “substance that is present in con- by names such as “NACA4412” which indicate the shape
centrations that may harm organisms (humans, plants family and proportions of the blade.
and animals) or exceed an environmental quality stan- Progress Ratio This ratio is related to the learning rate so
dard. The term is frequently used synonymously with con- that if the learning rate is 15%, then the progress ratio is
taminant” (United Nations, 2002). 85% (progress ratio is 100% minus learning rate).
Power Curve A plot of power output vs. wind speed, char-
acteristic of a particular WT model and configuration. R
Power Electronic Converter (PEC) A PEC is necessary on
fully variable-speed machines in order to convert the fre- Rated Capacity Refers to the nameplate capacity that
quency of the generated power from whatever it is being shows how much the turbine can produce when running at
generated at to the grid frequency. This is achieved by full load.
first rectifying then inverting the signal to the required fre- Rated Power The maximum power output possible from
quency. Older PECs generate problematic levels of har- the WT. This is dictated by the generator size and loads
monics, newer ones less so. that the WT can bear. Choice of rated power for a site is
Power Law An approximation used as an alternative to a balance dictated by the amount of energy available in
the log law for approximating the wind speed at a particu- the wind at different wind speeds and the cost of increas-
lar height in the surface layer. ingly large and powerful WTs.
Power Quality Utilities are committed to supplying elec- Rated Wind Speed The minimum wind speed at which a
tricity within narrow bands of frequencies and voltages WT will generate its rated power.
and with defined low levels of harmonics to consumers: Reactive Power If the voltage and current signals in an
this is described as the power quality. They are thus con- electrical network are not in phase, the out-of-phase com-
GLOSSARY
cerned that any generating device attached to their net- ponent gives rise to reactive power flow. Reactive power
works should inject power of sufficiently high quality. cannot be used, but still causes losses in an electrical
Variable speed and stall-regulated machines will smooth network, hence its flows should be minimised. Induction
over the effects of wind gusts, but the presence of power generators consume reactive power, and the network
electronic converters may increase the injection of har- operator may wish to be compensated for the losses this
monics onto the network. causes. Synchronous generators can be made to con-
Prevailing Wind Direction The direction from which the sume or produce reactive power as desired, which may be
wind comes for the largest proportion of the time. of use to the network operator.
Pressure Face The side of a WT blade upon which the air Rectifier A power electronic device used to convert alter-
flow around the blade causes a pressure rise. nating current into direct current. A rectifier will often form
Principal Component Analysis A particular way of fitting a part of the power electronics used with variable speed
curve through a set of data points which minimises the fit- WTs, along with an inverter.
ting error. Re-Energisation When a body of air has decelerated and
Productivity Productivity is used here as employees per is moving slowly (for example in a WT wake or towards the
output unit in fixed prices. The 2% increase in productivi- trailing edge of a blade), turbulent mixing with neighbour-
ty used as a basic assumption implies that 2% less peo- ing fast-moving airflows can re-energise the body of air,
ple are needed to produce the same output every year. If reaccelerating it.
additional cost reductions of turbines are assumed, this Reference Site A chosen location with known wind condi-
319
must partly be attributed to additional productivity increas- tions. The reference site is used to compare efficiency
es further reducing the need for employees. and power production of different turbines.
Reinvestments When a larger and more costly part of the Shear Air in contact with the ground is not moving: air at
turbine has to be replaced, eg. a gearbox or a blade. high altitudes is moving at high speeds. The gradual
Root The root of the blade is the section nearest the hub. increase of wind speed with increasing altitude is knows
Rotor The rotor is the combination of the hub and blades. as wind shear, and is particularly relevant over the area of
This term is also used to describe the rotating part of an a typical WT blade.
electrical generator. The design of a generator rotor Slip In induction generators, the slip is the difference in
depends on the type of generator in question. angular velocity between the rotor and the rotating mag-
Rotor Area See Swept Area. netic field, expressed as a fraction of the angular velocity
Rotor Disc The rotor disc is the imaginary circular surface of the rotating magnetic field.
swept out by the rotor blades as they rotate. Soft-Start If an induction generator is started by connect-
Roughness Class A classification system for the rough- ing it directly to the grid, it will initially draw a large current
ness of different surfaces. Examples include: landscape as the generator is magnetised; this large current could
with many trees and buildings (class 3-4); sea surface cause voltage drops in the local network. To avoid this, a
(class 0); concrete runway (class 0.5). Tables of rough- soft-start unit regulates the amount of current drawn by
ness classes for different surfaces are available. the generator, allowing it to magnetise more slowly and
Roughness Length The distance above ground level at preventing voltage dips on the network.
which the wind speed should theoretically be zero, due to Soft Tower A soft tower has a natural frequency lower
the roughness of the surface at that point. than the blade passing frequency but greater than the
rotational frequency of the WT.
S Soft-Soft Tower A soft-soft tower has a natural frequency
lower than the rotational frequency of the rotor.
Scale Parameter A parameter describing the height of a Solidity The frontal area of the blades on a WT divided by
Weibull distribution. the total swept area, expressed as a percentage.
Sea Breeze In the first half of the day as the land heats Squirrel Cage A description of the rotor of an induction
up, wind will tend to blow off the sea onto the land as generator.
heated air above the land rises (a weaker land breeze Stability The stability of the boundary layer above the
also occurs in the evening). Earth’s surface can be described as stable, neutral and
Separation An airflow is said to become separated from unstable. Stable conditions, when the surface of the Earth
the object around which it is moving when the flow ceas- is cooling, are characterised by little vertical mixing of the
es to follow the contours of the body but instead features air.. Unstable conditions occur when the Earth’s surface
turbulent mixing and flow reversal close to the surface. is heated and more vertical mixing occurs. Neutral condi-
Shadow Flicker If a WT comes between an observer and tions lie somewhere in between. We are largely concerned
the sun, the observer will experience a flickering as the with conditions of neutral stability in wind energy.
light passes between the rotating blades of the WT; this Stall Above a particular angle of attack, the air flow
can be very distracting. Wind farms should be planned to around the suction face of a blade will become separated.
avoid causing shadow flicker at nearby residences, and Once this happens, the drag on the blade will increase
wind farm design software can assist in doing this. dramatically and the lift will fall. The stalled section of
Shaft Brake A braking mechanism to stop the rotation of blade ceases to function as an efficient aerofoil.
the WT by arresting the motion of one of the shafts in the Stall Delay It is observed that sections of rotating WT
WT. As the high speed shaft has much lower torque, a blade close to the hub stall at higher angles of attack than
