Hide Notes 

Slide 1: Renewable Energy in Power Systems Leon Freris Centre for Renewable Energy Systems Technology (CREST), Loughborough University, UK David Infield Institute of Energy and Environment, University of Strathclyde, UK A John Wiley & Sons, Ltd, Publication
Slide 3: Renewable Energy in Power Systems
Slide 5: Renewable Energy in Power Systems Leon Freris Centre for Renewable Energy Systems Technology (CREST), Loughborough University, UK David Infield Institute of Energy and Environment, University of Strathclyde, UK A John Wiley & Sons, Ltd, Publication
Slide 6: This edition first published 2008 © 2008, John Wiley & Sons, Ltd Registered office John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, United Kingdom For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com. The right of the author to be identified as the author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Infield, D. G. Renewable energy in power systems / Leon Freris, David Infield. p. cm. Includes bibliographical references and index. ISBN 978-0-470-01749-4 (cloth) 1. Renewable energy sources. I. Freris, L. L. II. Title. TJ808.I54 2008 621.4–dc22 2007050173 A catalogue record for this book is available from the British Library. ISBN 978-0-470-01749-4 Set in 10 on 12 Times by SNP Best-set Typesetter Ltd., Hong Kong Printed in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire Cover image © Ted Leeming Reproduced by permission of Ted Leeming
Slide 7: Contents Foreword Preface Acknowledgements xi xiii xv 1 1.1 1.2 1.3 1.4 Energy and Electricity The World Energy Scene 1.1.1 History 1.1.2 World energy consumption 1.1.3 Finite resources 1.1.4 Energy security and disparity of use The Environmental Impact of Energy Use 1.2.1 The problem 1.2.2 The science 1.2.3 The Kyoto protocol 1.2.4 The Stern Report 1.2.5 Efficient energy use 1.2.6 The electricity sector 1.2.7 Possible solutions and sustainability Generating Electricity 1.3.1 Conversion from other energy forms – the importance of efficiency 1.3.2 The nuclear path 1.3.3 Carbon capture and storage 1.3.4 Renewables The Electrical Power System 1.4.1 Structure of the electrical power system 1.4.2 Integrating renewables into power systems 1.4.3 Distributed generation 1.4.4 RE penetration References 1 1 1 1 2 3 3 3 5 6 7 8 10 11 11 11 12 13 13 16 16 18 19 19 20 2 2.1 2.2 2.3 Features of Conventional and Renewable Generation Introduction Conventional Sources: Coal, Gas and Nuclear Hydroelectric Power 2.3.1 Large hydro 2.3.2 Small hydro 21 21 22 23 24 25
Slide 8: vi Contents 2.4 Wind Power 2.4.1 The resource 2.4.2 Wind variability 2.4.3 Wind turbines 2.4.4 Power variability 2.5 PV and Solar Thermal Electricity 2.5.1 The resource 2.5.2 The technology 2.5.3 Photovoltaic systems 2.5.4 Solar thermal electric systems 2.6 Tidal Power 2.6.1 The resource 2.6.2 Tidal enhancement 2.6.3 Tidal barrages 2.6.4 Operational strategies 2.6.5 Tidal current schemes 2.7 Wave Power 2.7.1 The resource 2.7.2 The technology 2.7.3 Variability 2.8 Biomass 2.8.1 The resource 2.8.2 Resource sustainability 2.9 Summary of Power Generation Characteristics 2.10 Combining Sources References 3 3.1 3.2 Power Balance/ Frequency Control Introduction 3.1.1 The power balance issue Electricity Demand 3.2.1 Demand curves 3.2.2 Aggregation 3.2.3 Demand-side management – deferrable loads Power Governing 3.3.1 Power conversion chain 3.3.2 The governor 3.3.3 Parallel operation of two generators 3.3.4 Multigenerator system 3.3.5 The steady state power–frequency relationship Dynamic Frequency Control of Large Systems 3.4.1 Demand matching 3.4.2 Demand forecasting 3.4.3 Frequency limits 3.4.4 Generation scheduling and reserve 3.4.5 Frequency control at different timescales 3.4.6 Meeting demand and ensuring reliability 3.4.7 Capacity factor and capacity credit Impact of Renewable Generation on Frequency Control and Reliability 3.5.1 Introduction 3.5.2 Aggregation of sources 27 27 28 30 33 36 36 37 38 40 42 42 43 43 44 45 47 47 48 49 50 50 51 52 53 53 55 55 55 56 56 57 58 59 59 60 61 62 63 64 64 65 67 68 68 70 71 72 72 73 3.3 3.4 3.5
Slide 9: Contents vii 3.6 3.7 3.8 3.5.3 Value of energy from the wind 3.5.4 Impact on balancing 3.5.5 Impact on reliability 3.5.6 Discarded/curtailed energy 3.5.7 Overall penalties due to increasing penetration 3.5.8 Combining different renewable sources 3.5.9 Differences between electricity systems 3.5.10 Limits of penetration from nondispatchable sources Frequency Response Services from Renewables 3.6.1 Wind power 3.6.2 Biofuels 3.6.3 Water power 3.6.4 Photovoltaics Frequency Control Modelling 3.7.1 Background 3.7.2 A modelling example Energy Storage 3.8.1 Introduction 3.8.2 Storage devices 3.8.3 Dynamic demand control References Other Useful Reading Electrical Power Generation and Conditioning The Conversion of Renewable Energy into Electrical Form The Synchronous Generator 4.2.1 Construction and mode of operation 4.2.2 The rotating magnetic field 4.2.3 Synchronous generator operation when grid-connected 4.2.4 The synchronous generator equivalent circuit 4.2.5 Power transfer equations 4.2.6 Three-phase equations 4.2.7 Four-quadrant operation 4.2.8 Power–load angle characteristic: stability The Transformer 4.3.1 Transformer basics 4.3.2 The transformer equivalent circuit 4.3.3 Further details on transformers The Asynchronous Generator 4.4.1 Construction and properties 4.4.2 The induction machine equivalent circuit 4.4.3 The induction machine efficiency 4.4.4 The induction machine speed–torque characteristic 4.4.5 Induction generator reactive power 4.4.6 Comparison between synchronous and asynchronous generators Power Electronics 4.5.1 Introduction 4.5.2 Power semiconductor devices 4.5.3 Diode bridge rectifier 4.5.4 Harmonics 4.5.5 The thyristor bridge converter 76 76 79 79 80 81 81 81 84 84 85 86 86 86 86 89 91 91 91 93 94 95 97 97 98 98 101 103 104 105 106 107 108 108 108 110 112 112 112 114 116 117 120 121 121 121 122 124 126 126 4 4.1 4.2 4.3 4.4 4.5
Slide 10: viii Contents 4.6 4.5.6 The transistor bridge 4.5.7 Converter internal control systems 4.5.8 DC–DC converters Applications to Renewable Energy Generators 4.6.1 Applications to PV systems 4.6.2 Applications to wind power References Power System Analysis Introduction The Transmission System 5.2.1 Single-phase representation 5.2.2 Transmission and distribution systems 5.2.3 Example networks Voltage Control Power Flow in an Individual Section of Line 5.4.1 Electrical characteristics of lines and cables 5.4.2 Single-phase equivalent circuit 5.4.3 Voltage drop calculation 5.4.4 Simplifications and conclusions Reactive Power Management 5.5.1 Reactive power compensation equipment Load Flow and Power System Simulation 5.6.1 Uses of load flow 5.6.2 A particular case 5.6.3 Network data 5.6.4 Load/generation data 5.6.5 The load flow calculations 5.6.6 Results 5.6.7 Unbalanced load flow Faults and Protection 5.7.1 Short-circuit fault currents 5.7.2 Symmetrical three-phase fault current 5.7.3 Fault currents in general 5.7.4 Fault level (short-circuit level) – weak grids 5.7.5 Thévenin equivalent circuit Time Varying and Dynamic Simulations Reliability Analysis References Renewable Energy Generation in Power Systems Distributed Generation 6.1.1 Introduction 6.1.2 Point of common coupling (PCC) 6.1.3 Connection voltage Voltage Effects 6.2.1 Steady state voltage rise 6.2.2 Automatic voltage control – tap changers 6.2.3 Active and reactive power from renewable energy generators 6.2.4 Example load flow 128 133 133 134 134 137 147 149 149 149 151 152 153 153 156 156 156 157 158 160 160 163 163 164 165 165 167 168 168 169 169 170 170 171 171 172 173 173 175 175 175 176 176 177 177 178 179 180 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 6 6.1 6.2
Slide 11: Contents ix 6.3 6.4 6.5 6.6 6.7 Thermal Limits 6.3.1 Overhead lines and cables 6.3.2 Transformers Other Embedded Generation Issues 6.4.1 Flicker, voltage steps and dips 6.4.2 Harmonics/distortion 6.4.3 Phase voltage imbalance 6.4.4 Power quality 6.4.5 Network reinforcement 6.4.6 Network losses 6.4.7 Fault level increase Islanding 6.5.1 Introduction 6.5.2 Loss-of-mains protection for rotating machines 6.5.3 Loss-of-mains protection for inverters Fault Ride-through Generator and Converter Characteristics References Power System Economics and the Electricity Market Introduction The Costs of Electricity Generation 7.2.1 Capital and running costs of renewable and conventional generation plant 7.2.2 Total generation costs Economic Optimization in Power Systems 7.3.1 Variety of generators in a power system 7.3.2 Optimum economic dispatch 7.3.3 Equal incremental cost dispatch 7.3.4 OED with several units and generation limits 7.3.5 Costs on a level playing field External Costs 7.4.1 Introduction 7.4.2 Types of external cost 7.4.3 The Kyoto Agreements 7.4.4 Costing pollution 7.4.5 Pricing pollution Effects of Embedded Generation 7.5.1 Value of energy at various points of the network 7.5.2 A cash-flow analysis 7.5.3 Value of embedded generation – regional and local issues 7.5.4 Capacity credit 7.5.5 Summary Support Mechanisms for Renewable Energy 7.6.1 Introduction 7.6.2 Feed-in law 7.6.3 Quota system 7.6.4 Carbon tax Electricity Trading 7.7.1 Introduction 7.7.2 The UK electricity supply industry (ESI) 183 183 184 184 184 185 186 186 187 187 187 188 188 189 190 190 192 193 195 195 195 195 197 198 198 200 201 203 204 205 205 205 206 207 208 209 209 210 212 213 215 215 215 216 217 217 218 218 218 7 7.1 7.2 7.3 7.4 7.5 7.6 7.7
Slide 12: x Contents 7.7.3 Competitive wholesale markets in other countries 7.7.4 The value of renewable energy in a competitive wholesale market References 8 8.1 8.2 The Future – Towards a Sustainable Electricity Supply System Introduction The Future of Wind Power 8.2.1 Large wind turbines 8.2.2 Offshore wind farm development 8.2.3 Building integrated wind turbines The Future of Solar Power 8.3.1 PV technology development 8.3.2 Solar thermal electric systems The Future of Biofuels The Future of Hydro and Marine Power Distributed Generation and the Shape of Future Networks 8.6.1 Distribution network evolution 8.6.2 Active networks 8.6.3 Problems associated with distributed generation 8.6.4 Options to resolve technical difficulties Conclusions References 223 226 229 231 231 232 232 233 238 240 240 241 242 243 244 244 245 246 246 249 250 253 253 253 255 255 256 257 258 259 260 261 263 265 265 265 266 266 267 269 269 271 272 273 275 276 277 8.3 8.4 8.5 8.6 8.7 Appendix: Basic Electric Power Engineering Concepts A.1 Introduction A.2 Generators and Consumers of Energy A.3 Why AC? A.4 AC Waveforms A.5 Response of Circuit Components to AC A.5.1 Resistance A.5.2 Inductance A.5.3 Capacitance A.6 Phasors A.7 Phasor Addition A.8 Rectangular Notation A.9 Reactance and Impedance A.9.1 Resistance A.9.2 Inductance A.9.3 Capacitance A.9.4 Impedance A.10 Power in AC Circuits A.11 Reactive Power A.12 Complex Power A.13 Conservation of Active and Reactive Power A.14 Effects of Reactive Power Flow – Power Factor Correction A.15 Three-phase AC A.16 The Thévenin Equivalent Circuit Reference Index
Slide 13: Foreword By Jonathon Porritt You can read the current state of awareness about climate change any which way you want. You can continue to ignore (or even deny) the overwhelming scientific consensus that has gradually emerged over the last few years. You can get totally lost in the intricacies of climate policy and the political controversies about who is doing or not doing what. But 20 years into the debate about climate change, one thing is overwhelmingly clear: the future of human kind depends in large measure on the speed with which we can massively expand the contribution of renewable energy to our overall energy needs. That the world is now on a collision course is not seriously disputed. The International Energy Agency constantly reminds people that overall energy use will at least double by 2030 and that most of that expansion will be powered by growth in fossil fuels. On the other hand, climate scientists now tell us that we will need to reduce emissions of CO2 and other greenhouse gases by at least 60% by 2050. It doesn’t remotely begin to add up. Which makes it hard to understand why so many people are still so crabby and cautious in defining the role for renewables. All their projections are based on ‘business-as-usual’ economic models – as if any of those are going to be terribly relevant for very much longer. Indeed, this is the one area where I believe it really is legitimate to talk about ‘going onto a war footing’ in combating the threat of runaway climate change. And that may not be so far off. For instance, if the price of oil stays at or around $100 a barrel, and the price of a tonne of CO2 rises rapidly over the next 3 or 4 years, much of the rubbish still being talked about renewables being ‘uneconomic’ will just wither away. That, however, is only the start of it. I have been giving lectures to CREST students for the best part of 10 years, and have learnt during that time that even if the technologies themselves are rapidly improving, and even if the political and economic context could be completely transformed, as I believe is now possible, the real challenge lies in accommodating high penetrations of these new technologies in the electricity supply system, by adapting existing networks and/or the creation of new infrastructure for transmission and distribution. That’s where much of the innovation (and huge amounts of new investment) will be needed over the next few years. And that is one of the greatest strengths of this hugely informative new book: connecting up all the dots so that a clear and utterly convincing picture emerges. And that means taking
Slide 14: xii Foreword proper account of the critical importance of energy efficiency (so often ignored in treatments of renewable energy), energy security, and the kind of governance systems which will be needed to drive forward so very different an energy economy. This is complex, challenging territory, for which reliable and very experienced guides are strongly recommended! Jonathon Porritt is Founder Director of Forum for the Future www.forumforthefuture.org. uk, Chairman of the UK Sustainable Development Commission www.sd-commission.org.uk, and author of Capitalism as if the World Matters; Revised Edition 2007 (in paperback), Earthscan – available through ’Forum for the Future’ website.
Slide 15: Preface There is worldwide agreement on the need to reduce greenhouse gas emissions, and different policies are evolving both internationally and locally to achieve this. On 10 January 2007 the EU Commission announced an Energy Package which was endorsed by the European Council. The objectives are that by 2020 EU greenhouse gases are to be reduced by 30 % if a global agreement is arrived at or by 20 % unilaterally. One of the vital components in the achievement of this goal is the intention to provide a 20 % share of energy from renewable energy (RE) sources in the overall EU energy mix. At present, wind power is the leading source of new renewable energy. World wind power capacity has been growing rapidly at an average cumulative rate of 30 % over the last ten years. About 20 GW of new capacity was installed in 2007 bringing the world total in that year to 94 GW. This annual investment represents around 25 billion euros by an industry that employs 200 000 people and supplies the electricity needs of 25 million households. This considerable expansion has attracted investment from major manufacturing companies such as General Electric, Siemens, ABB and Shell as well as numerous electricity utilities, notably E.ON and Scottish Power. The future of wind power over the next two decades is bright indeed. Generation of electricity from the sun can be achieved directly using photovoltaic (PV) cells or through solar concentration to raise steam and drive conventional turbines. Over the last few years considerable progress has been made in the reduction of the cost of PV generated electricity, with 2006 seeing the total value of installed capacity reaching 15 billion euros and with cell global production in that year approaching 2.5 GW. It is expected that further technology improvement and production cost reduction over the next decade will result in wide scale competitive generation from this source. Marine energy is an exciting, but less well developed technology. Tidal barrages, tidal stream turbines and wave energy devices are all in the experimental and pre-commercial stage but are expected to make a significant contribution by around 2015. Geothermal energy is now established in countries like Iceland with a significant accessible resource, and as the technology develops could be taken up more widely. Last but not least there are bioenergy and biofuels, important because they offer many of the advantages of fossil fuels, in particular being easily stored. Not surprisingly they are receiving much attention from policy makers and researchers both in the EU and North America. Most of this renewable energy will be converted into electricity. The renewable energy resource will be geographically highly distributed, and being mostly dependent on changing weather and climate cannot be directly controlled in the way fossil fuelled generation is. Electrical power networks were designed to operate from electricity generated in a few large power stations fuelled by coal, gas or uranium, fuels readily available on the international market and to varying degrees controllable. Significantly increasing the input from renewable energy sources requires a revision of the way power systems are designed and operated in
Slide 16: xiv Preface order to accommodate these variable sources better. This book is an introduction to this important topic. The material in this book is largely based on a Master ’s course module taught for over ten years at the Centre for Renewable Energy Systems Technology (CREST) at Loughborough University. The course as a whole was designed to provide general technical education in all major electricity generating renewable energy sources and their integration in electrical networks. Students taking this course normally have first degrees in numerate topics ranging from Physics or Engineering to Environmental Science. The course modules are therefore designed for students who, although they may be very knowledgeable in their speciality, will only have elementary knowledge of other topics. Likewise, this book assumes no previous knowledge in power systems engineering and guides the reader through the basic understanding of how a power system is put together and the way in which it ensures that the consumer demand is met from instant to instant. The characteristics of traditional and renewable energy (RE) resources are described with special reference to the variability of the latter and the way this impacts on their utility. These resources are available in a form that either has to be converted into electricity and/or their electrical output has to be conditioned before it can be fed into the grid. The book covers these aspects and stresses the importance of power electronic technology in the process of power conditioning. The power flows in an electricity network have to be appropriately controlled and the book addresses the way this is achieved when these new sources are integrated. The economics of renewable sources will determine their take-up by the market, and this issue is also addressed, and in some detail. Finally, an eye is cast on the future development of RE technologies and the way that power systems may evolve to accommodate them. An Appendix is available for readers who require a more mathematical coverage of the way electricity is generated, transported and distributed to consumers.