lighter and cheaper brake can be used if it is positioned would normally be expected. The reason for this is uncer-
on this shaft. However, should the gearbox fail, the high- tain.
speed shaft brake will be unable to stop the WT rotor. Stall Regulation The blades on a stall-regulated- WT can-
320
Shape Parameter A parameter describing the variation of not be pitched. To achieve control of power captured
a Weibull distribution about the mean. above rated wind speed, the blades are carefully designed
so that they progressively stall starting at the blade root typically consists of slots containing a series of coil wind-
and moving out towards the tip as the wind speed increas- ings, the energisation of which is electronically switched
es further. Stall-regulated WTs will respond much more to generate a moving field. The reluctivity of the rotor
quickly to gusts than pitch-regulated WTs (where mechan- results in a torque which tends to move the rotor in line
ic pitch adjustments take finite amounts of time to effect), with the energising coils, minimising the flux path. For an
improving the power quality output by the WT above rated SRG, mechanical energy is converted to electrical energy
wind speed. by the proper synchronisation of stator phase currents
Stall Strip A stall strip is a device attached near the max- with rotor position. By appropriate control of stator switch-
imum thickness of the chord of a stall-regulated blade ing it is possible to achieve variable speed operation for
which is intended to promote stall. the generator.
Stand-Alone This refers to a WT operating without being Synchronisation When using synchronous generators, it
attached to the electricity grid, for example in charging is vital that the alternating current produced is in phase
batteries or running in parallel with a diesel generator. with the alternating current on the network to which it is
Stator The non-rotating part of an electrical generator, connected; the process of achieving this (part of the WT
consisting of many windings of electrical cable. start-up programme) is known as synchronisation.
Stiff Tower A stiff tower has a first natural frequency high- Synoptic Describes variation in the typical frequency of
er than the blade passing frequency. arrival of different weather systems, generally with a peri-
Stream Tube An imaginary tube extending upstream and od of around four days.
downstream of the WT, containing all the air that interacts
with the rotor. The diameter of the stream tube is initially T
constant far upstream of the rotor, expands as air
approaches the rotor disc and slows, then continues to Teeter The blades on two-bladed wind WTs are usually, in
expand for a short distance after the rotor disc as the air fact, one single piece stretching the whole diameter of the
flow slows. rotor disc. To reduce stresses on the WT as a whole when
Suction Face The side of a WT blade upon which the air the blades pass the tower, this single piece is allowed to
GLOSSARY
flow around the blade causes a pressure drop. pivot at the centre; this is known as a teetered rotor.
Sulphur Dioxide (SO2) According to the United Nations Tilt The tilt angle is the angle between the horizontal and
(2002), SO2 is a “heavy, pungent, colourless gas formed the rotor axis.
primarily by the combustion of fossil fuels. It is harmful to Tip The tip of the blade is the end furthest from the hub.
human beings and vegetation, and contributes to the acid- Tip Brake These are fitted to some WTs to avoid runaway
ity in precipitation”. situations when other braking mechanisms have failed.
Surface Layer The first 100 m or so of the atmosphere, The tips of the blades are separate from the rest of the
in which most wind energy activity takes place. blade, and mounted on spring-loaded, threaded rods. As
Surface Roughness See Roughness Class. the rotor accelerates, centrifugal forces cause the tips to
Surface Winds The winds occurring near the surface of pull away from the body of the blade, and the thread on
the Earth (in the first kilometre or so), heavily influenced the rods causes the tip’s sections to rotate, spoiling the
by the properties of the Earth’s surface and obstacles in aerodynamics of the blade and regulating the speed of the
the region (see also Geostrophic Winds). rotor, preventing overspeed.
Swept Area The area swept out by the blades as the rotor Tip Loss Tip loss describes the loss in performance of a
rotates - the area of the rotor disc - is also known as the blade due to the fact that it is finite in extent and hence
rotor area. Often used as a proxy for a turbine’s power some air flow goes round the tip rather than over the
production. chord. It is borrowed from the propeller industry.
Switched Reluctance Generator (SRG) A SRG differs Tip Noise The noise made by the WT blades tips moving
321
from conventional machines in that it does not have any through the air. This increases with increasing speed, so
windings or permanent magnets on the rotor. The stator WTs near centres of population are often run more slowly
than say offshore WTs, especially at night when there is Turbulence Intensity A measure of the strength of the tur-
less background noise and people are sleeping. In gener- bulence of the wind compared to its underlying average
al, tip noise is only audible at low wind speeds when there speed. Defined as the standard deviation of the wind
is little other wind noise around. speed variations about the mean wind speed divided by
Tip Speed Ratio (λ) The ratio of the speed at which the the mean wind speed, using 10 minute or 1 hour averag-
blade tips are moving to the speed of the oncoming wind. ing periods.
A WT will typically reach peak efficiency at a particular Twist The tips of a WT blade move through the air faster
value of λ, regardless of wind speed (see variable speed) than the roots in the direction of rotation, but the oncom-
and this is often denoted λmax. ing wind speed is similar at all points along the blade.
Torque The “turning force” applied to an object. Measured Hence, to achieve the optimal angle of attack along its
in Newton metres, this is calculated as the distance from length, a WT blade is twisted so that the pitch of the blade
the axis of rotation at which a force is applied to a rotat- reduces from the root to the tip.
ing object, multiplied by the magnitude of that force.
Tower (Lattice, Tubular, Guyed Pole) The tower is the col- U
umn supporting the nacelle and rotor above the ground.
Towers are typically lattice (like an electricity pylon, consist- Upwind A WT which operates with the rotor upwind of the
ing of a network of thin struts) or tubular (a single tubular col- tower is said to be an upwind WT. Most modern WTs are
umn). For small WTs, the tower can also be a guyed pole. upwind.
Tower Shadow As wind flows around a WT tower, regions
of lower wind velocity are created just upstream and down- V
stream of the tower, particular in the case of tubular tow-
ers. As the blades pass through the tower shadow, they Value of Statistical Life (VSL) VSL is an approach that
experience a periodic dip in loading which causes fatigue measures a society’s willingness to pay to avoid addition-
damage to blades and can excite harmonics (resonance) al cases of death. This can be seen in spending for
in the blade. improved safety in the aircraft or car industry. In the EU
Tower Strike Occurs when extreme atmospheric condi- and the US, values of between one and 10 million US$ or
tions cause a blade to strike the tower as it passes € per life saved have been found in different studies.
Trailing Edge The sharper edge of a WT blade, to the rear Earlier versions of the ExternE project adopted a figure of
of the blade as it rotates. US$3 million per life saved for VSL calculations. In these
Transformer A piece of electrical equipment used to step calculations the age of a person saved does not matter.
up or down the voltage of an electrical signal. Most WTs Variable Speed The rotor on a variable speed WT will
will have a dedicated transformer to step up their voltage rotate at a speed calculated to make it as efficient as pos-
λmax – see Tip
output to grid voltage. sible at the prevailing wind speed (i.e. at