Slide 17: Acknowledgements This book contains input from other CREST staff besides the main authors. In particular, Dr Murray Thomson has provided much of the power electronics material of Chapter 4 and most of the content of Chapters 5 and 6. His comments and criticisms during the initial development of the book have been invaluable. In addition, Dr Simon Watson contributed most of the material found in Chapter 7. The material on dynamic demand control in Chapter 3 was the subject of a CREST Master ’s dissertation by J. A. Short. Finally, thanks are due to Dr Graham Sinden who gave us permission to use several diagrams from his recent work including some unpublished ones from his doctorate thesis and to Mr David Milborrow for permitting us to use in Chapter 7 several of his tables and figures. We are also grateful for the support of the Wiley staff in Chichester who have guided us in the process of preparing the manuscript for publication. Last, but certainly not least, we would like to dedicate this book to our respective partners Delphine Freris and Marion Peach who have had to put up with us slaving over the text in our spare time, rather than participating more fully in family life. Leon Freris and David Infield
Slide 19: 1 Energy and Electricity 1.1 The World Energy Scene 1.1.1 History Energy demand in the pre-industrial world was provided mostly by man and animal power and to a limited extent from the burning of wood for heating, cooking and smelting of metals. The discovery of abundant coal, and the concurrent technological advances in its use, propelled the industrial revolution. Steam engines, mechanized production and improved transportation, all fuelled directly by coal, rapidly followed. The inter-war years saw the rise of oil exploration and use. Access to this critical fuel became a key issue during the Second World War. Post-war industrial expansion and prosperity was increasingly driven by oil, as was the massive growth in private car use. More recently a new phase of economic growth has been underpinned to a great extent by natural gas. A substantial proportion of coal and gas production is used to generate electricity, which has been widely available now for over a century. Electricity is a premium form of energy due to its flexibility and ease of distribution. Demand worldwide is growing, driven by the explosion in consumer electronics, the associated industrial activity and the widening of access to consumers in the developing world. 1.1.2 World Energy Consumption The present global yearly primary energy1 consumption is, in round figures, about 500 EJ.2 This is equivalent to about 1.4 x1017 W h or 140 000 TW h . Dividing this figure by the number of hours in the year gives 16 TW or 16 000 GW as the average rate of world primary power Primary energy is the gross energy before its transformation into other more useful forms like electricity. The unit of energy in the SI system is the joule, denoted by J. Multiples of joule are kJ, MJ, GJ, TJ (T for tera denoting 1012) and EJ (E for Exa denoting 1018); the unit of power is the Watt (W) and represents the rate of work in joules per second. Electrical energy is usually charged in watt-hours (W h) or kW h. Joules can be converted into W h through division by 3600. 2 1 Renewable Energy in Power Systems Leon Freris and David Infield © 2008 John Wiley & Sons, Ltd
Slide 20: 2 Renewable Energy in Power Systems 1. other renewables 0.4% 1 2. nuclear 6.5% 2 7 3 4 6 5 3. natural gas 20.9% 4. large hydro 2.2% 5. coal 25.1% 6. biomass and waste 10.6% 7. oil 34.3% Figure 1.1 Percentage contribution to world primary energy consumption. The pie chart in Figure 1.1 shows the percentage contribution to world primary energy from the different energy sources according to data taken from the International Energy Agency (IEA) Key World Energy Statistics, 2006. The world demand for oil and gas is increasing significantly each year. The major part of this increase is currently taken up by India and China where industrialization and the demand for consumer products is escalating at an unprecedented pace. The world consumption in 2006 increased by more than twice Britain’s total annual energy use and is the largest global yearly increase ever recorded. China alone accounted for roughly 40% of this increase. The IEA forecasts that by 2030 demand for energy will be some 60% more than it is now. 1.1.3 Finite Resources It is extremely difficult to determine precise figures on the ultimate availability of fossil fuels. According to the major oil and gas companies, still significant new resources of oil are being developed, or remain to be discovered. A safe assessment is that there is enough oil from traditional sources to provide for the present demand for 30 years. The latest figures for global gas reserves indicate that these are approximately 50% higher than oil at some 60 years of current demand, and gas is far less explored than oil so there is probably more to be found. There are, however, unconventional hydrocarbon resources such as heavy oil and bitumen, oil shale, shale gas and coal bed methane – whose total global reserves have been assessed very roughly to be three times the size of conventional oil and gas resources. These are more expensive to extract but may become exploitable as the price of fossil fuels increases due to the steady depletion of the more easily accessible reserves. Fortunately for fossil fuel dependent economies, coal reserves are considered to be many times those of oil and gas and could
Slide 21: Energy and Electricity 3 last for hundreds of years. The downside of coal is its high carbon content, a topic to be discussed later. Much debate is currently focused on when the so-called peak oil and gas might occur. This is when the oil and gas extraction rate starts to fall and occurs well before resources run out. It is important because it signals that demand will most likely not be fully met, with prices rising significantly as a consequence. Certainly the UK’s North Sea reserves of oil and gas are fast declining with peak extraction having already occurred in 2003. Given the enormous investment in extraction and supply infrastructure, and the profits to be made, it would be surprising if those with vested interests did not work hard to maintain confidence in these sources. Fuel for nuclear fission is not unlimited and several decades ago this has prompted interest in the fast breeder reactor which in effect extends the life of the fuel. However, the political dangers inherent in the fast breeder cycle, with its production of weapons grade plutonium, has limited its development to a few prototype reactors which had major operational problems and are now defunct. The lifetime of uranium reserves for conventional fission at current usage has been estimated by some as around 50 years, but such calculations are very dependent on assumptions. If an extremely high ore price is tolerable, then very low grades of uranium ore can be considered as possible reserves. The DTI cites OECD/NEA ‘Red Book’ figures to claim that based on 2004 generation levels, known uranium reserves (at $130/kg) will last for around 85 years (see References [1] and [2]). 1.1.4 Energy Security and Disparity of Use Energy security is a major concern worldwide. A large part of the world’s oil is located in the Middle East and other politically unstable countries. The conflict between ‘Western’ and ‘Islamic’ cultures is at present exacerbating the anxiety over reliability of energy supply. Russia is a major producer of gas but recent events in Ukraine have made European countries aware how dependent they are on this single source. The USA is the world’s largest consumer of energy and is heavily dependent on imported oil. With economic growth seen as being intrinsically linked to cheap fuel it is difficult to imagine political parties, in the USA or elsewhere, proposing policies that require voters to drastically curtail their consumption and therefore alter their lifestyles. Another disturbing aspect is the disparity in consumption between rich and poor countries: the richest billion people on the planet consume over 50% of all energy, while the poorest billion consume around 4%. This is an added source of tension and of accusations that the developed countries are profligate in the use of energy. To excuse this high consumption on grounds of high industrial activity is simply wrong. Japan, for example, is the world’s second largest economy but has a per capita energy consumption half that of the US. 1.2 The Environmental Impact of Energy Use 1.2.1 The Problem Fossil fuels have one thing in common: they all create carbon dioxide when burnt. They are a key part of the Earth’s long term carbon cycle, having been laid down in geological periods
Slide 22: 4 Renewable Energy in Power Systems when the climate was tropical across much of the planet and atmospheric CO2 concentrations were very high. This storing of carbon through the growth of plant matter, and its subsequent conversion to coal, oil, peat and gas, dramatically reduced atmospheric CO2 levels and played an important role in cooling the planet to temperatures that could support advanced life forms. The concern now is that by unlocking this stored carbon climate change is being driven in the other direction, with global warming the direct result of an excessive greenhouse effect. Ice core samples indicate that the level of carbon dioxide in the atmosphere was more or less stable at 280 parts per million (ppm) over the last few thousand years up to the onset of the industrial revolution at the beginning of the nineteenth century. Subsequently, atmospheric CO2 levels rose, at first slowly as a result of coal burning but since the Second World War the release of CO2 has accelerated reflecting the exploitation of a wider range of fossil fuels. Current CO2 levels are 380 ppm and rising fast. CO2 is not the only pollutant created by fossil fuelled generation: combustion in air comprising 78% nitrogen by volume inevitably produces nitrogen oxides, NO, and NO2 and N2O, collectively known as NOx; and any sulfur content of the fuel results in SOx emissions. NOx and SOx together contribute to acid rain and as a result it is now common to reduce any SOx emissions from fossil fuelled power stations through flue gas desulfurization. The downside of this is reduced thermodynamic efficiency and some resulting increase in CO2 emissions. World coal reserves are substantial, but coal is a less attractive fuel from the point of view of CO2 emissions and also much more disruptive to extract. The cheapest coal is from opencast mines, but this process is immensely damaging to the environment. All forms of generation have some environmental impact, but these are not in general reflected in the cost of electricity; because of this, these additional environmental costs are known as externalities. Externalities are consequences of activity that are not normally a part of the economic analysis; for example the cost to society of ill health or environmental damage arising from pollution caused by a specific generating plant is not directly charged to the operator, i.e. it is external to the microeconomics of the plant’s operation. A number of European countries now seek to bring these externalities back into the economics of electricity generation by some kind of environmental levy or carbon tax. Carbon trading, discussed in detail in Chapter 7, is an alternative means of achieving this goal. The nuclear cycle is of course not without externalities, although the environmental costs are highly contested, contributing as they do to the economic attractions or otherwise of nuclear power. Radioactive waste disposal, radioactive emissions and final decommissioning and disposal of radioactive reactor components are rarely fully accounted for and thus fall to an extent into the category of externalities. There are also issues concerned with environmental damage associated with uranium mining, but in this regard it is similar to coal. If nuclear power is to mitigate global emissions, it is of vital importance to assess accurately how much CO2 will be displaced by nuclear power. This is a topic fraught with controversy. The well established 386 g CO2/kW h contributed by gas fuelled power stations will be taken as the benchmark. The emissions for nuclear power are quoted as 11–22 by OECD, and 10–130 by ISA, University of Sydney [3]. If the upper figures are valid, the contribution of nuclear power to CO2 mitigation may be seriously compromised. Clearly this is an issue that requires certainty.
Slide 23: Energy and Electricity 5 1.2.2 The Science The science of climate change is very well established and its primary goal is to understand the link between CO2 and other greenhouse gas concentrations and temperature rise. Work in this area has been carried out by the Intergovernmental Panel on Climate Change (IPCC), which was set up in 1988 by the World Meteorological Organisation and the United Nations Environment Programme. It involves scientists from 169 countries. Figure 1.2 shows the changing average global temperature, from 1850 to 2005. The bold curve is the smoothed trend while the individual annual averages are shown as bars. The temperatures are shown relative to the average over 1861–1900. The earth has warmed by 0.7 °C since around 1900, bringing the global temperature to the warmest level in over 12 000 years. All ten warmest years on record have occurred since 1990 and there is considerable physical and biological evidence confirming climate change. Most climate models indicate that a doubling of greenhouse gases since the pre-industrial period is very likely to result in a rise between 2–5 °C in global mean temperatures. This increased level is likely to be reached between 2030 and 2060. If no action is taken concentrations would be more than treble preindustrial levels by 2100, resulting in a warming of 3–10 °C according to the latest climate projections. Although the relationship between CO2 concentration, temperature change and undesirable climatic changes is very complex and thus hard to predict precisely, it is widely believed that the CO2 concentrations have to be stabilized if damaging global warming is to be avoided. The IPCC concluded in 2001 [4] that there is strong evidence that most of the warming observed over the last 50 years is anthropogenic in that it is attributed to human activities. This was supported by the Joint Statement of Science Academies (2005) and a report from the US Climate Change Science Programme (2006). An IPCC updated report which was published in 2007 confirmed this link with greater certainty. A summary of recent scientific research may be found in Reference [5]. Global Average Near–Surface Temperatures 1850–2005 1.0 Temperature Difference (°C) with respect to 1861–1900 0.8 0.6 0.4 0.2 0.0 –0.2 –0.4 1860 1880 1900 1920 Based on Brohan et al. (2008) 1940 1960 1980 2000 Figure 1.2 Temperature rise record. (© Crown copyright 2007, the Met Office)
Slide 24: 6 Renewable Energy in Power Systems 1.0 Temperature change(°C) observed model simulation 0.5 0.0 –0.5 1850 1900 1950 2000 Figure 1.3 Office) Natural factors cannot explain recent warming. (© Crown copyright 2007, the Met The basic evidence that provides confirmation that global warming is due to human-made factors was provided by a climate model developed at the Hadley Centre for Climate Prediction and Research in the UK. In Figure 1.3, the observed global temperature since the early 1900s is shown by the bold line. The climate model was driven over that period by natural factors such as output of the sun, changes in the optical depth of the atmosphere from volcanic emissions and the interactions between the atmosphere and oceans. The predictions of the model are shown by the fuzzy band. This clearly disagrees with the observations particularly since about 1970, that observed temperatures have risen by about 0.5 °C, but those simulated by natural factors have not changed at all. If the climate model is now driven by natural factors as previously but in addition by manmade factors – change in greenhouse gas concentrations and sulfate particles – the model simulation predictions in Figure 1.4 are in much better agreement with the temperature record. Climate modelling studies by other research centres have arrived at the same broad conclusions. 1.2.3 The Kyoto Protocol The effects of climate change are global and hence mitigation requires coordinated international effort. Signed in 1997, the Kyoto Protocol aims at reducing greenhouse gas emissions in the period 2008 to 2012 to 5.2% below those in 1990. Emissions of greenhouse gases by the US are currently 20% higher than in 1990 while the target figure in Kyoto was a cut of 7%. In the long run however it is prudent for industrialized countries to reduce their emissions by 60% by 2050 if the worst effects of climate change are to be mitigated with any confidence. This is a major challenge, to individuals, to governments and to supranational bodies. Greatest responsibility rests of course with the nations producing the largest CO2 emissions per capita and those moving fast up the emissions table. Table 1.1 illustrates the variation in emissions per head and how this is partly driven by the income per head. Emissions from China are expected to surpass those of the US by 2025 so there is much to be done.
Slide 25: Energy and Electricity 7 1.0 Temperature change(°C) observed model simulation 0.5 0.0 –0.5 1850 1900 1950 2000 Figure 1.4 Climate warming can be simulated when man-made factors are included. (© Crown copyright 2007, the Met Office) Table 1.1 Energy related CO2 emissions. (Reproduced with permission from Climate Analysis Indicators Tool (CAIT) version 4.0 (or 5.0). World Resources Institute, 2007, available at http://cait.wri.org) Country/grouping USA EU UK Japan China India World CO2 per head (tCO2) 20.4 9.4 9.6 9.8 3.0 1.1 4.0 GDP per head ($) 34 430 23 577 27 276 26 021 4 379 2 555 7 649 1.2.4 The Stern Report The economics of climate change mitigation are crucial in steering an optimal policy towards a given agreed goal. These issues were addressed in the Stern Report [6]. Although the experts are almost universally convinced that climate change is taking place they are uncertain as to what exactly will be its effects. The Stern Report is unique in the sense that it examines the probabilities of reaching certain temperature thresholds at different stabilization levels. These probabilities have only been established recently and provide the basis for the economics of the analysis of the risks and costs involved in taking a range of actions towards reducing the greenhouse emissions. The key message of the Stern Report is summarized below: • The lags in the climate change process must be recognized. What is going to happen to the climate over the next 20–30 years is already determined and irreversible. Actions over the next 20–30 years will affect what happens in the decades to come.
Slide 26: 8 Renewable Energy in Power Systems • • • • • Climate change threatens the basic elements of life, i.e. access to water, food, health and the use of land and the environment. There is still time to avoid the worst impacts of climate change if action is taken now. Stabilization at 550 ppm of all greenhouse gases is recommended, but this would involve strong action. The costs of stabilizing the climate are significant (1% of global GDP) but manageable. Delay would be dangerous and much more expensive, perhaps as costly as 20% of global GDP. Action demands an international response. The key actions should include: • • • • • Increase in efficiency of energy use. Strict emissions trading rules to support the transition to low carbon development paths. Extensive use of renewable and other low carbon technologies. Technology cooperation and fivefold increased in low carbon technologies R&D. Reduction in deforestation. The major focus of the Stern Report is the economics of climate stabilization. Figure 1.5 shows estimates of costs of low carbon technologies in 2015, 2025 and 2050 that may be used to constrain CO2 emissions. The costs are expressed as a central estimate, with a range, and as a percentage of the fossil fuel alternative in the appropriate year. Due to learning effects the costs fall over time. The ranges reflect judgements about the probability distribution of unit costs and the variability of fossil fuel prices. The 0% line indicates that the costs are the same as the corresponding fossil fuel option. As expected, the uncertainties are large even for short term predictions. Onshore wind is shown to be particularly attractive with photovoltaic (PV) cells becoming very attractive beyond 2025. On the basis of the costs of the low carbon technologies and assumptions on possible rates of uptake over time, the Stern Report estimates the distributions of emissions savings by technology for 2025 and 2050 for the desirable climate stabilization at 550 ppm. These estimates are shown in Figure 1.6. Energy efficiency and carbon capture and storage (CCS) play a major role in this scenario and will be discussed later in this chapter. Contributions from wind, solar, biofuels, hydro and distributed combined heat and power (dCHP) through electricity generation provide the remaining savings; and these are the technologies to be addressed in later chapters of this book. 1.2.5 Efficient Energy Use Figure 1.6 stresses that efficiency measures are projected to make the largest contribution in climate change mitigation. It is therefore a surprise that the important topic of rational and efficient use of energy is rarely pursued vigorously in national or supranational plans in spite of the fact that study after study has shown that this route provides the most cost effective way to meet sustainability goals. In most countries, regulations and financial incentives are now in place to encourage energy efficiency but their effect is modest and national energy consumption figures continue to
Slide 27: Energy and Electricity 9 –100 Electricity from gas with CCS Electricity from coal with CCS Nuclear power Electricity from energy crops Electricity from organic wastes Onshore wind Offshore wind Solar thermal (v. sunny regions) PV (sunny regions) dCHP using H from NG or coal with CCS Hydrogen from NG or coal (CCS) - industry Hydrogen from NG or coal (CCS) - distributed Electrolytic hydrogen - industry Electrolytic hydrogen - distributed Biomass for heat - distributed Bioethanol Biodiesel Hydrogen ICE vehicle – fossil H (+CCS) FC Hydrogen vehicle – fossil H (+CCS) FC Hydrogen vehicle – electrolytic H 0 Cost as % of fossil fuel option 100 200 300 400 500 600 Cost in 2015 Cost in 2025 Cost in 2050 Figure 1.5 Unit costs of energy from low carbon technologies: CCS stands for carbon capture and storage, dCHP stands for distributed combined heat and power. (Reproduced from Stern review website, copyright Cambridge University Press) Contributions to carbon abatement 2025 8 7 5 Contributions to carbon abatement, 2050 8 1 2 3 1. Efficiency 2. CCS 3. Nuclear 4. Biofuels 5. dCHP 6. Solar 7. Wind 8. Hydro 6 4 3 2 1 1. Efficiency 2. CCS 3. Nuclear 4. Biofuels 5. dCHP 6. Solar 7. Wind 8. Hydro 7 6 5 4 Abatement 11 GtCO2 Abatement 43 GtCO2 Figure 1.6 The distribution of emission savings by technology. (Reproduced from Stern Report, copyright Cambridge University Press) rise year on year. Energy efficiency must be the linchpin of any future energy strategy because [7]: • Using energy as efficiently as possible is the most cost effective way to manage energy demand, and thus to address carbon emissions. Saving energy is cheaper than making it.
Slide 28: 10 Renewable Energy in Power Systems • • • By reducing demand on gas and electricity distribution networks, energy efficiency will improve the security and resilience of these networks and reduce dependence on imported fuels. By reducing energy bills, energy efficiency will help businesses to be more productive and competitive. Improving the energy standards of homes has an important role in reducing spending on fuel by those in fuel poverty. Increasing energy end use efficiency is unattractive for energy companies driven by commercial imperatives to increase sales and profits. It thus falls to governments to implement policies that change these drivers. Regulations can be put in place for example that require utilities to encourage customers to use electricity efficiently. A more revolutionary approach envisages the utility being transformed into a supplier of energy services, owning appliances in people’s homes and thus being motivated to maximize the efficiency of these appliances. Whatever approach is finally adopted, the importance of reducing energy consumption should be the cornerstone of any CO2 mitigation programme. 1.2.6 The Electricity Sector Figure 1.7 shows the percentage of fuels used in the generation of electricity. Fossil fuels account for 64% of the fuels used in this sector with coal being the dominant source at approximately 40% and contributing nearly three quarters of CO2 emissions. At present, large hydropower plants account for the major part of the renewables sector. Under half of the electricity produced is used in buildings, about a third in industry, under one-tenth in energy production (e.g. refineries) and less than one-tenth in transmission and distribution. The world annual generation of electricity is in the region of 18 000 TW h representing an average rate of consumption of around 2000 GW. This electrical energy is generated in a very large collection of power stations driven mostly by fossil fuels. The electricity sector is the fastest growing source of emissions and estimated to increase fourfold between now and 2050. According to Stern this sector would need to be at least 60% decarbonized by 2050 1 1. Oil 10% 5 2 2. Nuclear 16% 3. Gas 15% 4. Hydro 19% 3 4 5. Coal 39% Figure 1.7 Contributions in the generation of electricity. (Data from Boyle, G., Renewable Energy, Oxford University Press, 2004)
Slide 29: Energy and Electricity 11 for atmospheric concentration to stabilize at 550 ppm, thereby reducing the risk of catastrophic climate change. 1.2.7 Possible Solutions and Sustainability Fundamental choices will have to be made in the years ahead. Societies are presently dependent on high and growing fossil fuel consumption. The possibility of weaning people from this dependence over a short timescale is completely unrealistic. The general shift in fossil fuels over the last two decades for both electricity generation and heating has been towards increased use of gas in place of coal, and to a lesser extent oil. This has helped to limit the growth in CO2 emissions as gas combustion releases less CO2 per unit of energy than coal. Political events, however, have generated anxiety in the EU and elsewhere in relation to increased dependence on this particular fuel. A possible alternative path is to revert to dependence on coal. This resource is abundantly available in many developed counties including the US, Australia, many EU countries, Canada, Russia and in developing countries such as China, South Africa and Turkey. Recent developments in CO2 capture or ‘sequestration’ for fossil fuels, discussed later in this chapter, give some hope that this source may be made more acceptable environmentally. A number of potentially carbon neutral sources exist: these include nuclear fission (and possibly fusion in the far future) and all sources that derive directly or indirectly from the sun, namely biomass, wind, solar (thermal and photovoltaic), hydroelectric and marine. Geothermal and tidal energy are also carbon neutral and often regarded as renewable on the grounds that the sources are so huge as to be virtually inexhaustible. In Chapter 2, the characteristics of all these conventional and emerging technologies are discussed in some detail. Finally, but no less important, other approaches essential in the move towards sustainability are a reduction in energy needs and improvements in the efficiency of energy use. The latter includes more efficient electricity generation. Although not the main focus of this book, these topics are briefly discussed in this chapter. The planet’s reserves of fossil fuels and minerals are of course finite, and thus the exploitation of coal, oil, gas and uranium are not sustainable in the longer term. Fortunately, renewable energy (RE), being derived from naturally occurring energy flows, is inexhaustible and has no long term detrimental effect on the environment. As such it will in time become the basis of the energy supply system, and probably the sole means by which electricity is generated. 1.3 Generating Electricity 1.3.1 Conversion from other Energy Forms – the Importance of Efficiency Figure 1.8 shows the ways in which various types of energy can be converted into electricity. At present, the path generating the bulk of electricity worldwide is shown by the bold lines that lead through combustion from chemical to thermal, from thermal to mechanical and finally to electrical power conversion. The bottleneck of this path is the limited thermodynamic efficiency determined by the Carnot cycle. Older thermal generating stations have
Slide 30: 12 Renewable Energy in Power Systems Nuclear Fission (Fusion) Heat engines Thermal n < 60% Gravitational (Hydro tidal) Wind wave Electric generators Mechanical n = 90%+ Photovoltaic Electrical n = 90%+ Fuel cells Solar thermal Solar Chemical (Coal, oil, gas, biomass, hydrogen) Figure 1.8 Conversion from a variety of energy forms into electricity efficiencies between 35 and 40% although in the last two decades conversion has been substantially improved to over 50% through the development of combined cycle gas turbines (CCGTs) a technology discussed in Chapter 2. It follows that when coal, oil or gas is used only 35–50% of the primary energy is successfully converted, the remaining being discharged into the environment in the form of waste heat. One way of getting around the Carnot limit is simply to make use of the waste heat. This is the principle of combined heat and power (CHP), used extensively in Scandinavia and of growing importance elsewhere. In such schemes, the waste energy from the thermal generation of electricity is distributed through heat mains to local industry and/or housing. This requires substantial infrastructure and is therefore only viable if the power station is reasonably close to the heat users. An alternative arrangement made possible by recent developments is to transport the fuel (mainly gas) to the consumer using the existing supply infrastructure and install the CHP system at the consumer ’s premises. Such systems are known as micro-CHP and are discussed in Chapter 8. Direct paths that bypass the Carnot bottleneck are also available. The leading example of this approach is the fuel cell, which now borders on commercial viability in a number of forms: solid oxide, molten carbonate, and proton exchange membrane (PEM) to name the main ones. Another direct path is through photovoltaics, a technology that perhaps is the most promising in the near to far future. The conversion efficiencies of the various routes indicated in Figure 1.8 are dealt with in greater detail in Chapter 2. 1.3.2 The Nuclear Path The topic of electricity generation from nuclear power elicits strong emotions from supporters and critics. Although nuclear power supplies only the equivalent of 5.7% of the world primary
Slide 31: Energy and Electricity 13 energy at the time of writing this book, some believe this should be expanded massively. They argue that it is an attractive source of electricity, having very low carbon emissions. After the Three Mile Island and the Chernobyl accidents there was a period of nearly ten years during which almost no new nuclear capacity was constructed. However, the recent concerns regarding fossil fuel security have prompted a number of countries to consider new building programmes. China and India are planning to build several tens of reactors each and the USA is posed to do the same. In contrast within Europe, only Finland has embarked on the construction of a new nuclear plant while, Sweden, Switzerland and Germany all have moratoriums in place leading to a phasing out of nuclear power. France on the other hand, remains committed to nuclear power which contributes about 80% of its present electricity needs. In the UK the 2003 government White Paper was critical of the nuclear option, but by 2006, with concerns about a possible energy gap, the government’s position had changed. It is now supportive in principal of a new nuclear programme. A key concern is that major investments in nuclear will deprive renewable energy sources of the finance they need to expand. Reflecting its importance, the debate over nuclear power is extensive and there is voluminous literature. Reference [8] provides a good entry point for those interested. 1.3.3 Carbon Capture and Storage Figure 1.6 indicates, that according to Stern, by 2025 and 2050 about 20 and 40% respectively of carbon abatement is expected to be provided by the emerging technology of carbon capture and storage (CCS) provided that the technoeconomic and environmental issues can be satisfactorily dealt with. CCS has the significant advantage of reconciling the necessary use of fossil fuels in the medium term with the necessity of serious cuts in CO2 emissions. A large scale demonstration project was being planned in northeast Scotland, a joint venture by BP, Shell and ConocoPhilips [9]. Unfortunately, this project was recently abandoned, but the technology is being vigorously pursued in other projects worldwide. Figure 1.9 illustrates the geological storage options for CO2. With gas and oil prices likely to rise significantly, extracting CO2-free energy from coal is also attracting substantial attention. Such technologies take a number of forms, but the socalled integrated gasification combined cycle (IGCC) process is in the forefront. This involves the production of a synthetic gas (syngas) obtained from coal through gasification. Syngas is composed mainly of hydrogen and carbon monoxide and is the fuel source to a high efficiency plant operating in the combined cycle mode. Buggenum, a 253 MW plant in the Netherlands operates on this principle and is the cleanest coal based plant in Europe. To date, no IGCC plant involving carbon capture has been built although this principle is to be used as a basis for a zero carbon emission 275 MW plant funded by the US Department of Energy which is being built and should be up and running in 2013. 1.3.4 Renewables Figure 1.10 provides an overview of the earth’s main energy paths that can be tapped to generate sustainable electricity. The main source of easily accessible renewable energy
Slide 32: 14 Renewable Energy in Power Systems Geological Storage Options for CO2 Ultra clean fuels Clean electricity Coal ENHANCED COALBED METHANE UNMINEABLE COAL SEAMS ENHANCED OIL RECOVERY DEEP SALINE AQUIFERS Figure 1.9 Geological storage options for CO2. (Source: World Coal Institute) PATH Direct heating Direct radiation 120000 TW Water evaporation Absorbed by earth Heating of atmosphere Photosynthesis Gravitational forces (Moon–sun) Earth’s core 3TW 10TW RE Technology Thermal electric Photovoltaics Hydropower Wind/Wave converters Biofuels Tidal schemes Geothermal schemes Sun Figure 1.10 Renewable energy flow paths
Slide 33: Energy and Electricity 15 is the sun. On average the rate of solar radiation intercepted by the earth’s surface is about 8000 times as large as the average rate of world primary energy consumption. With the present world population this amounts to a staggering average power of 20 MW per person. The figure shows that this energy flux can be accessed directly using solar thermal or photovoltaic technology, or indirectly in the form of wind, wave, hydro and biofuels. Two other energy sources are often regarded as renewable in view of their sustainable nature: energy in the tides caused by the gravitational fields of the moon and the sun which can be tapped using tidal barrages or tidal stream technology; and geothermal energy from the earth’s core accessible in some locations through hot springs, geezers or boreholes. The available average power from these resources is a small fraction of that available from the sun. A substantial proportion of the incident radiation is reflected back to space. Over the last several millennia and up to the onset of the industrial revolution, energy inputs and outputs have been in equilibrium at a global temperature level suitable for the development of the earth’s biosphere. Exploiting the incident energy through the application of renewable energy technology does not disturb this balance. Intercepted natural energy flows, for example converted to electricity and then converted again by consumers into mechanical, chemical or light energy, all eventually degrade into heat. Most renewable energy forms are readily converted to electricity. Solar energy, geothermal energy and biomass can also be used to supply heat. Renewable energy can in principal provide all the energy services available from conventional energy sources: heating, cooling, electricity and, albeit with some difficulty and cost, transport fuels. It has the additional advantage that being a naturally distributed resource, it can also provide energy to remote areas without the need for extensive energy transport systems. It is worth noting that it is not always necessary to convert the renewable energy into electricity. Solar water heating and wind-powered water pumping are fine examples of systems that can work very well without involving electricity at all. However, the major contribution that renewable energy will be increasingly making in supplying people’s needs will be in electrical form. Renewable energy is currently experiencing dramatic growth. Wind power and solar PV are leading the growth with global companies such as GE and Siemens entering the wind energy market, and BP and Shell playing a major role alongside Japanese companies like Sharpe and Sanyo in PV. In China five of the largest electrical aerospace and power generation equipment companies have recently begun to develop wind turbine technology. Most large oil companies have expanded their research and development in ethanol and biodiesel production from biomass. The fastest growing RE technology is currently grid connected PVs with 40% annual year on year growth, but the RE technology that has made the largest contribution to date (excluding conventional hydro) is wind power with over 60 GW installed in EU countries and 95 GW worldwide by the end of 2007. At least 48 countries have national targets for RE supply including all 25 EU countries. Figure 1.11 shows the intended increase in contribution of the EU countries from 2002 to 2010. The EU has Europe-wide targets of 21% electricity and 12% of total energy by 2010. Table 1.2 shows the intentions of the EU over a wider period, with expected contributions in TW h per annum from various RE technologies.