Troposphere The first 11km of the atmosphere, in which Speed Ratio). Variable speed WTs will produce less tip
weather occurs. noise at low wind speeds as they will be rotating more
Turbine Lifetime The expected total lifetime of the tur- slowly, and can be up to 10% more efficient than fixed
bine, normally 20 years. speed WTs. However, they are more complex than fixed
Turbulence The stochastic (random) motion of air charac- speed WTs and require power electronic converters
teristic of all natural winds. Though wind can be said to (PECs) to convert their output to grid frequency. A variety
have an underlying steady direction and speed, on a small of partially-variable speed WTs are available, bringing
scale different parcels of air will move in random direc- most of the efficiency gains with only some added com-
tions at random speeds. There is energy in turbulence plexity and no PEC.
that can be captured, but it also causes damage to WTs Velocity Deficit The amount by which the wind speed is
322
after long periods of exposure due to the constantly vary- reduced in the wake of a WT, as compared to the free
ing random loads a WT experiences. stream.
Vertical Axis Wind Turbine (VAWT) An alternative design mounted on the end of a rod, free to rotate in the hori-
of wind turbine where the rotor rotates about a vertical zontal plane about a pivot.
axis, rather than a horizontal one. It proved difficult to
scale these designs up effectively, so modern VAWTs are Y
of limited size; also, VAWTs are less efficient than
HAWTs. As the concept of yaw is irrelevant to VAWTs; they Yaw Angle As the direction of the wind is constantly
can however be rather simpler than HAWTs. changing, it is normally the case that the wind does not
Vortex Generator Vortex generators are small mechanical strike the rotor disc at right angles. The angle between the
devices which are attached to the suction surface of the rotor disc and the incident wind is known as the yaw angle
stall-regulated blade which generate local vortices. These (often denoted by Greek letter γ).
vortices re-energise the boundary layer and hence prevent Yaw Control Another means of WT power control (along-
stall. When viewed from above they are often V-shaped side pitch- and stall-regulation), primarily used in very
with the sharp end of the V pointing towards the leading small WTs. The WT is deliberately misaligned from the pre-
edge. vailing wind direction, reducing the area of the rotor disc
seen by the oncoming wind and thus the power output of
W the WT.
Yaw Drive This controls the direction in which the nacelle
Wake As wind passes through the rotor disc, it is slowed (and hence rotor) of the WT is pointing. Electrical motors
as energy is extracted from it, and vorticity is introduced are used to ensure that the WT is facing the prevailing
into the air by the blades. This stream of slowed air can wind direction at all times.
extend up to 10 rotor diameters behind the WT and is Yaw Error The amount by which the WT rotor is misaligned
known as the wake. WTs placed in the wake of others will from the prevailing wind direction.
experience lower wind speeds and increased turbulence, Years of Life Lost (YOLL) The YOLL approach takes into
reducing power output and increasing wear and tear. account that due to different causes people of very differ-
Weak Grid An area of the electrical grid where the voltage ent age groups may be at risk. In the case of a chronic dis-
GLOSSARY
and power quality is likely to be significantly influenced by ease leading to the death of very old people, only the
the presence of large loads or power supplies, such as a years of life lost due to the disease as compared to the
wind farm. It may be necessary to perform grid reinforce- average life expectancy are taken into account. For each
ment if a wind farm is to be connected to the grid where year of life lost approximately 1/20st of the value of sta-
it is currently weak. tistical life is used.
Weibull Distribution A probability distribution specific to a
given location describing the probabilities that the wind Z
will blow with particular strengths.