Slide 34: 16 Renewable Energy in Power Systems Renewables Contribution to Electricity Production, 2002-2010 90 80 70 2002 60 % share of national total 2010 50 40 30 20 10 0 Austria Sweden Finland Portugal Denmark Italy Spain France Germany Greece Ireland Netherlands Czech United Poland Republic Kingdom Belgium Hungary Figure 1.11 RE contribution to European electricity production. (Source: Oxford Intelligence) Table 1.2 Contribution of renewables to electricity generation (1995–2020). (Source: Eurostat http:// epp.eurostat.cec.eu.int) 1995 Eurostat Wind (TWh) Photovoltaics (TWh) Biomass (TWh) Hydro (TWh) Geothermal (TWh) Total RES in EU 15 Total electricitya (TWh) Share of RES (%) a 2000 Eurostat 22.4 0.1 39.2 322 4.8 388 2574 15.1 2010 Projections 168 3.6 141 355 7.0 675 3027 22.3 2020 Projections 444 42 282 384 14 1166 3450 33.8 4 0.03 22.5 290 3.5 320 2308 13.9 EU trends to 2030. 1.4 The Electrical Power System 1.4.1 Structure of the Electrical Power System Electricity is widely used because it is a supremely flexible form of energy. It can be readily and efficiently transported and is easily converted to other forms of energy. Mechanical energy can be provided by very efficient motors, light energy by increasingly efficient light
Slide 35: Energy and Electricity 17 Superheaters Chimney Mechanical dust arrestors Conveyor house Main flue Electrical precipitator Coal bunker Boiler Cooling tower LP Turbine HP Generator Power station Generator transformer 22 kV Air Steam (trom boiler) Cooling water (outlet) Coal Condensed steam (returned to boiler) Condensers Current to transformer Cooling water (inlet) Step-up transformer Induced draught fan 33 kV Stepdown Large industrial consumer Small industrial consumer Domestic and commercial consumer Distribution substation Stepdown R Y S N E Forced draught fan Pulverised fuel mill Subtransmission system Bulk power substation Transmission system Circuitbreaker 132 kV (UK) 230/115 kV (USA) Step-down transformer 400 or 275 kV (UK) 500 or 345 kV (USA) Cable Secondary distribution 415 V 3-phase 240 V 1-phase (UK) 110 V 1-phase (USA) Primary distribution 11 kV Distribution transformer Figure 1.12 Pictorial view of the components of a large power system fittings, heat energy by 100% efficient resistive elements, and power supply to electronic and IT (information technology) hardware through very efficient power conditioning units. Figure 1.12 shows a diagrammatic layout of a typical electrical power system from the point of generation to the point of consumption. The figure depicts a coal fired power station as this represents the majority of world stations. The energy conversion chain follows the chemical → thermal → mechanical → electrical path depicted in Figure 1.8. Coal is pulverized and fed into a boiler where it is mixed with forced air and combusted. The boiler is a complex structure consisting of many stages of energy extraction from the combusted fuel. The flue gases are guided through equipment that removes solid particles and sulfur (desulfurization is not shown in the figure) before being released into the atmosphere. The highly purified water in the boiler is converted into superheated steam which is passed through several turbine stages on the shaft of a turbogenerator. The low pressure low temperature steam from the outlet of the turbine is condensed into the purified water which in this closed system is pumped back into the boiler. The condensing process unfortunately needs a substantial amount of external cooling water. In the figure, this water is provided from a pond at the bottom of a cooling tower. The hot water from the condenser is sprayed at the top of the tower and transfers its heat to the air that passes up the venturi shaped tower. The lost water must be made up from some external source such as a local river. The energy generated at the power stations is transmitted to consumers by overhead transmission lines and underground cables that possess ohmic resistance. The energy loss due to the unavoidable resistive heating of a line or cable is proportional to the square of the current I it carries. Additionally, for a given power transfer, which is proportional to VI, the current is inversely proportional to the voltage (other things being equal). Thus, the loss decreases
Slide 36: 18 Renewable Energy in Power Systems with the square of the voltage V. The downside of operating at higher voltages is that costs of insulation and other power system equipment increase substantially. Thus, for bulk transmission of power over long distances, higher voltages are most economic whereas, for local distribution of modest power to numerous connection points, lower voltages are most economic. The economics also dictate that it is worthwhile to have several intermediate voltages. This multiple voltage arrangement results in network transmission losses confined to within 5–10% of the throughput power. The bulk of global electricity is generated in large (>500 MW) power stations at around 20 kV. This is then stepped up by transformers to an extra high voltage (EHV) level such as 400 kV and carried by the transmission system to the bulk supply points, where it is stepped down to a high voltage (HV) level of around 100 kV. Some very large industrial consumers are connected at this level but most power is transformed down again to medium voltage (MV) levels such as 30 kV, then to 10 kV and finally to the low voltage (LV) level of 400 V, also referred to as the distribution system, which provides 230 V, when the connection is single-phase. In the USA and a few other countries, the LV level is 200 V three-phase, 115 V single-phase. The voltages used vary from country to country but the power system structure follows closely the layout of Figure 1.12. In Figure 1.12, a circuit-breaker is shown after the generator step-up transformer. This is a component part of an extensive protection network which permeates all levels of the power system. Faults on the network may result in low resistance paths that cause excessive currents capable of damaging equipment. The protection devices, circuit breakers at high voltage levels and fuses at domestic distribution level, operate to isolate the faulty part of the network and prevent equipment damage. The effect on the protection system of introducing renewable energy sources will be discussed in Chapter 6. 1.4.2 Integrating Renewables into Power Systems The term grid is often used loosely to describe the totality of the network. In particular, grid connected means connected to any part of the network The term national grid usually means the EHV transmission network. Integration specifically means the physical connection of the generator to the network with due regard to the secure and safe operation of the system and the control of the generator so that the energy resource is exploited optimally. The proper integration of any electrical generator into an electrical power system requires knowledge of the well-established principles of electrical engineering. The integration of generators powered from renewable energy sources is fundamentally similar to that of fossil fuelled powered generators and is based on the same principles, but, renewable energy sources are often variable and geographically dispersed. A renewable energy generator may be described either as standalone or grid-connected. In a standalone system a renewable energy generator (with or without other back-up generators or storage) supplies the greater part of the demand. In a grid-connected system, the renewable energy generator feeds power to a large interconnected grid, also fed by a variety of other generators. The crucial distinction here is that the power injected by the renewable energy generator is only a small fraction of that generated by the totality of generators on the grid. The distinction between standalone and grid-connected generators is a useful one but
Slide 37: Energy and Electricity 19 is not always clear-cut. Sometimes confusion arises when the word grid is used to refer to a relatively small standalone electrical network. This is not necessarily wrong (though it may indicate delusions of grandeur!) but it should always be clear as to the extent of the grid being referred to. The point on the network to which a renewable energy generator is connected is referred to, for reasons to be explained later, as the point of common coupling (PCC). 1.4.3 Distributed Generation Power systems have developed over the years to supply a varying demand from centralized generation sourced from fossil and nuclear fuels. Unless nuclear fusion proves successful, which will not be known for over 50 years, there is universal agreement that by the end of this century the majority of our electrical energy will be supplied from RE sources. Generators powered from renewable energy sources (except large scale hydro and large offshore and onshore wind farms) are typically much smaller than the fossil fuelled and nuclear powered generators that dominate today’s large power systems. Small generators cannot be connected to the transmission system because of the cost of high voltage transformers and switchgear. Also, the transmission system is often a long way away as the geographical location of the generator is constrained by the geographical availability of the resource. Small generators must therefore be connected to the distribution network. Such generation is known as distributed or dispersed generation. It is also known as embedded generation as it is embedded in the distribution network. In traditional power systems power invariably flows from the large centralized power stations connected to the EHV network down through the HV and LV systems to be distributed to consumers. In power systems with distributed generation power may travel from point to point within the distribution system. This unusual flow pattern has some serious implications in the effective operation and protection of the distribution network. Distributed generation will be discussed in Chapter 6. It may be concluded that present power systems will gradually have to evolve and adapt so that, in the far future, a managed demand will be supplied from distributed, mostly variable, RE generation. This transformation will be aided by the liberal use of power electronic interfaces capable of maximizing the effectiveness of RE sources, controlling power flows and ensuring reliability of supply. Some of these issues are discussed in the last chapter. 1.4.4 RE Penetration The proportion of electrical energy or power being supplied from renewable sources is generally referred to as the penetration. It is usually expressed as a percentage. When fuel or CO2-emission savings are being considered, it is useful to consider the average penetration: Average penetration = Energy from renewable energy powered generators ( kW h ) Total energy delivered to loads ( kW h ) In this case, the energy (kW h) is measured over a long period of time, perhaps a year. At first sight, it might seem more natural to express the denominator (total energy delivered to loads) as: total energy from all the generators (including fossil fuelled generators). However,
Slide 38: 20 Renewable Energy in Power Systems in standalone systems, there may be dump loads (loads where energy is dumped as heat) to consider and, in grid-connected systems, there is often interest in the penetration in a given geographical area, in which case it may be termed the local penetration. For other purposes, including system control, it is necessary to consider the instantaneous penetration: Instantaneous Penetration = Power from renewable energy powered generators ( kW ) Total power delivered to loads ( kW ) Since the electrical output from some generators operating from renewable energy sources is variable, the maximum instantaneous penetration will normally be much higher than the average penetration. References [1] Barnaby, F. and Kemp, J. ‘Too hot to handle? The future of civil nuclear power ’, Briefing paper, July 2007, Oxford Research Group. [2] Cited in ‘The future of nuclear power ’, DTI/Pub 8519/4K/05/07.NP, p.156; www.dti.gov.uk/files/file39197. pdf. [3] See www.pmc.gov.au/umpner/docs/commissioned/ISA_report.pdf. [4] IPCC, Third Assessment Report of the IPCC, Cambridge University Press, 2001. [5] Schellnhuber, H.J. (ed). Avoiding Dangerous Climate Change, Cambridge University Press, 2006 [6] Stern Review Report on the Economics of Climate Change, Cambridge University Press, 2006. Following criticisms of the report a rebuttal was published by Stern and his team and can be found in http://www.hm-treasury. gov.uk/media/5E1/FB/stern_reply_worldeconomics.pdf. [7] ‘Plan for action on energy efficiency’, April 2004, UK government follow-up to the ‘Energy White Paper ’. [8] ‘The role of nuclear power in a low carbon economy’, Sustainable Development Commission, UK, March 2006. [9] Gordon, S. ‘Carbon cure’, IEE Power Engineering, December 2005/January 2006.
Slide 39: 2 Features of Conventional and Renewable Generation 2.1 Introduction The purpose of this chapter is to describe the essential features of the different electricity generation plant, so that the role of the renewable sources and how they might be integrated into the electricity supply system can be better understood. The main features of conventional sources are dealt with in one section and in less detail. Such sources supply the overwhelming proportion of energy in power systems in which renewables are now being integrated but are themselves not the focus of this book. Such plant will remain for a long time to come and in the transition to a sustainable supply system they will need to complement the growing proportion generated from renewable sources. It is therefore essential that the reader has a basic understanding of conventional sources and their characteristics. Each of the renewable energy sources is treated in greater length and separately to reflect their rather diverse characteristics. First the rather obvious fact should be remembered that due to conservation of energy the time varying power demand of the consumers should be continuously matched by the generated power. In this situation, generation is said to be load following. How this is done and what happens when exact balance is not achieved is the topic of the next chapter. Conventional sources, taken here to include both fossil fuelled and nuclear generation, in general produce a given output when required to by the operator and so can be dispatched,1 whereas most renewable sources generate according to the time varying strength of the renewable energy source. These time variations are particularly noticeable in the case of wind, solar, wave and tidal power, although they are different in the extent to which the variations can be predicted, a factor that will turn out to be significant for efficient operation of the power system. Conventional generation in itself is neither instantly dispatchable nor completely flexible; it takes time to prepare large thermal plant for full load output and such plant are 1 Dispatching is the planned allocation of plant to meet expected future loads on the system. Renewable Energy in Power Systems Leon Freris and David Infield © 2008 John Wiley & Sons, Ltd
Slide 40: 22 Renewable Energy in Power Systems not permitted to operate at too low a load. Nuclear plant in particular offers very limited capability to follow changes in load. The remainder of this chapter explores in detail the characteristics of the different sources of generation with specific regard to their contribution to meeting the demands of a large power system. All the renewables are dealt with except geothermal power, which is very geographically limited, and where it does exist can be treated very much like fossil fuelled thermal power. It is after all simply a way of raising steam using large naturally occurring thermal gradients. 2.2 Conventional Sources: Coal, Gas and Nuclear The diverse characteristics of generators operating from different fossil fuel sources define their function in an integrated power system. Nuclear powered stations, dealt with in more detail at the end of the section, are generally inflexible and designed to run at constant power. This mode of operation is known as base load since it contributes to that fraction of the load that can be regarded as always present. Conventional coal plant efficiency, as reviewed in Chapter 1, is in the range 30–40% and depends significantly on how it is operated.2 The higher efficiencies would correspond to the best plant in base load operation; increased cycling 3 and low load operation of the plant significantly lowers operational efficiency. Peak system loads generally occur for only short periods of time and it turns out that such loads are best met by plant of low capital cost;4 the high operational costs associated with typically low efficiency generation of this peaking plant is acceptable since total operation time tends to be limited. Thermal conversion has been substantially improved over the last two decades through the development of combined cycle gas turbines (CCGTs). In this arrangement the gas, usually natural gas, is burnt at temperatures of around 1000 °C to drive a turbine and the exhaust gases are subsequently used to raise steam for a traditional steam turbine. Both turbines drive separate generators feeding power to the grid. By increasing the temperature at which the fuel is burnt efficiencies of around 50% can currently be achieved, with plant efficiencies towards 60% being projected for the next generation. Such plant are currently highly controllable, but because of their high efficiency they are constrained by fuel supply contracts that dictate running on high load. Most gas powered stations are therefore run at constant power and thus also contribute to the base load, although it is worth noting that recent rises in gas prices within Europe have recently changed the situation to one where it is, at least for the time being, economically preferable to run coal fired stations at base load in place of CCGT plant. Most conventional modern coal plants burn pulverized coal; future plant using gasified coal could reach a considerably higher overall efficiency. 3 Plant cycling is the process by which a plant is run up from cold to meet load and subsequently run down when not required. During the transitional periods plant is generally operating at far from optimal conditions and energy is required to prepare the plant for generation. All this contributes to poor overall efficiency. 4 Open cycle gas turbines provide cheap peaking capacity but are relatively expensive to operate. The impact of plant costs on their role in the system is explored in more detail in Chapter 7. 2
Slide 41: Features of Conventional and Renewable Generation 23 The design of generation plant does affect its time response and the extent that it can contribute to the regulation demands of the power system. Load following capabilities reflect in the main the thermal mass of the central plant elements. Thus coal stations with large boilers can take many hours to reach full output from cold. In contrast, gas turbines can reach rated power in minutes and combined cycle plants lie somewhere between coal fired and direct gas fired plant. Innovations in plant design aimed at improving conversion efficiency can have the effect of reducing flexibility. This important issue is discussed in detail in Reference [1]. Although the term nuclear power formally covers both fission and fusion, fusion power is still decades away and may never be technologically viable or cost effective, and it certainly cannot be regarded as conventional. Nevertheless, considerable international research has been and continues to be directed towards this technology, in particular through the recently agreed ITER consortium. In contrast fission based reactors have been generating a significant proportion of electricity worldwide. It remains a controversial technology due to the dangers of radiation and the challenge of radioactive waste disposal. A number of different technologies have been developed including boiling water reactors and gas cooled reactors. Such plant sustain a fission chain reaction within a controlled environment. Fission takes place in the reactor core which is normally contained within a pressure vessel safety shield. Moderators, usually graphite or water, slow down the neutrons and help regulate the chain reaction. Control rods are inserted into the core to regulate the nuclear reaction; these are made of a material such as boron which absorbs neutrons. The core also comprises fuel rods of fissile material. A coolant, normally water or gas, passes through the reactor. It then passes to the boiler where steam is raised. Electrical generation is provided by steam turbines and in this regard nuclear power stations are similar to large coal fired stations; it is simply the source of heat for raising the steam that is different. In general nuclear reactors are regarded as inflexible and they normally operate on base load. The fact that capital and installation costs far outweigh fuel and operational costs is further motivation for operating such plant as continuously as possible. Load factors are around 80% but would almost certainly fall if they were used to any extent to load follow, as advocates of nuclear power in France have claimed is possible. The activity of load following is sometimes confused with the steady reduction of nuclear power station output over some weeks, which is possible, indeed advisable, if the station is being run down prior to maintenance. 2.3 Hydroelectric Power As mentioned in Chapter 1, hydroelectric generation is an indirect form of solar energy. Incident solar radiation evaporates water from the sea, and to a lesser extent from land areas, and the warmed water vapour rises; as it ascends it expands and cools, eventually condensing in the form of clouds. Some of the resulting rain falls on high ground. This water has thus gained potential energy as a result of solar input. Hydro power is the result of extracting some of this energy as the water flows back towards the sea. Large scale hydro makes use of large reservoirs, usually created by damming rivers. Water is allowed to flow out of the reservoir in a controlled manner, turning turbines that drive electrical generators as it does so.
Slide 42: 24 Renewable Energy in Power Systems The storage of water in the reservoir allows generation to be timed to meet the demands of the power system. Energy storage capacity5 is limited so the aim is usually to generate at times of high load and so maximize the income generated. Since water availability is limited and seasonal, complex algorithms making use of rainfall prediction are used to optimize hydro operation. Some countries, for example Norway and Switzerland, have bountiful hydro resources and as a result there are times when electricity is so cheap it is almost given away. This has encouraged the development of industries that require abundant cheap electricity. Energy intensive processes such as aluminium smelting and silicon production are often located in countries with plentiful hydro resources for exactly this reason. Dams are attractive because they can provide a large head (equal to the fall in height of the water) but whether they can be built depends on the local topography. Rivers are the natural way in which water loses potential energy and it is possible to extract a proportion of this by means of so called low head schemes which are usually small scale. It is also possible to place turbines in a river flow directly with no dams or penstocks,6 extracting only a very small amount of the predominantly kinetic energy as the river flows by. Such turbines operate at effectively zero head and the installations are known as run of river schemes. 2.3.1 Large Hydro Large scale hydro is a well developed and widely used form of generation. Depending on estimates, between 20 and 25% of the world’s large scale hydro potential has already been developed, although the resource reasonably located geographically in relation to electricity demand has according to many commentators already been largely exploited. Hydroelectric stations currently contribute about 20% of world electricity generation. Large scale hydro power is operationally the most desirable of all renewable energy sources with respect to availability and flexibility of supply. As explained above, water can be stored in reservoirs and used when required, either continuously if the reservoir is large or when most required by the demand for electricity. The advantage of this storage arrangement is compounded by the natural capability of hydro plant to respond within minutes to demand increase or decrease. Such plant is therefore invaluable as a means of flexible generation to follow both predicted and unexpected changes in consumer demand. This feature is so valuable that in countries such as the UK where the topography is unsuitable for the installation of large conventional hydro, pumped storage schemes driven by the bidirectional transfer of water between two reservoirs have been developed. The downside of large hydro schemes is that they involve substantial upfront capital investments with profits accrued over long periods in the future. Their development can also be environmentally undesirable because of the flooding of large areas and the displacement of populations, such as has occurred with the construction of the Three Gorges scheme in China. However, this established renewable energy technology is unlikely to provide substantially increased contributions in the future since many of the attractive sites have already been developed; the areas with remaining potential lie mainly in the former Soviet Union and the 5 6 Capacity for a store is the energy that can be stored (i.e. MW h), rather than the rating. A penstock is a sluice or gate used to control the water together with the pipe to take the water to the turbine.
Slide 43: Features of Conventional and Renewable Generation 25 developing countries [2]. The optimal way of integrating this technology in power systems has been covered in many books and scientific papers published over the last century and will not be dealt with further here. Reference [3] presents an up to date approach to hydro scheduling. 2.3.2 Small Hydro A small-scale hydro is commonly defined as being smaller than 5 MW. At even lower powers (<100 kW), the so-called microhydro is subdivided into dammed and run of the river schemes in which there is no storage; a further distinction is into high and low head. It is estimated that the world potential of small/microhydro is around 500 GW of which roughly only onefifth has been exploited to date. A particular attraction of small/microhydro is that the resource is often located in remote rural upland areas unserved by a conventional electricity supply, but where the size of communities and their energy requirements are consistent with the available supply. The flow in a given river will vary greatly throughout the year, generally having, in the Northern hemisphere, high values during the winter months and low values during the summer months. In more tropical climates the flow is likely to relate to monsoon conditions. These seasonal constraints are important. For example, a correctly sized small scale hydropower scheme in the UK will not be expected to have sufficient water to run continuously throughout the summer months. For this reason most small hydro systems are grid connected. Figure 2.1 shows the mean daily flow for the river Barle in Somerset for 1980 from January through to December. The following observations can be made: • • • For two months in the summer the flow dropped to very low values, less than 1 m3/s. The river flooded (over 20 m3/s for this river) on several occasions. The river level tends to rise fairly quickly but this level reduces gradually; i.e. there is a sharp ‘leading edge’ followed by a slow decay. This is a general characteristic of rivers. Mean Daily Flow on River Barle (Somerset) Jan - Dec for 1980 35 30 25 Flow m3/s 20 15 10 5 0 Figure 2.1 The mean daily flow on the River Barle at Dulverton Somerset in 1980. (Provided by Robin Cotton for CREST MSc notes, Loughborough University, 1998)
Slide 44: 26 Renewable Energy in Power Systems Small scale hydro schemes show little variability in output from minute to minute, but can change substantially over hourly or daily cycles due to sudden rainfall. As already mentioned, the output is also highly sensitive to the time of year. If a large number of small scale hydro schemes are connected to an integrated electricity network their influence on the minute to minute operation of the power system would be negligible due to the statistical smoothing effect due to the aggregation of uncorrelated short term variations. Turbine Designs All hydo systems exploit the potential energy of the water. Engineers use the term head, H, for the height through which the water is allowed to fall; often in practice this becomes the net or effective head, reflecting frictional losses. Potential energy is simply mgH where g is the acceleration due to gravity; for a volume V of water density ρ this is VρgH. The power in kilowatts associated with a flow rate Q m3/s falling through an effective head of H metres is thus given by: P = Qρ gH Two different approaches to turbine design exist: impulse turbines like the Pelton wheel extract kinetic energy from the flow through impact on cups mounted on the turbine wheel, while reaction turbines like Francis and Kaplan designs, which in contrast to impulse turbines, operate submerged in water. Impulse turbines are suited to high pressure/head and can have efficiencies around 90%, whereas reaction turbines run faster and are suited to lower heads and have higher efficiencies. Turbine efficiency is conventionally plotted, as in Figure 2.2, as a function of the specific speed NS = n (P)H5/4, where P is the output power in kW, H is the effective head in metres and n is the rotational speed in revolutions per minute (rpm). The formula for the specific speed can be understood by considering a turbine scaled down to produce 1 kW of power at a head of 1 m for then its runner will rotate at the specific speed equal to its speed in rpm, i.e. NS = n. Figure 2.2 shows example efficiency curves for different turbine types. The speed of the runner blade relative to the water striking it is critical for the efficiency of the turbine. Propeller type turbines run best when their blade tips move faster than the water. In the case of the 96 94 Efficiency, % 92 90 88 86 84 82 80 0 Francis turbines 1 - jet Pelton turbines 76 152 228 304 380 456 Kaplan turbines 532 608 684 Figure 2.2 Efficiency curves for different turbine types
Slide 45: Features of Conventional and Renewable Generation 27 Kaplan, the blades should move at twice the water velocity, whereas the skirted Francis is most efficient when the two speeds are roughly equal. The Pelton theoretically performs best when its blades are moving at half the water velocity. Reaction turbines use guide vanes at the rotor inlet to adjust the direction of the inlet flow and the flow rate in order to vary the shaft power in response to the electrical load. The turbine runner has a series of vanes whose complicated geometry is designed to extract maximum power under design conditions. These vanes cannot be moved instantaneously, which has implications for the power quality achievable from such turbines. For standalone applications where fast control is essential, electronic control of the loads is used to maintain turbine speed. From Figure 2.2 it is clear that efficiency depends critically on specific speed, and this in turn depends on the effective head. Thus if the effective head varies significantly, the speed of the rotor needs to be adjusted to maintain high efficiency. Most turbines, however, operate at a fixed speed and efficiencies may thus vary to some extent. Small hydro systems can be grid-connected by driving an appropriately sized induction generator (see Chapter 4) provided that there is sufficient flow control. Flows should be controlled by the guide vanes to limit the output of the generator to the rated value. 2.4 Wind Power 2.4.1 The Resource Winds result from the large scale movements of air masses in the atmosphere. These movements of air are created on a global scale primarily by differential solar heating of the earth’s atmosphere. Therefore, wind energy, like hydro, is also an indirect form of solar energy. Air in the equatorial regions is heated more strongly than at other latitudes, causing it to become lighter and less dense. This warm air rises to high altitudes and then flows northward and southward towards the poles where the air near the surface is cooler. This movement ceases at about 30 °N and 30 °S, where the air begins to cool and sink and a return flow of this cooler air takes place in the lowest layers of the atmosphere. The areas of the globe where air is descending are zones of high pressure. Conversely where air is ascending, low pressure zones are formed. This horizontal pressure gradient drives the flow of air from high to low pressure, which determines the speed and initial direction of wind motion. The greater the pressure gradient, the greater is the force on the air and the higher is the wind speed. Since the direction of the force is from higher to lower pressure, the initial tendency of the wind is to flow perpendicular to the isobars (lines of equal pressure). However, as soon as wind motion is established, a deflective force is produced due to the rotation of the earth, which alters the direction of motion. This force is known as the Coriolis force. It is important in many of the world’s windy areas, but plays little role near to the equator. In addition to the main global wind systems there is also a variety of local effects. Differential heating of the sea and land also causes changes to the general flow. The nature of the terrain, ranging from mountains and valleys to more local obstacles such as buildings and trees, also has an important effect. The boundary layer refers to the lower region of the atmosphere where the wind speed is retarded by frictional forces on the earth’s surface. As a result wind speed increases with
Slide 46: 28 Renewable Energy in Power Systems height; this is true up to the height of the boundary layer, which is at approximately 1000 metres, but depends on atmospheric conditions. The change of wind speed with height is known as the wind shear. It is clear from this that the available resource depends on the hub height of the turbine. This has increased over recent years, reflecting the scaling-up of wind turbine technology, with the hub heights of the multimegawatt machines now being over 100 m. The European accessible onshore wind resource has been estimated at 4800 TW h/year taking into account typical wind turbine conversion efficiencies, with the European offshore resource in the region of 3000 TW h/year although this is highly dependent on the assumed allowable distance from shore. A recent report suggests that by 2030 the EU could be generating 965 TW h from onshore and offshore wind, amounting to 22.6% of electricity requirements [4]. The world onshore resource is approximately 53 000 TW h/year, taking into account siting constraints. To see these figures in context note that the UK annual electricity demand is in the region of 350 TW h and the USA demand is 3500 TW h. No figure is currently available for the world offshore resource, and this itself will be highly dependent on the allowable distance from shore. Of the new renewables wind power is the most developed. On very windy sites wind farms can produce energy at costs comparable to those of the most economic traditional generators. Due to advances in technology, the economies of scale, mass production and accumulated experience, over the next decade wind power is the renewable energy form likely to make the greatest contribution to electricity production. As a consequence, more work has been carried out on the integration of this resource than any of the other renewables and, naturally, this is reflected in the amount of attention given to wind power integration in this book. 2.4.2 Wind Variability The wind speed at a given location is continuously varying. There are changes in the annual mean wind speed from year to year (annual) changes with season (seasonal), with passing weather systems (synoptic), on a daily basis (diurnal) and from second to second (turbulence). All these changes, on their different timescales, can cause problems in predicting the overall energy capture from a site (annual and seasonal), and in ensuring that the variability of energy production does not adversely affect the local electricity network to which the wind turbine is connected. In Figure 2.3 each graph shows the wind speed over the time periods indicated. Wind speed measured continuously over 100 days is shown on the first graph followed by graphs, which in sequence zoom in on smaller and smaller windows of the series. It is easy to see the much larger relative variability in the longer time series (synoptic) as compared with the time series covering hours or less (diurnal, turbulence). This information is summarized in the spectral density presentation in Figure 2.4. In a spectral density function the height indicates the contribution to variation (strictly the variance) for the frequency indicated. A logarithmic scale as used here is the norm, and allows a very wide range of frequencies/timescales to be represented easily. The y axis is scaled by n to preserve the connection between areas under any part of the curve and the variance. The area under the entire curve is the total variance. It can be seen that the largest contribution to variation is the synoptic variation, confirming the interpretation of Figure 2.3. Fortunately these variations, characterized by durations of
Slide 47: Features of Conventional and Renewable Generation 29 Figure 2.3 Wind speed measured 30 m above flat terrain: vertical axis is wind speed, 0–20 m/s. (Reproduced with permission of Risø National Laboratory for Sustainable Energy) VanDerHoven 5 4.5 Normalised Variance. 4 3.5 3 2.5 2 1.5 1 0.5 0 -3 -2 -1 0 1 2 3 Log (Base 10) of time scale, cycles per hour VanDerHoven 4 days Semi-diurnal 5 mins 1 min. Figure 2.4 Power spectrum of wind speed variation
Slide 48: 30 Renewable Energy in Power Systems Frequency distribution of Wind Speed 0.06 0.05 Relative Frequency 0.04 0.03 0.02 0.01 0.00 typically 3 to 5 days, are slow in the context of the operation of large power systems. Apparently more difficult to deal with is the impact of short term variations due to wind turbulence, which are clear on the right hand side of Figure 2.4. However, as will be shown later, the aggregation effects will reduce this problem considerably. Fortuitously it is the timescales at which there is least variation, the so called spectral gap between 10 minutes and an hour or two, that pose the greatest challenge to power system operation. The essential characteristics of the long term variations of wind speed can also be usefully described by a frequency or probability distribution. Figure 2.5 shows the frequency distribution for a year of 10 minute means recorded at Rutherford Appleton Laboratory, Oxfordshire, UK. Its shape is typical of wind speeds across most of the world’s windier regions, with the modal value (the peak) located below the mean wind speed and a long tail reflecting the fact that most sites experience occasional very high winds associated with passing storms. A convenient mathematical distribution function that has been found to fit well with data, is the Weibull probability density function. This is expressed in terms of two parameters, k, a shape factor, and C, a scale factor that is closely related to the long term mean. These parameters are determined on the basis of a best fit to the wind speed data. A number of mathematical approaches of differing complexity are available to perform this fitting [5, 6]. 2.4.3 Wind Turbines The power in the wind than can be extracted by a wind turbine is proportional to the cube of the wind speed and is given in watts by: P= 1 ρ AU 3Cp 2 0.25 1.25 2.25 3.25 4.25 5.25 6.25 7.25 8.25 9.25 10.25 11.25 12.25 13.25 14.25 15.25 16.25 17.25 18.25 19.25 20.25 21.25 22.25 23.25 24.25 25.25 26.25 27.25 28.25 29.25 Wind Speed (m/s) Figure 2.5 Example wind speed frequency distribution
Slide 49: Features of Conventional and Renewable Generation 31 Figure 2.6 The Vestas V90, 3 MW wind turbine. (Reproduced with permission of Vestas Wind Systems A/S) where ρ is the air density, A is the rotor swept area, U is the wind speed and Cp is the power coefficient that represents the aerodynamic efficiency of the rotor. The variability in power output from one wind turbine would therefore be expected to substantially exaggerate the variability shown in the time histories of Figure 2.3. Wind turbines are designed to generate their rated or nameplate output at a rated wind speed Ur. For wind speeds below a cut-in wind speed Uco the wind turbine is not operational as the developed aerodynamic torque is not sufficient to overcome the frictional losses of the drivetrain and generate a useful power. For wind speeds above rated the power is controlled aerodynamically to maintain the output at the rated value until some limiting wind speed value is reached, known as the cut-out wind speed Uco at which point the turbine is shut down. The relationship between power and wind speed is known as a power curve. A power curve for a 3 MW wind turbine illustrated in Figure 2.6 is shown in Figure 2.7. For this machine Uci = 3.5 m/s, Ur = 15 m/s and Uco = 25 m/s, values which are typical of large modern turbines. This power characteristic combined with temporal variations in wind speed produces time varying electricity generation once the long term wind speed variations have been expressed in terms of a frequency or probability distribution of the sort shown in Figure 2.5. This can be combined with the power curve to indicate the probabilities of different power outputs,
Slide 50: 32 Renewable Energy in Power Systems 3,500 3,250 3,000 2,750 2,500 2,250 2,000 1,750 1,500 1,250 1,000 750 500 250 0 0 Power curve V90-3.0 MW Power (kw) 5 10 15 Wind speed (m/s) 20 25 Figure 2.7 Power curve for the Vestas V90, 3.0 MW turbine. (Reproduced with permission of Vestas Wind Systems A/S) and the overall average output of the turbine or wind farm. Details of this procedure can be found in Reference [7]. Average outputs lie typically in the range of 0.25–0.45 of the rated output depending on the mean wind speed at the site in question. These so called load factors or capacity factors are much lower than would be expected for conventional generators and are discussed in some detail in Chapter 3. The aerodynamic manner in which the wind turbine rotor extracts energy from the wind is described in a number of textbooks [8–10]. There is a well defined upper limit to the aerodynamic efficiency Cp of a rotor; this is known as the Betz limit and is approximately 0.59. It reflects the fact that the air is not forced to flow through the rotor (e.g. as in a ducted turbine) but can flow around it instead. Conventionally Cp, is plotted as a function of the tipspeed ratio λ (defined by λ = ΩR/U where Ω is the angular velocity of the rotor, R the radius and U the incident wind speed). The ratio λ is defined in this way to provide a generalized representation of the wind turbine rotor performance that is applicable to all combinations of incident wind speed and rotor rotational speed. A typical Cp–λ characteristic is shown in Figure 2.8. It is apparent that to operate at peak efficiency (for large modern wind turbines Cp is normally in the range 0.4–0.5), the tip speed ratio must be held constant for maximum output and this requires the rotor speed to be controlled in proportion to the wind speed. This is one reason why most larger modern wind turbines are designed to operate at variable speed (see Chapter 4 for a more complete discussion). This is attractive from an integration perspective as the rotor has inertia available to absorb or release energy when accelerating or decelerating respectively, thus smoothing short term variations in wind speed. Consequently its electrical power output varies less and can be more easily accommodated by the electrical system. The Cp–λ of Figure 2.8 assumes a fixed blade configuration. If the blade orientation is changed, then the efficiency will change too. In fact if the blades are pitched or feathered, i.e. rotated about their axis so as to reduce the angle between the blade chord and the resultant incident wind, then the lift forces on the blade that produce the torque will reduce. This will
Slide 51: Features of Conventional and Renewable Generation 33 0.5 0.4 0.3 Cp 0.2 0.1 0 0 5 λ 10 15 Figure 2.8 Cp–λ curve for a three-bladed rotor result in a Cp–λ curve that sits below the one shown in Figure 2.8. The more the blades are pitched away from the optimal position, the lower the output at any given combination of wind and rotor speed. Wind turbines that can control their output by pitching the blades in this manner are called pitch-controlled. An alternative is to allow the process of stall to limit the wind turbine output. Such stall regulated machines have a fixed blade pitch and run at a nominally fixed rotor speed so that wind speeds above rated result in stalling of the blades, thus limiting power output around the rated value. Stall regulation was widely used for the smaller turbines (50–500 kW), but this passive approach to power control lacks the potential for operators to adjust the turbine output in response to external conditions. In part as a consequence of this, practically all MW sized wind turbines are now pitch-controlled in some way. An approach known as active stall control combines pitch and stall regulation. Dynamic power control is by stall but the blades are slowly adjusted to make sure stall occurs at the correct power level. The mechanical shaft power created by the wind turbine rotor is converted to electricity by an electrical generator with a conversion efficiency that may reach 98% for large generators. Variable speed operation requires a frequency conversion through a power electronic converter, a process that reduces somewhat the overall efficiency. The generator and frequency conversion aspects are discussed in Chapter 4. 2.4.4 Power Variability The short, medium and long term variations in wind speed shown in Figure 2.3 affect power system operation in different ways. This is sufficiently important, given the significant and growing proportion of wind capacity in a number of power systems, to merit more detailed discussion. Variability from Second to Second At low/medium wind speeds the electrical output from a single wind turbine could vary substantially. In Chapter 5 it will be shown that this may have a detrimental effect on the power
Slide 52: 34 Renewable Energy in Power Systems system. However, when wind turbines are clustered in wind farms, there is physical spacing between them and the turbulence seen by each wind turbine is different and to a great extent uncorrelated. The electrical output from wind farms therefore exhibits substantially lower relative variability than that from a single wind turbine. At the planning stage, appropriate analytical studies are carried out to ensure that the variability expected from a wind farm at a particular site will not adversely affect the power system. Variability from Minute to Minute Figures 2.3 and 2.4 indicate that the character of wind is such that if the second to second turbulence is removed, the average wind speed from 10 minute period to 10 minute period remains effectively constant. In Chapter 3 will be shown that this ‘persistence’ nature of such averaged wind speeds is particularly important in integrating wind generated electricity in power networks. In practice the output of turbines can be regarded as uncorrelated on the timescale of minutes and as with the faster variations the affect of aggregation is to smooth out variations at these higher frequencies. Variability from Hour to Hour and from Day to Day Figure 2.9 shows the actual records of wind speed at 13 geographically dispersed wind sites in the UK. As expected from Figure 2.4, there is substantial variability at each location over Figure 2.9 UK site specific and average hourly wind speed over 72 hours, with 50 or more records per hour. (Reproduced from Sinden, G.E., 2007, DPhil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment)
Slide 53: Features of Conventional and Renewable Generation 35 1.00 0.90 Wind Speed Correlation-r 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0 200 400 600 800 1000 1200 Distance Between Recording Sites-km Figure 2.10 UK onshore wind power correlation by distance between sites based on UK long term averages. (Reproduced from Sinden, G.E., 2007, DPhil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment) the 72 hour recorded period. Variability now becomes significant and effective integration of wind power in an electrical power network would have been problematic if the total resource were to fluctuate in sympathy with the fluctuations in one site. Fortunately this is not the case due to the geographic diversity of the sites. The benefits of geographic diversity are clearly illustrated by the bold line in Figure 2.9, which represents the average wind speed from all sites. This average is substantially smoother (i.e. exhibits much reduced variability) than the wind at the individual wind farm sites. With large scale exploitation of the wind resource, wind farms are installed inevitably across a variety of geographically dispersed sites. As suggested above, this has a major beneficial effect in terms of overall variability. Just as the output from a wind farm has less short term variability than a single wind turbine due to their dispersion across the site, so the aggregate output from several geographically dispersed wind farms has less longer term variability than the output from a single wind farm. This reflects the fact that distant localities experience variations in wind due to shifting weather patterns that are time shifted in relation to one another, and also to an extent distinct. Figure 2.10 presents the correlation between pairs of onshore wind sites in the UK as a function of the distance between the sites, and demonstrates that sites very far apart exhibit low cross-correlation. The data from Reference [11] were recorded over a period of 15–20 years. Seasonal Variability Seasonal and monthly average wind speeds vary significantly over most of the world. Figure 2.11 shows the seasonal changes of monthly averaged wind speeds for Billings, Montana in the USA. The trend of higher wind speeds during winter compared to summer is typical of the Northern hemisphere. The figure also indicates that there is variability from year to year.