Wind Atlas An atlas mapping the wind resource across an Zone of Visual Influence (ZVI) In planning a new wind
area, for example The European Wind Atlas. farm, it is important to consider the effect it will have on
Wind Resource A reference to the quantity of energy the local landscape. Wind farm software will analyse top-
potentially available from the wind in a particular place (as ographical maps of the area surrounding a planned wind
in “The wind resource of the British Isles could supply all farm and highlight the areas from which the new wind
our electricity several times over”). farm will be visible - known as the ZVI.
Wind Rose A circular diagram giving a visual summary of
the relative amounts of wind available in each of a num-
ber of direction sectors (often 12) at a given location, and
the speed content of that wind.
323
Wind Vane An instrument for measuring the direction in
which the wind is blowing, usually consisting of a vane
Technical Units
kW kilowatt 1,000 Watts

MW megawatt 1,000,000 Watts (1,000 kW)
GW gigawatt 1,000,000,000 Watts (1,000 MW)
kWh kilowatt hour 1,000 Watt hours

MWh megawatt hour 1,000 kilowatt hours
GWh gigawatt hour 1,000 megawatt hours
324
REFERENCES
Anon (2002) La mayoría de los albaceteños considera pos- Centre for Energy Policy and Technology (2001) External
itiva la llegada de la energía eólica. Diario La Verdad. Cost and Environmental Policy in the United Kingdom and
Octubre 2002. the European Union, Centre for Energy Policy and
Technology, Imperial College, London. Available from:
Awerbuch, S. (2003) The true cost of fossil-fired electricity, www.british-energy.com/environment/downloadable
Power economics, May 2003. _items/other/external_cost_and_environmental_policy.
doc [accessed 14th May 2003].
BirdLife (2002) Windfarms and Birds: An analysis of the
effects of windfarms on birds, and guidance on environ- CERES (2002) Consulta a la Població de la Província de
mental assessment criteria and site selection issues. Tarragona sobre diversos aspectes relacionats amb la
[online]. Strasbourg. Available from: http://www.coe.int/ instal-lació de Parcs Eòlic.
t/e/Cultural%5FCo%2Doperation/Environment/Nature%5
Fand%5Fbiological%5Fdiversity/Nature%5Fprotection/sc2 Commission for Energy Regulation/Office for Electricity
2_inf30erev.pdf [Accessed 19th September 2003]. Regulation NI (2003) The Impacts of Increased Levels of
Wind Energy Penetration on the Electricity Systems of the
Black and Veatch (2003) Strategic Assessment of the Republic of Ireland and Northern Ireland: Final Report
Potential for Renewable Energy: Stage 1, European Bank (www.cer.ie).
for Reconstruction and Development.
Delft University et al., (2001) Concerted Action on
British Wind Energy Association (2003) New Power for Offshore Wind Energy in Europe. Offshore Wind Energy:
Britain. Government’s expansion of wind power backed by Ready to Power a Sustainable Europe, December.
public [online]. BWEA Press Release. Available from:
www.britishwindenergy.co.uk/view/news/round2.html CRES (2001).
[Accessed 8th July 2003].
.
Brocklehurst, F. (1997) A Review of the UK Onshore Wind
Energy Resource, ETSU-R-99. Damborg, S. (1998) Public Attitudes Towards Wind Power.
REFERENCES
Danish Wind Industry Association. [online], publications.
Brower, M.C. et al. (2000) Validation of the Forewind Available from: www.windpower.dk/articles/surveys.htm
Model in Wind Forecasting Application, EWEA Special [Accessed 2nd June 2003].
Topic Conference in Kassel.
Danish Wind Industry Association (1995) Employment in
BTM Consult (2003) International Wind Energy Development Wind in Denmark, Wind Power Note No. 2 (in Danish).
– World Market Update 2002, March.
Danish Wind Industry Association (1997) The Energy
Burch, S.F. and F. Ravenscroft (1992) Computer Modelling Balance of Modern Danish Wind Turbines, Wind Power
of the UK Wind Energy Resource: Overview Report, ETSU Note No. 16.
WN7055. More on NOABL database at: www.bwea.com/
noabl/index.html. Danish Wind Industry Association (2002) Bag NESA’s
groenne elskranke, Vinformation, 30.
Danish Wind Industry Association (2003) Danish wind power