Slide 54: 36 Renewable Energy in Power Systems 8 Average wind speed, m/s 7 6 5 4 3 18.5 year average 1965 1966 1967 J F M A M J J Month A S O N D Figure 2.11 Seasonal changes of monthly average wind speeds. (Reproduced from Reference [10] with permission of John Wiley & Sons, Inc.) 2.5 PV and Solar Thermal Electricity 2.5.1 The Resource The average intensity of light outside the atmosphere (known as the solar constant) is near to 1353 W/m2. Attenuation by the atmosphere results in peak intensity at sea level of around 1 kW/m2, giving a 24 hour annual average of 0.2 kW/m2 averaged over the planet’s surface. As this overall energy density is relatively small, large areas will be needed for a significant energy production. For example, in order to produce a gigawatt of power, an area of nearly 5 km2 would be needed (assuming a conversion efficiency of 20%). However, in countries such as the UK (with low irradiance7 and high population density) the existing and appropriate roof space is sufficient in principle to generate enough electricity to cover a significant proportion of electricity consumption. Common sense tells us that irradiation varies regionally, with the changing seasons, and hourly with the daily variation of the sun’s elevation. Many locations do not experience unbroken sunshine. Cloud cover can significantly reduce the net radiation and cause relatively fast variations in intensity, in some cases significant variations from minute to minute or even over seconds. Seasonal variation on the earth’s surface is illustrated here by comparing two different sites. The magnitude will obviously vary from site to site but the basic principles are the same. Sites of similar latitudes should, in principal, have a similar solar resource. The earth’s tilt angle leads to a variation of the seasonal irradiance. Figure 2.12 illustrates the differences in average monthly irradiation for a site close to the equator, Kisangani (Congo, latitude –0.31 °) and a site at higher latitude, Sutton Bonnington (UK, latitude 52.5 °). As expected, Kisangani receives twice as much irradiation in the course of the year. While 7 Sometimes the terms solar radiance, radiation intensity and insolation (in older text books) can be found although irradiance is the most widely accepted. The SI units for irradiance are W/m2.
Slide 55: Features of Conventional and Renewable Generation 37 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Kisangani beam Kisangani diffuse Sutton bonnington, beam Sutton bonnington, diffuse Irradiation [W/m2] J F M A M J Month J A S O N D Figure 2.12 Comparison of Irradiation at Kisangani (Congo) and Sutton Bonnington (UK) Sutton Bonnington experiences a strong seasonal variation of ±73%, the irradiation in Kisangani varies only by ±13% around the yearly average. The annual average irradiation peaks in the north-east of Africa at around 300 W/m2. Other favoured regions can be found all over Africa, the north of South America and the south of North America, and South-West Asia. Australia experiences solar irradiation levels above 250 W/m2 across nearly half the continent. The irradiation reduces, generally speaking, the further the distance away from the equator. It can be well below 100 W/m2 on a 24 hour average. Obviously, these values are averaged over the year and will vary significantly with the seasons. Most northern European, North American and north Asian countries are within this latter region and present and future systems are and likely to be rooftop installations in the kilowatt scale connected to the local 230 or 115 V network. 2.5.2 The Technology There are two main technologies for the conversion of sunlight into electricity. Photovoltaic (PV) cells depend on the use of semiconductor devices for the direct conversion of the solar radiation into electrical energy. Efficiencies of the typical commercial crystalline PV cells are in the range 12–18% although experimental cells have been constructed that are capable of over 30%. In contrast, solar thermal systems depend on intermediate conversion of solar energy into thermal energy in the form of steam, which in turn is used to drive a turbogenerator. To obtain high temperatures, thermal systems invariably use concentrators either in the form of parabolic troughs or thermal towers. At present, generation of electricity by either technology is substantially more expensive than traditional means. Due to the considerable potential of cost reductions in PV systems it is believed that in the future, perhaps in a decade
Slide 56: 38 Renewable Energy in Power Systems or so from the time of writing, PV systems will be providing a sizeable proportion of the renewable energy contribution. Some countries, notably Japan and Germany, have created substantial home based markets for PV cells, and the last couple of years have seen a rapid growth in multimegawatt installations in Europe. Despite this encouraging trend, the relatively high present cost of PV systems are likely to limit them to modest contributions to overall electricity supply in the immediate future. In sunnier places such as California, Australia, North Africa and the Mediterranean, where the peak electricity demand occurs in summer due to tourism and air-conditioning loads, large scale multimegawatt PV plants or solar thermal plants are more attractive. Nevertheless, significant technology and cost breakthroughs in these two solar technologies will be needed if they are to make sizeable contributions to electricity supply. The characteristics of PV and solar thermal plants are rather different and are next outlined. 2.5.3 Photovoltaic Systems At the heart of a PV system is the PV module. Detailed descriptions of the different PV technologies and the basics of solar cell operation can be found in a wide range of textbooks, for example References [13] and [14]. PV modules produce output determined mainly by the level of incident radiation. They are characterized for given external conditions, by an I–V curve of the type shown in Figure 2.13. The power, IV, depends on the operating point and is maximized for operation near to the knee of the I–V characteristic, known as the maximum power point (MPP). Power electronics is used to convert the DC (direct current) output of the PV modules to AC (alternating current) for injection into the network (more about this in Chapter 3). The quality of a cell can be judged by the squareness of the I–V characteristic. This is quantified in terms of the ratio of the voltage at open circuit (i.e. where the I–V curve meets the voltage axis) times the closed circuit current (i.e. where the I–V curve meets the current axis), divided by the power at the MPP. This ratio is known as the fill factor. 3 2.5 Current (Amperes) 2 1.5 1 0.5 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Voltage (Volts) MPP Figure 2.13 An example I–V curve
Slide 57: Features of Conventional and Renewable Generation 39 300 1000W.m2, 60°C 1000W.m–2, 25°C Current density (A.m–2) 200 100 200W.m–2, 25°C 0 0 Operating voltage 0.2 0.4 Voltage (V) 0.6 0.8 Figure 2.14 Impact of radiation and temperature on the I–V characteristic 1200 1000 Irradiance, W/m2 800 600 400 200 0 0 2 4 6 8 10 12 14 Time, h 16 18 20 22 24 Figure 2.15 Time variation of radiation through a summer ’s day in Loughborough, UK As shown in Figure 2.14, radiation and temperature affect the shape of the I–V curve and the voltage at which the MPP occurs. The power electronic converter is also used to control the operating voltage as near as possible to the MPP and track this as it changes with radiation level and to a lesser extent with module temperature. Commonly both of these power electronic functions are combined in the inverters used for grid connection. PV modules have negligible energy storage capability, and as far as the power system is concerned their output varies instantaneously with changes in radiation. Hence a passing cloud, as occurs in Figure 2.15 just before 18.00, will result in the sudden collapse in output from a PV system. As with wind turbines, a geographical spread of PV systems will substantially mitigate such short term effects. It has been speculated that as power systems become increasingly interconnected, as is currently happening in Europe, it may eventually be possible to smooth out the diurnal
Slide 58: 40 Renewable Energy in Power Systems Figure 2.16 50 MW parabolic trough plant with thermal storage at Andasol, Spain. (Reproduced from Concentrating Solar Power – from research to implementation. © European Communities, 2007) variations in PV output.8 In such a vision significant intercontinental power flows might occur; for example, power generated from massive PV arrays in the Sahara during hours of daylight could be transmitted to areas of the planet then in darkness. Of course this would involve power systems spanning substantial proportions of the planet, something that may never turn out to be cost effective. 2.5.4 Solar Thermal Electric Systems Solar thermal electricity generation systems most commonly use solar concentrators to produce high temperatures that can drive heat engines with acceptable conversion efficiency. Two main types of generator suitable for large scale generation have so far been demonstrated, both requiring direct or beam radiation. Climates with cloudy or overcast conditions, where most of the radiation is diffuse, are thus unsuited to this technology. The two technologies are: • Large collections (or solar farms) of parabolic trough reflectors, Figure 2.16, focus solar radiation on to a line receiver containing a piped heat transfer medium. The medium can be a thermal oil, capable of withstanding high temperatures, or alternatively pressurized water. This is collected and passed through a heat exchanger/boiler where steam is raised for supply to the turbines. Operational temperatures vary between 350 and 400 °C and system sizes up to 80 MW have been built. To increase steam turbine operating 8 The benefits of interconnection are central to the operation of power systems and will be returned to again and again throughout this book.
Slide 59: Features of Conventional and Renewable Generation 41 Figure 2.17 Aerial view of 11 MW PS10 solar tower plant at Sanlúcar la Mayor, Spain. (Reproduced from Concentrating Solar Power – from research to implementation. © European Communities, 2007) • temperature, and thus efficiency, steam from the solar system can be further heated using conventional fuels. In solar power towers a central receiver is mounted on top of a tower which is surrounded by a field of concentrating mirrors (heliostats) which track the sun. The heliostats reflect and concentrate the light on to a receiver where the energy is absorbed by the heat transfer medium, which could be water, a molten salt or any other suitable high temperature heat transfer liquid. Temperatures as high as 1000 °C can be achieved. System sizes up to 200 MW have been considered, but to date the largest systems constructed are considerably less. Figure 2.17 shows one of the larger systems currently operating, the 11 MW PS10 solar tower plant at Sanlúcar la Mayor, Spain. Solar power towers are expected to be more economic than solar farms for larger sized plant, say over 100 MW. Since both these technologies depend on direct sunlight for efficient operation they function for only part of the time, and then not always at maximum output. However, because they both involve a thermal intermediary stage, they can be combined with fossil fuel combustion (a hybrid) and depending on design details it may be possible to incorporate thermal storage.
Slide 60: 42 Renewable Energy in Power Systems Table 2.1 Performance of parabolic trough and power tower systems. (Source: http://www.eere. energy.gov/consumerinfo/pdfs/solar_overview.pdf) Parabolic trough Size Operating temperature (°C/°F) Annual capacity factor Peak efficiency Net annual efficiency Status Storage available Hybrid designs a Power tower 10–200 MW 565/1049 20–77%a 23%(predicted) 7(demonstrated)–20% Demonstration Yes Yes 30–320 MW 390/734 23–50%a 20%(demonstrated) 11(demonstrated)–16% Prototype Limited Yes The top end capacity factors are for systems including substantial thermal storage. The primary advantage of these adaptations is to the capability to supply power when the sun does not shine, in other words to be dispatchable. A smaller scale approach, suitable for units in the range 10–50 kW, is possible using parabolic concentrators. These also work at high temperatures of 600–1000 °C and might be suitable for distributed electricity generation. Thermal-to-electric efficiency is in the range of 20–40% depending on the design, with the resulting overall solar-to-electricity conversion efficiency in the range 13–25%. Table 2.1 (adapted from material available from the website given in Reference [14] when it was accessed in January 2006) compares the performance of the parabolic trough and power tower systems. 2.6 Tidal Power 2.6.1 The Resource The moon and sun’s gravitational fields cause the natural rise and fall of coastal tidal waters. Since the moon is closer to the earth, albeit much less massive, it has a dominant effect upon tides. As the moon is 2.2 times more influential than the sun, it could be considered that tidal energy is mostly a form of lunar energy! The earth rotates on its axis once every 24 hours. In the earth’s frame of reference the sun orbits the earth once every 24 hours. The moon orbits the earth once every 29 days approximately. In the earth’s frame of reference, the moon appears to orbit the earth once every 24 hours and 50 minutes. This difference in periods between the apparent orbits of the sun and moon leads to phase changes with larger spring tides during in-phase behaviour and smaller neap tides when the sun and moon are out of phase. Spring tides can be twice as large as neap tides. A 1991 study commissioned by the EU estimated that the technically feasible energy resources from tidal barrages across the EU could be as much as 105 TW h/year (from 64 GW of installed capacity). This resource is unevenly distributed across Europe with the UK
Slide 61: Features of Conventional and Renewable Generation 43 (47.7%) and France (42.1%) sharing the bulk of the resource and Eire (7.6%) accounting for most of the rest. It has been estimated that exploitation of all practicable estuaries in the UK could lead to electricity generation of up to around 20% of demand in England and Wales. 2.6.2 Tidal Enhancement Over much of the surface of the oceans, the tidal range (the vertical rise and fall) is rather small, less than one metre, but in certain places, there is an enhancement of the range. Enhancement may be due to the following. Funnelling The tide is gradually constrained from the sides and so increases in height – or the reverse, later in the cycle. Resonance The estuary has a resonant period equivalent to the tidal period. The length and depth of the estuary are very important for resonance. Coriolis Effect The spinning of the earth leads to the Coriolis effect mentioned earlier with respect to wind. The tide is influenced by this and in some locations tends to increase in height at high tide and be drawn away from the coast at low tide, with the net effect of enhancing the tidal range. These enhancements are highly predictable, allowing the output from a tidal scheme to be determined many years ahead. 2.6.3 Tidal Barrages Constructing a barrage across an estuary and allowing tidal waters alternately to fill the estuary through sluice gates and then to empty it through turbines can generate energy. A barrage constructed across an estuary is equipped with a series of gated sluices and a bank of low head axial turbines. Where it is necessary to maintain navigation to the upper part of the estuary, a ship-lock may be required. Tidal barrages are a currently available technology, but very few exist worldwide. The best known example is the 240 MW scheme at La Rance in France, and smaller installations have been made in Nova Scotia, Russia and China. The UK has a number of attractive sites due to its high tidal ranges, the largest potentials being on the Severn estuary (8600 MW capacity) and the Mersey (700 MW capacity). However, the scale of these installations and the calculated long payback periods make the required investments unlikely in the context of the
Slide 62: 44 Renewable Energy in Power Systems privatized electricity supply industry. Furthermore, environmental considerations may present a barrier to large scale developments. 2.6.4 Operational Strategies Ebb generation is the simplest mode of operation for a tidal barrage scheme. The operating cycle consists of four steps: • • • • Sluicing on the flood tide, to fill the basin. Holding the impounded water until the receding tide creates a suitable head. Releasing the water from the basin to the sea via turbines, on the ebb tide, until the tide turns and rises to reduce the head to the minimum operating point. Holding until the tide rises sufficiently to repeat the first step. Ebb generation with flood pumping is a modification of this mode which allows increased energy output. By using the turbines in reverse as pumps, the basin level and hence the generating head can be raised. The energy required for pumping must be imported but since the pumping is carried out against a small head at high tide and the same water is released later through the turbine at a greater head, this can produce a net energy gain with some limited ability to re-time output. The energy gain through pumping could be small but useful and typically in the range 3–13%. Flood generation is the reverse of ebb generation and is rarely suggested alone, possibly because it offers little storage opportunity as the basin is often being filled by a river, which reduces the total energy capacity. Two-way generation (ebb and flood) is possible with reversible turbines and is used at La Rance (together with flood pumping). The additional energy recovered may not justify the extra cost and complexity of the turbines. Most schemes propose ebb-only generation. The full resource could only be captured if the barrage of the scheme has a very large two-way generating flow capacity in order to achieve rapid filling and emptying of the basin. In practice, it is neither economic nor is there enough physical space to install enough turbines to come anywhere near full resource capture. Typically only about one-third of the gross power can be harnessed, so after turbine and generator losses the mean electrical output will at best be about a quarter of the gross power. The other main reason for preferring ebb generation is the nature of the amenity that results. In such a scheme, the water level in the basin never falls below the mean sea level, so the basin is available for recreational activity much like an inland reservoir. On the other hand, it is bad news for the bird population for whom exposed mudflats at low tide provide a major food resource. High initial capital cost and polarized opinions on the resulting environmental changes are the main reasons why tidal schemes seldom get beyond the feasibility study stage. Tidal energy barrages are expected to have very long lifetimes. Their design life could be about 120 years, but with normal maintenance and replacement of turbine generators at 40 year intervals, their lifetime could effectively be unlimited.