2002 [news online]. Available from: www.windpower.
325
org/en/news/stat2002.htm [accessed 19th November 2003].
Danish Wind Turbine Owners’ Association (2003) Lav vilka- EMU and Hammarlund Consulting (2003) European
ar om eller aflys licitation, Naturlig Energi, December 2003. Thematic Network on Wave Energy: Public Acceptability.
Final Report. Available from: www.emu-consult.dk/
Danmarks Vindmølleforening (2002) Danskernes mening includes/public-acceptability.pdf [Accessed 8th August
om Vindmøller. Fakta om Vindenergi, M6 [online]. 2003].
Available from: www.dkvind.dk/fakta/ fakta_pdf/M6.pdf
[Accessed 15th July 2003]. Ender, C. (2003) Windenergienutzung in der
Bundesrepublik Deutschland – Stand 31.12.2002, Dewi
Dany, G et al. (2003), Energiewirtschaftliche Tagesfragen, Magazin Nr. 22 February 2003, pp.7-19.
Heft 9, September.
Energy Information Administration (1995) Electricity
Dewi (2002) Studie zur aktuellen Kostensituation 2002 der Generation and Environmental Externalities: Case Studies
Windenergienutzung in Deutschland – draft, July 2002. (Washington, DC: EIA). Available from: www.eia.doe.gov/
cneaf/electricity/external/external.pdf [accessed 14th
Dobesch, H. and G. Kury (1997) Wind Atlas for the Central May 2003].
European Countries. Österreichische Beiträge zu
Meteorologie und Geophysik, Heft 16 (Wien, Austria: Ernst, B. and K. Rohrig (2000) Online-supervision and
Zentralanstalt für Meteorologie und Geodynamik). Prediction of 2,500 MW Wind Power.
ECN (2003) The Dutch Renewable Electricity Market in 2003. Eurelectric (2002) Statistics and Prospects for the
European Electricity Sector (1980-1990, 2000-2020), EUR-
ECOTEC (2002) Renewable Energy Sector in the EU: Its PROG 2000.
Employment and Export Potential, Report to DG Environment,
ECOTEC, Birmingham UK 2002. EurObserv’ER (2003) Wind Energy Barometer. European
Commission [online], Energy. Available from:
Eggersglüß, W (2002) Das steht und das dreht sich. In: http://www.observ-er.org/comm/baro153.pdf [Accessed
Ministerium für Finanzen und Energie des Landes 5th August 2003].
Schleswig-Holstein. Stimmen zur Windenergie. p. 8-10.
Available from: http://landesregierung.schleswig- European Commission (1994) Externalities of Fuel Cycles
holstein.de/coremedia/generator/Aktueller_20Bestand/ “ExternE” Project. Summary Report (Brussels: European
FM/Brosch_C3_BCre_20_2F_20Publikation/PDF/ Commission).
Windstimmen,property=pdf.pdf [Accessed 15 July 2003].
th
European Commission (1997) Renewable Sources of

EHN (1998) Un 81% de los Navarros estima positiva la Energy, White Paper.
instalación de parques eólicos y un 3% los considera per-
judiciales. Pamplona. European Commission (1999) ExternE. Externalities of ener-
gy. National Implementation Vol. 10 (Brussels: European
EHN (2001) Medio ambiente y entorno social. In: EHN Commission).
Informe Anual 2001.
European Commission (2002a) Towards a European
EHN (2003) Report on the impact of wind farms in Navarra Strategy for the Security of Energy Supply, Green Paper.
on birdlife.
326
Elliott, D. et al. (2003) Wind Energy Resource Atlas of
Armenia.
European Commission (2002b): Eurobarometer. Energy: Focken, U. et al. (2000) Reduction of Windpower
Issues, Options, and Technologies Science and Society. Prediction Error by Spatial Smoothing Effects, EWEA
Luxembourg, European Commission. Available from: Special Topic Conference in Kassel.
http://europa.eu.int/comm/research/energy/pdf/euro-
barometer_energy_en.pdf [Accessed 6th August 2003]. Friedrich, L. et al. (1989) Externen Kosten der
Stromerzeugung.
European Commission (2003) European Energy and
Transport Trends 2030. Garrad Hassan and Partners (2001) Wind Farmer Theory
Manual.
European Opinion Research Group (2003).
Garrad Hassan, Germanischer Lloyd, Windtest (1995)
Eurostat (2000) The Input-Output Tables: Eurostat Data Study of Offshore Wind Energy in the EC.
Shop Handbook.
Gesellschaft für Reaktorsicherheit (1989) Deutsche
Eurostat (2003) Electricity Statistics. Risikostudie Kernkraftwerke Phase B. Eine zusammen-
fassende Darstellung 2, Auflage, Köln.
EWEA (2000) Special Topic Conference on Wind Power
Forecasting System, Kassel. Giebel, G. et al. (2000) The Zephyr-Project The Next
Generation Prediction System, EWEA Special Topic
EWEA (2003a) Wind Power Targets for Europe: 75,000 MW Conference in Kassel.
by 2010.
Gipe (1995) Wind energy comes of age. John Wiley &
EWEA (2003b) Survey for Wind Energy - The Facts, sum- Sons, Inc.
mer 2003.
Grubb, M.J. and N.I. Meyer (1993) Wind Energy:
EWEA (2003c) Wind Force 12: A Blueprint to Achieve 12% resources, systems, and regional strategies. In: Johansson,
of the World’s Electricity From Wind Power by 2020 T.B. et al. (eds) Renewable Energy, Sources for Fuels and
REFERENCES
(Brussels: European Wind Energy Association/Greenpeace), Electricity.
May.
Günther, W (2002) Touristische Effekte von On- und
EWEA (2003d) The European Wind Energy Association’s Offshore- Windkraftanlagen in Schleswig-Holstein. In:
comments on COM (2003) 131 - Consultation on the Ministerium für Finanzen und Energie des Landes
Commission’s Communication on developing an action Schleswig-Holstein. Stimmen zur Windenergie. p. 61-63.