Slide 63: Features of Conventional and Renewable Generation Water lever/m 45 5 Water level in basin –5 7 6 5 4 3 2 1 1 2 3 4 Time/days 5 6 7 Figure 2.18 Water level and electrical power output of the proposed Severn Barrage over a spring– neap tide cycle. (Source: the Watt Committee on Energy now disbanded. Reproduced with permission of Oxford University Press) Power Variability One of the main disadvantages of tidal schemes is the pulsed nature of their electrical output. Figure 2.18 shows the expected electrical output from the 7 GW scheme proposed for the Severn estuary assuming ebb generation only. Power is generated for five to six hours during spring tides and for three hours during neap tides. This pattern may, but is unlikely to, match the pattern of demand in an integrated power system. On the positive side the power availability from such a scheme is highly predictable. In a large integrated system such as the UK’s this predictability will ease the task of scheduling alternative generation to take up the demand when the output from the scheme falls to zero. 2.6.5 Tidal Current Schemes The direct production of electricity from tidal streams is a relatively new concept. It relies on a different approach to conventional tidal barrage schemes and is attractive in that it does not require such massive infrastructural investment. Tidal stream technology extracts energy directly from the currents that flow in certain locations, driven by the rise and fall of tides in the vicinity. These currents usually have a low velocity (1 m/s), though this can be enhanced by the local topography. In particular, the velocity can be magnified greatly in straits between islands or between islands and the mainland. Tides can be predicted with very high accuracy; hence after measurements at a site, the energy available for conversion can be forecast with confidence. Figure 2.19 shows a tidal cycle together with the associated stream current velocity. Tidal height varies approximately sinusoidally and, as already mentioned, is more or less completely predictable. For any site the tidal current velocity has four peaks and four troughs per day, as in the figure. Power/GW
Slide 64: 46 Renewable Energy in Power Systems Tide height and current velocity - South Lundy Island 3.0 2.5 2.0 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Tide height (m) 1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 -2.0 Tide height Current velocity Time (24 hours) Figure 2.19 Tide height and current velocity. (Reproduced from Sinden, G.E., 2007, DPhil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment) The technology to extract energy from the tidal current is conceptually simple: a turbine is placed in a suitable tidal flow, which turns the generator through a gearbox as in Figure 2.20. It is similar to a submerged wind turbine, except that the greater specific gravity of seawater results in much higher energy densities in tidal streams than is found in winds of the same velocity. However, the water velocities available in tidal streams (typically rated velocities of 2–3 m/s on good sites) are much lower than the air velocities used by wind turbines. Although power output is proportional to the cube of the velocity, tidal stream rotors generally produce significantly greater output than wind turbines of the same size because of the massively increased water density. Compared to wind, tidal flow velocities are expected to have little turbulence and thus vary in a smooth manner, reducing fatigue loads on the rotor and generating electricity with little short term variation. At different sites, the peaks and troughs occur at different times. Power being proportional to the cube of current velocity, this implies that maximum and minimum electricity generation is possible at different times. By combining generation from sites that are out of phase it is theoretically possible to smooth out the diurnal variability. The tidal stream resource is highly site-specific, and there is a limited number of tidal stream sites around the world worth exploring. The accessible tidal stream resource for the most suitable sites in the UK (including the Channel Islands) is estimated to be approximately 36 TW h/year. The power variability from such schemes will be similar to that from barrages. A second type of variability also affects tidal velocities. This is the spring–neap cycle, which occurs over 14 days and changes tide ranges (the difference in height between high and low tides) from a minimum to a maximum, then back to a minimum. Unlike diurnal variations, the spring–neap cycle affects all sites in the same way at the same time, so there is no scope for smoothing by combining generation from many sites. It is concluded that patterns of tidal stream power output differ significantly from electricity demand cycles; maximum potential output sometimes coincides with peak demand but at other times minimum output does. However at the likely levels of penetration Current velocity (ms-1)
Slide 65: Features of Conventional and Renewable Generation 47 Figure 2.20 Artist’s impression of axial flow tidal current turbines. (Reproduced with permission of Marine Current Turbines Limited) from this resource, the magnitude of this variability is small compared to demand variations. 2.7 Wave Power 2.7.1 The Resource The passage of wind over the surface of the sea results in the gradual transfer of energy into the water to produce waves, so wave energy is also an indirect form of solar energy. Wind power typically has densities in the range 1.2–1.8 kW/m2. Waves with a typical power density of 50 kW per metre of wave front or crest length are in effect a highly concentrated form of solar energy. The distance over which this process of wave generation occurs is called the ‘fetch’ and longer fetches produce larger, more powerful waves as do stronger winds and extended periods of wind. Sea waves are characterized by their wave height (H), period (T) and crest length. The power per metre of crest length is proportional to the period and to the square of the wave height. Wave records indicate that the heights and periods of ocean waves vary continuously over time, which results in wave power varying almost continuously. However, the conversion of the energy of a wave face into electricity output tends to partly smooth this variability. Because of these irregularities the calculation of the power density requires an averaging process. The significant wave height HS is approximately equal to the average of the highest
Slide 66: 48 Renewable Energy in Power Systems Table 2.2 Technically achievable wave power resource UK potential (GW) Shoreline Near shore Offshore 0.03 0.3–0.7 7–10 World potential (GW) 1–50 10–500 200–5000 one-third of the waves and the zero up-crossing period Te is defined as the average time counted over ten crossings or more between upward movements of the surface through the mean level. It can be shown that the average power in one metre of wave crest is proportional to Te and to the square of HS. The world wave resource is not yet fully analysed but there is no doubt that it is sizeable. Table 2.2 gives an estimate of the scale of the technically achievable resource. 2.7.2 The Technology The conversion of wave power into electricity requires a device that intercepts the waves and converts a proportion of their energy first into mechanical and then into electrical form. The conversion of wave energy into mechanical energy demands a central stable structure incorporating an active element which moves relative to it under the forces exerted by the waves, and can react against the central structure to produce forces and displacements that generate mechanical power. Wave energy research and development has been taking place for over 25 years now, and significant progress has been made towards the development of viable technologies able to exploit the large energy potential of the world ocean wave climates. Ocean waves are often powerful, but with extremely low frequencies, of about 0.1 Hz (equivalent to 6 rpm or to periods of around 10 s), and the success in generating electricity demands that this frequency is raised to 500–1500 rpm. The technology for wave energy conversion is still at an early stage compared to wind and photovoltaics and a variety of different approaches are currently under development. Because of the nature of the resource and for efficient conversion, the swept volume of the device must be of the order of several tens of cubic metres per metre of device width. The devices are therefore physically large and have to be designed to withstand without damage extreme waves that may occur very rarely. The transition of a concept from a model tested in a laboratory wave tank to a working prototype requires considerable expenditure and it is only very recently that serious funding from governments and industry has been forthcoming. It is not the intention here to describe in detail the various concepts and associated hardware. The devices can be classified in a variety of ways depending on their intended location or their geometry and orientation. At least twelve different devices are being developed worldwide and it is still too early to guess which of them will be capable of providing energy competitively and reliably. As an example of the ingenuity of recent developments, Figure 2.21 shows an artist’s impression of a Pelamis array. The structure of the Pelamis comprises four semi-submerged cylinders linked by hinged joints. The relative motion between the
Slide 67: Features of Conventional and Renewable Generation 49 Figure 2.21 Artist’s impression of Pelamis array. (Reproduced with permission of Pelamis Wave Power) cylinders is resisted by hydraulic rams that pump oil into high pressure tanks which then is used to drive an electrical generator. While the UK is fortunate in having a good wave climate, the political climate has not always favoured the technology. Attitudes, however, are changing, prompted by the need to address global climate change, by the long term resource security of fossil fuels and by the increasingly competitive economics of wave energy. UK R&D teams conducted much of the early work, but many other countries are now active in wave energy development. Commercial involvement is now significant and schemes or concepts from the Netherlands, Norway, Australia, Sweden, Denmark, the USA as well as the UK are now being developed. 2.7.3 Variability The power output from a single wave power device will follow an exaggerated variation of the wave height trains because of the square relationship between power and wave height. Not unlike offshore wind farms, a substantial wave power installation will consist of an array of such devices, the outputs from which when added together, because of spatial effects, will be relatively smoother than that from one device. If this were extended to hundreds of such devices in geographically dispersed arrays, the overall output will be smoother still. The short term variability of the output from wave power converters is unlikely to present integration problems, assuming that the outputs from arrays are connected to a grid voltage level capable of absorbing variable generated power without adverse network effects.
Slide 68: 50 Renewable Energy in Power Systems Average montly capacity factor for wave power 50% 45% Average monthly capacity factor 40% 35% 30% 25% 20% 15% 10% 5% 0% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Monthly average wave power Average annual wave power Monthly average wind power Month Figure 2.22 Average monthly wave power in the UK. Average monthly wind power capacity factor [15] is shown for reference purposes. (Reproduced from Sinden, G.E., 2007, DPhil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment) The wave resource, not unlike the wind resource on which it depends, also varies on a day-to-day and season-by-season basis; in general wave conditions are more energetic in the winter than in the summer. For example, about half of the annual wave power at all UK sites occurs during the winter months of December, January and February, as shown in Figure 2.22. This exceptional match between availability and demand for the UK would also hold for many countries with coastal areas facing the Atlantic or Pacific in the northern hemisphere. Figure 2.22 also depicts the seasonal variability of wind to illustrate, by comparison, the pronounced variability of wave power. Wave power levels [16] are predictable to varying degrees over different timescales. At short forecast periods (a few hours), information on the wave conditions offshore, coupled with the power available from wave devices in the previous hour, can be used to estimate future output. For longer forecast periods, wave models provide estimates of wave power up to 5 days ahead. 2.8 Biomass 2.8.1 The Resource Biomass differs considerably from the other renewable sources in that it takes the form of a fuel that can be stored and used for electricity generation when required, in the same way as fossil fuels. However, unlike fossil fuels, biomass is often limited by the energy density of
Slide 69: Features of Conventional and Renewable Generation 51 the stored fuel. Therefore, it must be produced and consumed locally, as energy consumption associated with transportation over long distances might even exceed that of the fuel itself. This means that biomass power generating units are relatively small compared to conventional plant, (relying on local supply chains for feedstock) and possess the characteristics of small embedded generating units. There are three basic thermochemical conversion technologies that use solid biomass as a primary fuel for the production of electricity, namely direct combustion, gasification and pyrolysis. In addition, the use of liquid biomass (such as sewage sludge) for the production of methane via anaerobic digestion is increasingly common. Electricity production using solid biomass fuels is still a developing industry and as a consequence is not competitive on price with electricity from fossil fuels without some kind of government fiscal or policy support. However, it is competitive with nuclear power and possibly new-build clean coal power stations, but not with modern gas fired power stations within the current regulatory and economic climate. With the correct support, as currently in the UK, co-firing of coal with biomass is commercially attractive. In the longer term, gridconnected biomass generation (using the full range of possible technologies) may become competitive; the greatest potential is for small scale embedded generation using gasification, pyrolysis or high speed steam engine based plant. In the short term, small scale (100–500 kWe) dedicated plants for use on farms or by rural industry has the greatest potential. In the medium term when increased demand for electricity could be causing the grid to become overloaded and unreliable, then larger (1–20 MWe) embedded biomass generation plant providing endof-grid support may become an attractive alternative to reinforcing the grid. 2.8.2 Resource Sustainability Biomass has become less important as countries have industrialized and now accounts for less than 3% of energy in the developed world. In contrast developing countries remain highly reliant on wood and other natural biomass with over 30% of their energy needs being supplied from these sources. Of course with growing populations this supply is not sustainable. Ironically, the industrialized countries need to make increased use of biomass, and the less developed regions limit their use of this resource to a sustainable level. One of the key features of biomass is that the energy expended in growing it, i.e. planting, watering, use of chemicals and pesticides to enhance yield, harvesting, drying etc., is not negligible. Specifically for ethanol production used as fuel for transport, the refineries themselves are fired by fossil fuels to ferment the crop and to purify ethanol from the product of fermentation. A US Department of Agriculture report found that the energy from corn bioethanol was a mere 8% in excess of the input production energy and a recent paper in Science [15] found that the energy ratio was net-positive when the energy savings from ‘co-products’ for cattle feed were included. Efforts are now being made to produce bioethanol from cellulosic crops and not from fermentation. This promises to produce twice the amount of ethanol per hectare of crop. The scene is a lot brighter if biomass is used to generate electricity, especially in CHP plants in small decentralized power stations. The benefits are compounded if the crops are grown organically, if possible, and used locally. The choice of crop is also vital in the effectiveness of CO2 mitigation. Table 2.3 compares the ratio of energy out to energy in for a
Slide 70: 52 Renewable Energy in Power Systems Table 2.3 Crop Energy ratio of several crops Ratio 32.5 30.0 8.5 8.8 3.8 Miscanthus Willow Hemp (straw) Wheat (grain) Oilseed rape number of crops based on DTI report URN 01/797, [16]. Miscanthus and willow are the preferred crops in European and other similar latitudes. Currently, the main use of miscanthus in the UK is in co-firing with coal in existing power stations. Large scale biofuel production is not only energy intensive but it could have an adverse environmental and social impact. Such development requires substantial water resources with the result that water tables in areas of intense cultivation have been lowered to unacceptable levels. Expansion of biofuel crops could also speed up tropical deforestation with the associated lowering of CO2 absorption and threat to extinction of thousands of species of animals and plants. If such crops are encouraged through subsidies, food shortages may occur if land previously used to produce food is lost. It could be concluded that the biomass path, unless used wisely, may cause serious environmental impacts. 2.9 Summary of Power Generation Characteristics This chapter has reviewed the main forms of electricity generation. For the renewable energy forms that are the main topic of this book, the resource characteristics have been presented and discussed in some detail since they impact significantly on the manner in which these sources can be integrated into the power system. This chapter concludes by summarizing the key characteristics of the different forms of generation in Table 2.4. In this table typical unit sizes for the different forms of electricity generation are referred to. These unit sizes are in all cases the nameplate rating of the prime mover/generator package. Power stations may contain one or several such units. The table also describes the nature of the energy resource in terms of availability. Traditional generators supplied from coal, oil, gas or nuclear energy score highly here as the primary energy resource is continuously available. However, their capability to be dispatched, which requires that their output can be changed automatically or at will by control engineers to follow the demand variations, depends crucially on their thermal/mechanical nature and differs from plant to plant as described in Chapter 3. One characteristic shared by all the renewable sources, excluding biomass and tidal, is their variability and relative unpredictability. This presents a challenge in integrating such sources in electrical power networks that have been designed to operate with traditional generators whose availability appears certain. In reality, however, no plant can be completely available; there is always some probability of breakdown, the need for maintenance, etc. This topic is complex and often misunderstood. It is dealt with in some detail in Chapter 3.
Slide 71: Features of Conventional and Renewable Generation 53 Table 2.4 Generator characteristics by energy source Typical unit size 500 MW 500 MW Up to 500 MW 100 MW Up to 500 MW Up to 500 MW Up to 100 MW Up to 40 MW (at present in UK), in future larger units Up to 5 MW 1 MW kW (in UK) 1 kW, domestic up to 100 kW commercial No commercial examples yet No recent Variablea No No No No No Yes Usually No Predictable Yes Yes Yes Yes Yes Yes Usually Yes Dispatchable Yes No Yes Yes Yes Yes No, because it is heat led Yes Energy source Coal Nuclear Gas CCGT Gas open cycle Hydro with reservoir Pumped storage hydro CHP Energy crops and municipal solid waste Wind Landfill gas Run-of-river hydro Photovoltaic cell Yes No Yes Yes Not accurately Yes Not accurately Not accurately No Yes No No Wave Yes Tidal a Yes Not accurately over long term Yes No No Whether output varies in time. 2.10 Combining Sources Many power systems at present make use of renewable sources, most commonly hydro and wind power. For now it is sufficient to note that, as mentioned earlier, geographical diversity of the wind resources assists integration. Geographical diversity also applies to the other renewable resources and is important to their integration into power systems. The relationship between the different renewable sources, spatially and temporally, is also directly relevant to the issue of integration. Tidal current velocities are not correlated with wind speeds, and wave energy, although an integrated form of wind energy, is not strongly correlated with ten minute or even hourly wind speeds. Solar power has its own unique pattern of variation, and biomass is flexible. How renewable sources of generation can be combined to meet load demand is a critical issue and is discussed in later parts of the book. References [1] Star, F. ‘Flexibility of fossil fuel plant in a renewable energy scenario: possible implications for the UK’, in Renewable Electricity and the Grid, (ed. G. Boyle), Earthscan, 2007. [2] Moreira, J.R. and Poole, A.D. ‘Hydropower and its constraints’, in Renewable Energy, (ed. T.B. Johansson et al.), Earthscan, 1993.
Slide 72: 54 Renewable Energy in Power Systems [3] Bardsley, W.E. and Choudhry, S. ‘An approach to estimating hydro power system income gain from computerized water scheduling’, Natural Resources Research, September 2000, 9(3), 215–222. [4] ‘Wind energy: a vision for Europe in 2030’, Report from TPWind Advisory Council, European Wind Energy Technology Platform, 2007. [5] Conradsen, K. and Nielsen, L.B. ‘Review of Weibull statistics for estimation of wind speed distributions’, J. Climate and Applied Meteorology, August 1984, 23, 1173–1183. [6] Seguro, J.V. and Lambert, T.W. ‘Modern estimation of the parameters of the Weibull wind speed distribution for wind energy analysis,’ J. Wind Engineering and Industrial Aerodynamics, 2000, 85, 75–84. [7] CREST MSc Wind I notes, Loughborough University, 2007. [8] Sharpe, D. Chapter 3 in Wind Energy Handbook (eds T. Burton, D. Sharpe, N. Jenkins, E. Bossanyi, et al), John Wiley & Sons, Ltd, Chichester, 2001. [9] Hansen, M.O.L. Aerodynamics of Wind Turbines, James & James, London, 2000. [10] Manwell, J., McGowan, J. and Rogers, A. Wind Energy Explained, John Wiley & Sons, Ltd, Chichester, 2001. [11]Sinden, G. ‘Wind power and the UK wind resource’, Environmental Change Institute, University of Oxford, 2005. [12] Green, M.A. Solar Cells: Operating Principles, Technology, and System Applications, Prentice-Hall, Englewood Cliffs, New Jersey, 1982. [13] Nelson, J. The Physics of Solar Cells, Imperial College Press, London, 2003. [14] http://www.eere.energy.gov/consumerinfo/pdfs/solar_overview.pdf. [15] Farrell, A., et al. ‘Ethanol can contribute to energy and environmental goals’, Science, January 2006, 311, 27. [16] UK DTI report URN 01/797.
Slide 73: 3 Power Balance / Frequency Control 3.1 Introduction The essential function of an electrical power system is to meet the energy demand of consumers. This chapter describes the power demand characteristics of consumers and the technical and operational principles involved in the reliable delivery of power to them from a variety of conventional and renewable energy generators. The detailed regulatory issues of control and operational requirements adopted in different countries where electricity systems have been liberalized are addressed in Chapter 7. 3.1.1 The Power Balance Issue Consider first the case in which the power system consists of a single generator driven by a prime mover and supplying a load. In Chapter 4 it will be shown that the frequency of the generated voltage is directly proportional to the rotational speed of the generator. The generator has rotational inertia and it is assumed that the prime mover is fitted with a governor. The function of a governor is to sense any changes in speed and to adjust the fuel supplied to the prime mover so that the speed (and therefore the frequency) is controlled. The law of energy conservation requires that at any instant the power demanded by the load, usually referred to as demand or just load, is supplied by the generator and/or by energy stored within the system. If the load is suddenly increased, the extra energy demand is initially supplied by the rotational inertia of the generator through a decrease of its speed. This decrease in speed is also reflected by a proportionate decrease in frequency. The governor senses this decrease in speed and increases the fuel supply to arrest the fall in speed and frequency. Exactly how far the frequency falls, how quickly it recovers and the frequency of the new equilibrium state depends on the governor characteristics and the frequency Renewable Energy in Power Systems Leon Freris and David Infield © 2008 John Wiley & Sons, Ltd
Slide 74: 56 Renewable Energy in Power Systems dependence of the load. All these issues will be discussed later in more detail. For now it is sufficient to recognize the interdependence between demand and frequency. Complex power systems consisting of multiple interconnected generators supplying a large number of consumers respond much as the single generator, but with all the generators and loads contributing to the system response during demand changes. Just as for a single generator supplying a load, the frequency of a complex interconnected system is the same at all parts of the network. For now, the idea of a unique system frequency should be accepted without question, but it will be properly explained in Chapter 4. A power system is, of course, never in equilibrium because the demand varies continuously as consumers switch on or off their loads. It can be concluded that frequency shifts are an indicator of the imbalance between supply and demand at a particular instant. Frequency drifts downwards when demand exceeds supply and vice versa. Conventionally, power systems are run so that their frequency remains within narrow bounds because: • • • • This ensures that electric motors operate at virtually constant speed. A fixed speed is required in many consumer applications where an AC electric motor is used to drive a device at an approximately constant rate, e.g. a pump in a washing machine or a lathe in an industrial workshop. In electronic applications the mains frequency can be used as a basis for timing various processes. Transformers are sensitive to frequency variations and may be overloaded if the frequency drifts substantially from the nominal. Finally and most importantly, in traditional power stations the performance of the generators is dependent on the performance of all the auxiliary electric motor drives that deliver fuel and air to the boiler, oil to bearings and cooling services to several systems. If these auxiliaries underperform due to low speed caused by low frequency, power station output can be reduced. As will be discussed later this phenomenon could lead to a runaway situation with cascaded shutdown of power stations and blackouts. For a near-constant frequency to be maintained it is necessary that the supply of power accurately tracks the variations in demand. How and why electricity demand changes is the subject of the next section. 3.2 Electricity Demand 3.2.1 Demand Curves Figure 3.1 shows the highly variable nature of the electricity demand over a day of a typical individual house in the UK, with a minimum demand of a few watts, an average between 0.5–1 kW and the maximum in the range of 5–10 kW, i.e. 10 to 20 times the average load. Using a dedicated electricity generator to supply this house alone would be hopelessly expensive. The generator system would have to be large enough to meet the maximum demand, but most of the time it would be running at a very small fraction of its rated capability. Any fuel driven prime mover operated in this way would be woefully inefficient. Energy storage,
Slide 75: Power Balance / Frequency Control 57 7 6 5 4 3 2 1 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Time (hours) 00 Consumption (kW) Figure 3.1 Demand curve of an individual house such as batteries, could improve generation efficiency, but would have to be quite large, and therefore expensive, in relation to the average demand. If each household in the UK met its own maximum demand of say 5 kW, 100 GW of plant would be needed for this sector alone. 3.2.2 Aggregation The smoothing benefit arising from aggregation is of vital importance to electricity utilities. The more uncorrelated the demand among consumers, the more effective the overall smoothing. For a large power system this statistical effect is dramatic and is illustrated by the characteristic demand profile shown in Figure 3.2. This shows a typical demand curve over a day for the whole of England and Wales and should be contrasted with Figure 3.1. Not only is the peak to mean ratio considerably reduced but this curve is noticeably much smoother than that of the individual house. As a consequence it is much easier to predict, and the generation required to supply this aggregate load can be scheduled and controlled very efficiently, as will be discussed later. The value of interconnection to form large power systems should now be clear: it allows demand aggregation and the benefits that stem from this, primarily through the easier matching of supply and demand. Some proponents of renewable energy suggest that national grids will become redundant once generators are located near to consumers, but this is a misconception unless an unprecedented breakthrough in energy storage technology is achieved. Indeed, given the intrinsic variability of many dispersed renewable energy sources, interconnection may well prove to be even more valuable in the future.