plan for environmental technology of 25 March 2003. Available from: http://landesregierung.schleswig-
holstein.de/coremedia/generator/Aktueller_20Bestand/
EWEA (2003e) Eastern promise, Wind Directions, March FM/Brosch_C3_BCre_20_2F_20Publikation/PDF/
2003. Windstimmen,property=pdf.pdf [Accessed 15th July 2003].
EWEA (2003f) Wind Directions, Focus on Public Opinion, Hohmeyer, O. (1996) Social costs of climate change strong
September /October, p.16-33. sustainability and social costs. In: Hohmeyer, O. et al. (eds)
Social Costs and Sustainability (Berlin and Heidelberg:
Fischedick, M. et al. (2000) Nach dem Ausstieg: Springer-Verlag).
Zukunftskurs Erneuebare Energien (Stuttgart: Hirzel).
327
Hohmeyer, O. (1988) Soziale Kosten des Energieverbrauchs Koistinen, J. (2002) Survey concerning influence on bird
(Berlin and Heidelberg: Springer-Verlag). life. For Ministry of Environment. Finland.
Hohmeyer, O. et al. (eds) (2000) Chance Automausstieg. Koomey, J. and F. Krause (1997) Introduction to
Perspektiven für neue Arbeitsplätze an Atomstandorten Environmental Externality Costs. Available from:
(Hamburg: Greenpeace). enduse.lbl.gov/Info/Externalities.pdf [accessed 14 May th
2003].
IDAE (2003) Current Status and Prospects for Wind Energy
in Spain 2002 (Madrid: IDAE). Krieg, R. (1992) Vindatlas for Sverige.
International Energy Agency (1999) Annex XI: Krieg, R. (1999) Verifiering af Beraknad Vindenergiproduktion.
Recommended practices for wind turbine testing 11. Wind
speed measurement and use of cup anemometry, 1 (Paris: Krohn, S. (1998) Creating a Local Wind Industry:
IEA). Experience from Four European Countries, Helios Centre
for Sustainable Energy Strategies.
IEA International Energy Agency (2003) World Energy
Investment Outlook 2003 (Paris: IEA). Krohn, S. (2000) Renewables in the EU single market – An
economic and a policy analysis.
IEC 6-1400 Part 12, Wind turbine generator systems Part
12: Wind Turbine Power Performance Testing. Kull, A Estonia Wind Atlas, University of Tarfu, Institute of
Geography.
IPCC (2001) Climate Change 2001: Mitigation. Summary
for policy makers and technical summary of the IPCC WGIII Kull, A. and G. Steinrucke (1996) Estonia Wind Atlas,
Third Assessment Report (Geneva: IPCC Secretariat). University of Tarfu, Institute of Geography.
IPCC (2002) Workshop on Carbon Dioxide Capture and Landberg, L. (2000) Prediktor – an On-line Prediction
Storage. System, EWEA Special Topic Conference in Kassel.
Ivanov, P. et al. (1997) Renewable Energy Potential and Landberg, L. et al. (2003) Wind Resource Estimation – An
Sustainable Development of Bulgaria, The European Overview, Wind Energy Journal Vol. 6, No 3.
Congress on Renewable Energy Implementation, Asir
palace, Vouliagmeni, Athens, Greece, 5-7 May 1997, Lange, B. et al. (2003) The Wind Speed Profile at Offshore
pp.446-55. Wind Farm Sites, Conference Proceedings, Offshore Wind

Energy in the Mediterranean and Other European Seas,
Jensen, P.H. et al. (2002) Økonomi for vindmøller i Naples, April 2003.
Danmark - Etablerings-, drifts- og vedligeholdelsesomkost-
ninger for udvalgte generationer. Manwell, J.F. et al (2002) Wind Energy Explained. Theory,
Design and Application. Wiley.
Kemmoku, Y. et al. (2000) Method of Forecasting the Wind
Velocity Using Weather Data over Wide Area. Marie, A. et al. (1996) A summary of research conducted
into attitudes to wind power from 1990 - 1996. The British
Kjaer, C. (1999) That was then - this is tomorrow, Wind Wind Energy Association [online]. Available from:
Power Monthly, December 1999. www.bwea.com/ref/survsum.html [Accessed 8th August
328
2003].
Martf, I. et al. (2000) First Results of the Application of a OECD (1998) Improving the Environment Through
Wind Energy Prediction Model in Complex Terrain, EWEA Reducing Subsidies.
Special Topic Conference in Kassel.
Official Journal of the European Communities (2001)
Measnet: International Network for Harmonised and Community Guidelines on State Aid for Environmental
Recognised Measurements in Wind Energy (2003) Protection (2001/C 37/03). Available from:
www.measnet.com. europa.eu.int/eur-lex/pri/en/oj/dat/2001/c_037/
c_03720010203en00030015.pdf [accessed 19th
Milborrow, D.J. (2003) Wind Power Monthly, April 2003, p. November 2003].
51.
Ottinger, R.L . et al. (1990) Environmental Cost of
Moehrlen, C.S. et al. (2000) Wind Power Prediction and Electricity (Pace University Center of Environmental Legal
Power Plant Scheduling in Ireland, EWEA Special Topic Studies, New York: Oceana Publications, Inc.).
Conference in Kassel.
Petersen et al. (1981) Wind Atlas for Denmark, A Rational
MORI Scotland (2002) Tourist attitudes towards wind Method for Wind Energy Siting.
farms. Summary report. Available from: www.bwea.com/
pdf/MORI.pdf [Accessed 20th July 2003]. Podesta et al. (2002) The Wind Map of Italy. At Eurosun
2002, June 23-26, 2002, Bologna, Italy.
National Wind Coordinating Committee NWCC (2002)
Permitting of Wind Energy Facilities. A Handbook. Available Podesta et al. (2002) The Italian Wind Atlas. At In-Vento-
from: www.nationalwind.org/pubs/permit/permitting 2002, 15-18 September 2002, Milan, Italy.
2002.pdf [Accessed 11th July 2003].
Raftery, P. et al. (1999) Understanding the Risks of
Neij, L. (1997) Use of experience curves to analyse the Financing Wind Farms, proceedings of the EWEC
prospects for diffusion and adaption of renewable energy Conference, Nice, France.
technologies, Energy Policy, 23(13).
REFERENCES
Rathman, O. (2003) A New Coastal Wind Atlas for the
Neij, L.; Dannemand Andersen, P.; Durstewitz, M.; Helby, Baltic Countries, EWEC.
P.; Hoppe-Kilpper, M.; Morthorst, P.E., (2003) Experience
curves: A tool for energy policy assessment. (Lund Redlinger, R.Y., Dannemand Andersen, P. and P.E.
University, Department of Technology and Society, Morthorst (2002) Wind Energy in the 21st Century (Palgrave).
Environmental and Energy Systems Studies, Lund (SE),

2003) (IMES/EESS Report, 40) 139 p. Renewable Energy Research Laboratory et al.(2002) Wind
turbine noise issues. Available from: www.ceere.org/
NFO System Three (2002) Investigation into the potential rerl/publications/whitepapers/WindTurbineNoiseIssues.p
impact of wind farms on tourism in Scotland. Final Report. df [Accessed 9th July 2003].
Nielsen, T.S. and H. Madsen (2000) WPPT – A Tool for Rennings, K. (1996) Economic and ecological concepts of
Wind Power Prediction, EWEA Special Topic Conference in sustainable development: External costs and sustainability
Kassel. indicators. In: Hohmeyer, O. et al. (eds) Social Costs and
Sustainability (Berlin and Heidelberg: Springer-Verlag).
OECD/IEA (1998) Projected Costs of Generating Electricity
329
– Update 1998, OECD Nuclear Energy Agency and
International Energy Agency, Paris.
Risø, EMD (1999) Wind Map of Denmark, Danish Energy Troen, I. and E.L. Petersen (1989) European Wind Atlas,
Agency. Risø National Laboratory, Denmark.
Risø National Laboratory (2002) OffshoreWakes: TrueWind Solutions (2003) Wind Energy Resource Maps of
Measurements and Modelling, proceedings of the ENDOW the Republic of Ireland, ESBI International.
workshop.
TWE-Report 970615-2 (1997) Power Curve Measurement
Sagrillo, M. (2003) Putting Wind Power’s effect on birds in for Vestas V47 660 kW Wind Turbine at O Gammelby Tripod
perspective. American Wind Energy Association Wind Wind Energy ApS.
Energy [online], Technical Info. Available from:
www.awea.org/faq/sagrillo/swbirds.html [Accessed 9th UK Department of Trade & Industry (2002), Quantifying
July 2003]. the System Costs of Additional Renewables in 2020, Ilex
Energy Consulting/UMIST, October.
Sander, J. et al. (2003) A New Wind Atlas of Poland,
EWEC, Madrid. Umweltbundesamt (2002) Vergleich externer Kosten der
Stromerzeugung in Bezug auf das Erneuerbare Energien
Schwahn, C. (2002) Landscape and policy in the north sea Gesetz.
Marsches. In: Pasqualetti, Martin J. et al. (2002) Wind
power in view. Energy landscapes in a crowded world. United Nations (2002) Environmental glossary[online].
Academic Press. Environment Statistics. Available from: unstats.un.org/
unsd/environment/default.htm [accessed 24th November
Sifo Research and Consulting (2002) Telephone inter- 2003].
views 29th January – 4th February. Project 3825150 (SoCe).
Van Wijk, A.J.M. and J.P. Coelingh (1993) Wind Potential
Starkov et al. (2000) Russian Wind Atlas. in the OECD Countries, University of Utrecht.
Steinrücke et al. (1996) Zukünftige Nutzung der VDEW (2000) VDEW-Statistik 1998. Leistung und Arbeit.
Windenergie in der Republik Estland, Die Erde, 127/1996. (einschließlich CD-ROM), Frankfurt, VWEW Verlag- und
Wirtschaftsgesellschaft der Elektrizitätswerke, p.21.
System Three Social Research (2000) Public attitudes
towards wind farms in Scotland. Results of a residents sur- Vector, A.S. (2001 ). The Norwegian Wind Atlas, revision 1.
vey. Scottish Executive, Central Research Unit.
Vector, A.S. (2003) The Norwegian Wind Atlas, revision 2.

Tammelin, B. (1991) Wind Atlas for Finland, Finnish
Meteorological Institute. VEÖ (2003) Verband der Elektrizitätsunternehmen
Österreichs
Tande, J.O.G. (2003) Grid integration of wind farms, Wind
Energy Journal Vol. 6, No.3. Waldl, H-P. and G. Giebel (2000) The Quality of a 48-Hours
Wind Power Forecast Using the German and Danish
Transmission System Operator, Ireland (2002) Weather Prediction Model, EWEA Special Topic
Generation Adequacy Report 2003-2009. Conference in Kassel.
Traup, S., and Kruse, B. (1996) Wind und Windenergie- Watson, R. and L. Landberg, The Irish Wind Atlas (forthcoming).
330
potentiale in Deutschland – Winddaten fur Windenergie-
nutzer.
Weinreich, S. (2002) Nachhaltige Entwicklung im
Personenverkehr unter Berücksichtigung der externen
Kosten, PhD thesis at the University of Flensburg.
Western EcoSystems Technology Inc. (2001) Avian

Collisions with Wind Turbines: A Summary of Existing
Studies and Comparisons of Avian Collision Mortality in the
United States. Available from: http://www.nationalwind.
org/pubs/avian_collisions.pdf [Accessed 9th July 2003].
Wiser, Bolinger and Holt (2000) Customer Choice and

Green Power Marketing: A Critical Review and Analysis of
Experience to Date (University of California: Lawrence
Berkeley National Laboratory).
World Commission for Environment and Development

(WCED) (1987) Brundtland Report.
Youth, Howard (2003) Winged messengers. The decline of

birds. Worldwatch paper 165.
REFERENCES
331

WETF-wind Power PDF

Uploaded by

Copyright:

Available Formats

WETF-wind Power PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

WETF-wind Power PDF

Uploaded by

Copyright:

Available Formats

VOLUME 1 - TECHNOLOGY

CHAPTER 1 TURBINE TECHNOLOGY 7

1.1 Evolution of Commercial Wind Technology 7

1.2 Technology Trends 19

1.3 Recent Developments 23

1.4 Technology Status 32

WIND ENERGY - THE FACTS - TABLE OF CONTENTS

1.5 Concluding Remarks and Future R&D 37

CHAPTER 2 WIND RESOURCE ESTIMATION 38

2.2 Regional Wind Resources 38

2.3 Wind Atlases 39

2.4 Energy Estimates 41

2.6 Offshore Wind Farm Design and Resource Estimation 58

2.8 Future Developments 63

CHAPTER 3 WIND FARM DESIGN 65

3.2 Preliminary Layout Design 65

3.3 Detailed Layout Design 65

3.4 The Infrastructure 66

3.7 Commissioning and Operation 68

3.8 Concluding Remarks 69

CHAPTER 4 TRANSMISSION AND DISTRIBUTION NETWORKS 70

4.2 Setting the Scene 70

4.4 Considerations for Wind Energy 74

4.5 Issues for High Wind Penetration in Europe 80

4.6 Concluding Remarks 82

CHAPTER 5 RESEARCH AND DEVELOPMENT 84

5.1 Wind Industry Research and Development Overview 84

5.2 Socio-Economic, Policy and Market Issues 85

5.3 Environmental and Social Impacts 86

5.4 Turbine and Component Design Issues 87

5.5 Testing, Standardisation, Certification and Safety 88

5.6 Grid Integration, Energy Systems, and Resource Prediction 88

5.7 Operation and Maintenance 89

5.8 Location of Wind Farms 90

5.9 Offshore Wind 90

5.10 Multi-Megawatt Turbines 91

WIND ENERGY - THE FACTS - TABLE OF CONTENTS

CHAPTER 2 COST AND INVESTMENT STRUCTURES 97

CHAPTER 3 OPERATION AND MAINTENANCE COSTS OF WIND POWER 100

CHAPTER 4 THE COST OF ENERGY GENERATED BY WIND POWER 103

CHAPTER 5 DEVELOPMENT OF THE COST OF WIND POWER 105

CHAPTER 6 FUTURE DEVELOPMENT OF THE COST OF WIND POWER 106

CHAPTER 7 COSTS OF CONVENTIONAL POWER PRODUCTION 108

CHAPTER 8 EXTERNAL COSTS OF POWER PRODUCTION 110

CHAPTER 1 INDUSTRY STATUS 114

1.1 Introduction 114

1.5 Domestic Market 123

1.6 Export Market 124

1.7 Industry Segments 125

CHAPTER 2 EMPLOYMENT IN THE WIND TURBINE SECTOR 127

2.1 Direct and Indirect Employment Effect 127

WIND ENERGY - THE FACTS - TABLE OF CONTENTS

CHAPTER 3 EMPLOYMENT PREDICTION AND METHODOLOGY 133

3.1 Direct and Indirect Employment 133

CHAPTER 4 SCENARIOS FOR EMPLOYMENT IN THE WIND TURBINE SECTOR 137

CHAPTER 5 DEVELOPMENT AND INNOVATION 139

INTRODUCTION TO VOLUME 4 143

CHAPTER 1 EXTERNALITIES 147

CHAPTER 2 ENVIRONMENTAL BENEFITS OF WIND ENERGY 158