Slide 76: 58 Renewable Energy in Power Systems 50000 45000 40000 35000 Demand (MW) 30000 25000 20000 15000 10000 5000 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 00 Time (hours) Figure 3.2 Illustration of the smoothing effects of aggregation. (Source: National Grid plc) 3.2.3 Demand-side Management – Deferrable Loads Two important linked concepts are those of deferrable loads and demand-side management. A deferrable load consumes a certain amount of energy to provide a service but is flexible in terms of exactly when that energy is supplied because it possesses either an internal storage capacity or a large thermal inertia or because the consumer is flexible about the time when he or she requires the energy service. Demand-side management is a technique used by utilities to regulate remotely the demand required by deferrable loads so that their connection to the grid is scheduled according to the availability or cost of power. The use of off-peak electricity tariffs for night-storage heating is a familiar example in the context of the UK and helps to level out the demand difference between night and day so that a greater proportion of the generation can run at a constant output. Similar incentives are offered to commercial and industrial users, but, on the whole, consumers connected to the large reliable power systems found in affluent countries rarely give much thought to when they use electricity. For the most part, the generation in these systems is designed and controlled to meet the load, rather than the other way round. Conversely, users of standalone electricity systems often time their use of electricity to suit the available generation. In fossil fuelled systems, it is usually a matter of levelling out the load. In systems relying mainly on wind or solar power, the timing of use is obviously related to the availability of the resource. Exploiting the deferability of loads is a useful tactic in any power system; it is especially valuable in systems relying on variable renewable energy sources, and can be far cheaper than employing energy storage.
Slide 77: Power Balance / Frequency Control 59 3.3 Power Governing 3.3.1 Power Conversion Chain To understand the impact of RE sources on power systems it is essential to understand the way in which thermally driven generators are controlled. This is also of direct application to biomass fuelled systems, where in many cases the plant is similar to a standard fossil fuelled plant. Figure 3.3 shows the power flow diagram for a simple fuel fired generator [1], supplying electricity to consumers as depicted in Figure 1.12 in Chapter 1. Energy is first converted from chemical into thermal form in the boiler, from thermal into mechanical form in the turbine, from mechanical into electrical form in the generator and finally back to thermal, mechanical, light or chemical form by the action of the consumers. For the frequency to remain fixed the fuel energy input to the boiler must be controlled to balance the variable consumer demand. Importantly, the chain of power conversion units in Figure 3.3 contains two stages with intrinsic energy storage. The boiler consists of kilometres of piping that carry superheated high pressure steam. Because of its large physical size it constitutes a substantial thermal store, in fact it contains enough energy to supply the turbine at full output for a few minutes. The turbogenerator itself, because of its substantial rotating mass and its high rotational speed, also contains inertial stored energy, in this case only of the order of a few full-output seconds. These stored energies can be used to follow variations in demand over short periods. Rotational energy in the turbogenerator is instantly available in response to decreases in system frequency. A sudden increment in demand will result in a frequency drop that will extract a part of the stored energy from the spinning mass. The more significant stored energy in the boiler is also available, with some delay, to supply any increments of electrical generation through the action of the governor. Fuel Boiler Turbine Generator Transmission system Load Energy from chemical Thermal Mechanical Electrical Thermal mechanical chemical light Main feedback control loops Boiler Governor Figure 3.3 Power flow diagram for a thermal power station [1]. (Reproduced with permission of McGraw-Hill)
Slide 78: 60 Renewable Energy in Power Systems 3.3.2 The Governor The steam flow from the boiler to the turbine of a large conventional fossil fuelled generator is regulated by a valve of considerable size and weight. For the turbogenerator to respond quickly to a requirement to increase or decrease its output, this valve has to be opened or closed as quickly as possible. This is achieved through a hydraulic actuator. The control signal for the hydraulic actuator is provided by the governor. In the past governors were of mechanical nature but modern generators are fitted with electronic governors. The governor ’s function is to measure the rotational speed of the generator, to compare it to the reference value (50 or 60 Hz) and, based on the error signal, to instruct the hydraulic actuator to open or close the steam valve. Figure 3.4 shows the action of the governor of a single generator supplying a load with 50 Hz AC. Governors are designed to operate as proportional control systems; i.e. an error must be present between the set point frequency and the actual frequency for the governor to alter the fuel or steam supply. This proportional control is characterized by the line in Figure 3.4 which has a fall or droop of 4% across the operational range. This 4% value has been found through extensive experience to be appropriate for stable governing and is widespread. Governors have the inbuilt facility that allows an operator to adjust the frequency at which the characteristic intercepts the frequency axis, known as the set point. For the line aa the set point is 52 Hz. With this set point, the system frequency is 52 Hz when the generator is supplying no load and is reduced incrementally to 50 Hz when the load increases to rated generator power. Changing the set point to 51 Hz moves the line to bb. Set point adjustment is of considerable importance because it allows power system operators to decide how the demand is shared by the generators on the grid, but more will be given on this later. f(Hz) 52 a b 50 a b 48 0 p 100% Figure 3.4 Frequency/power governor characteristic
Slide 79: Power Balance / Frequency Control 61 3.3.3 Parallel Operation of Two Generators When two generators are connected in parallel and are jointly supplying the demand in a small power system, the load is shared according to the set points of their governors. The best way to illustrate this sharing mechanism and the resulting system frequency is by means of an example. In the small power system of Figure 3.5(a) two generators A and B rated at 50 and 100 MW respectively supply a load of 100 MW. Both generators are fitted with governors having a droop of 4% and a no-load set point of 52 Hz. Lines aa and bb in Figure 3.5(b) show the frequency–power (f–P) characteristics of the two generators. The division of load between the generators when they are supplying 100 MW will now be determined. Let line cc located Δ f below 52 Hz represent the frequency at which this sharing takes place. (a) Using trigonometry PA/Δf = 50/2 and PB/Δf = 100/2. Therefore PA/PB = 1/2, but as PA + PB = 100 MW, then PA = 33.3 MW and PB = 66.6 MW In this case the demand is shared in proportion to the generator ratings (b) To determine the system frequency, from PA/Δf = 50/2 Δf = 33.3 × 2/50 = 1.33 and hence the frequency at which line cc intercepts the f axis is f = (52 − 1.33) = 50.67 Hz (c) Suppose that generator A is more modern and therefore more efficient in terms of fuel used per generated kW h. The set point adjustment of the governors could be used to arrive at a new more economic load-sharing condition that maximizes the contribution from generator A. One way to implement this is to shift the set point of generator B so that its characteristic is displaced from bb to BB, with the result that each generator is supplying 50 MW. The system frequency is now 50 Hz and the new set point of generator B can be found from δf/50 = 2/100. Hence δf = 1 and the new set point is 51 Hz 50 MW A B 100 MW 52 Δf f(Hz) b a B δf 100 MW 50 PA (a) (b) 50 PB 100 P(MW) B c a c b’ Figure 3.5 Parallel generators. (Reproduced with permission of McGraw-Hill)
Slide 80: 62 Renewable Energy in Power Systems 3.3.4 Multigenerator System The two-generator load-sharing paradigm can be extended to explain the behaviour of a typical power system where a very large number of generators are interconnected through a transmission network to feed a multiplicity of loads. If all generators have the same standard governor droop characteristic and a common set point, then they will all share load in proportion to their ratings. In fact, for the economic reasons illustrated in the example of Figure 3.5 and discussed at length in Chapter 7, generators with the lowest cost of energy production are run with their governor set points adjusted at full output (e.g. characteristic bb in Figure 3.5). As such generators are incapable of providing increments in power their governors are unresponsive to demand increases. A number of generators that are less economic to operate, or for other reasons to be explained later, are run with governor set points that results in them operating at part load. Such generators are capable of responding to demand increases. It can be shown that in such well integrated systems and to a good approximation all the generators with active governors can be lumped together into one very large generator having a rating equal to the sum of the ratings of the constituent generators and possessing the typical 4% droop due to the combined governor action of all the machines. This equivalent generator has a rating several times the value of the typical rating of generators on the network. The reason for this will be explained later. The f–P characteristic of such an equivalent generator is shown by aa in Figure 3.6, where f is the system frequency at which the present steady demand Pd on the equivalent generator is supplied. Because of the large size of this equivalent generator, aa intercepts the P axis far to the right. Suppose that an additional generator is connected to the network where the 4% droop characteristic is given by bb. At the system frequency f the additional generator injects into the system a power ΔP. The power balance principle now demands that the system frequency rises by an amount such that the power injection by the equivalent generator is reduced by ΔP. The figure shows that this requires a very small frequency rise Δf. It can be concluded that in a large interconnected power system, an increase or decrease of power from a single generator will have a small effect on frequency. A power system that is fed by active governor generation capacity overwhelmingly larger than the rating of a single generator is known as an infinite bus. The frequency of such a a f b Pd Δf b ΔP ΔP f a P Figure 3.6 Frequency power characteristic of an infinite bus
Slide 81: Power Balance / Frequency Control 63 system is taken to be independent of the power injected by a single generator. In Figure 3.6 this ‘independent’ frequency is given by f. Therefore, the power injected by each individual generator is defined by the intersection of its characteristic with the frequency line f. By shifting line bb parallel to itself, i.e. adjusting the governor set point, the power injected by each generator can be regulated to the desired level. 3.3.5 The Steady State Power–Frequency Relationship [1] The new frequency at which a power system will stabilize after a sudden change in generation or demand depends on: (a) the combined effect of the droops of all active governors; (b) the total load on the system and its frequency dependence. The small-increment frequency–power behaviour of a power system can be described approximately through a simple mathematical relationship. (a) When the frequency changes by −Δf (a frequency drop due, for example, to a sudden demand increase), governor action increases the power output of generators by, say, ΔPG. (b) Consumer demand consists of a variety of loads. All resistive loads i.e. heating and lighting as well as electronic devices such as PCs, are, in general, insensitive to small frequency variations. However, the large majority of electrical drives, whether in household appliances or most importantly in industry, are frequency sensitive unless they are connected to the mains through a power–electronic interface. A large induction motor drive will run marginally slower at a lower frequency, provide less mechanical power and therefore absorb lower electrical power. The decrease in total system demand can be defined by −ΔPD for a drop in frequency −Δf. It can now seen be that for a drop in frequency of −Δf, the effective increase of available power ΔP is partly due to the increase of power from generators and partly due to the released demand from loads. Therefore, ΔP = ΔPG + ΔPD for a frequency change of − Δf . If it is assumed that the small variation ΔP is linearly related to the frequency change Δf, the power frequency relationship is therefore given by ΔP = − K Δf (3.1) where K is a constant expressed in MW/Hz and is known as the power–frequency, or frequency response characteristic, or system gain. The negative sign in the equation indicates that a drop in frequency results in an increase in the net power available within a system. This relationship gives a rough value of the resulting system frequency after the initial transients have subsided in the case of a sudden major loss of generation or less usually of a sudden increase in demand. The relationship describes the situation maybe a minute after the
Slide 82: 64 Renewable Energy in Power Systems incident (see Figure 3.10, discussed later) therefore K is not related to rotating inertia effects. The value of K for a specific power system depends on the governing characteristics of the generators and the combined frequency–power sensitivity of the loads. Its value varies during the day as the share of load may shift between frequency sensitive and insensitive consumers. The value of K can be estimated by tripping out a large generator and measuring the resulting frequency drop after the transients have subsided. Preferably, values of K can be derived at random times from the records of system response after the loss of a generator due to a fault. To ensure stable operation after the most severe contingency that the utility considers as credible, i.e. probable, K should be such that the system frequency does not drop below a value that leads to serious underperformance of power station auxiliary drives. Of course, the utilities cannot protect the system from contingencies that have very low probability, e.g. the simultaneous loss of two large power stations. If such events were to occur, measures are in place to maintain the integrity of the system by means of unavoidable power interruption to some consumers, but more of this will be given later. 3.4 Dynamic Frequency Control of Large Systems 3.4.1 Demand Matching Figure 3.2 showed the daily variation of the aggregate load for a typical utility. The pattern of low load at night and the peaks in the morning and afternoon is a feature of such demand curves. Superimposed on this daily cycle are faster random variations. Power systems operation is conventionally broken down into different timescales ranging from seconds to days. During second to minute load variations partly loaded plants respond through governor action. Generating plants responding in this timescale are said to be providing continuous service or frequency response. The next time scale is load following, which involves the connection/disconnection of plant to balance the anticipated load increases/decreases. This timescale covers approximately 10 minutes to several hours during which decisions are taken in response to the trend in demand on the basis of plant operation economics. For example, in anticipation of the early morning increase in demand, the system operator either directly or through bidding mechanisms (Chapter 7) is responsible for ensuring that sufficient capacity is available to meet these large but relatively slow and predictable changes in demand. Fossil fuelled generators, and especially plant fuelled by coal, can be readily made to follow demand. Such generators are brought on line to supply any escalating demand according to cost effectiveness and flexibility. This results in the layering of generation shown in Figure 3.7. Nuclear power being inflexible occupies the bottom of the pile. CCGT generation and efficient coal occupy the next layer. These two layers supply what is known as the base load with other coal fired generation and imports from adjacent networks required to do the load following. Pumped hydro, when availble, is used to extract energy when demand is low and to deliver it during peak demand. For load following to be effective it is essential to have accurate predictions of the demand curve. Programming generation at this timescale is known as unit commitment or generation scheduling.
Slide 83: Power Balance / Frequency Control 65 50 45 40 35 Generation (GW) 30 25 20 15 10 5 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Time (hours) Pumped Hydro Generation Pumped Hydro Pumping Other Coal Gas Nuclear Figure 3.7 Demand curve of the England and Wales power system (Source: National Grid plc) The reader should not misunderstand the nature of the layering in Figure 3.7. For example, it would be wrong to assume that a constant load like office lighting is supplied from a nuclear power station while a toaster is supplied from coal! Rather a power system could be likened to a bathtub in which water is fed from all the generators and is extracted by loads. The level of water in the bathtub would represent the frequency of the system which the operators endeavour to keep constant. Some generators, e.g. nuclear, provide a constant inflow of water. Other generators are instructed to supply water when the level is detected to be falling. In such an analogy there is no way someone can tell which ‘generated water molecule’ reaches which load. In other words, electrons cannot be labelled. 3.4.2 Demand Forecasting Accurate forecasts of demand are required because: • • • Electrical energy cannot yet be stored economically. The largest proportion of generating plant is thermal in nature. An unfortunate feature of this plant is the considerable delays involved in preparing the cold generators for connection to the power system (several hours) and the restrictions in the rate at which a steam driven turbogenerator can be loaded after connection. These operational delays are dictated by the thermal/mechanical safety requirements of massive boilers and of turbogenerator sets. Thermal generators using steam turbines have an upper limit of power generation equal to their nameplate rating, but also a lower limit dictated by cavitation problems in the turbine blades at low throughputs of steam. Consequently, when a turbogenerator is connected to the network it should be loaded to a level at least equal to the minimum recommended by the manufacturers (from 30 to 50% of rated power).
Slide 84: 66 Renewable Energy in Power Systems Figure 3.7 shows that there are periods during the day (e.g. 6 to 7 am) when the rate of demand growth is considerable. To maintain system frequency, the injected power must closely track the trajectory of the demand curve. Unfortunately, because of the sluggishness of the thermal plant, this tracking cannot be done unless preparative action is taken some hours before the event. It may be concluded that there is an absolute necessity to carry out a demand forecasting activity in order to prepare and progressively load plant as required. Utilities have invested considerable effort in forecasting the daily pattern of demand. Through years of experience they have evolved sophisticated mathematical techniques to correlate demand to the aggregate of the national habits and to other factors such as weather. All methods are essentially based on the fact that demand exhibits regular patterns. Forecasting techniques adjust past demand to present weather and other conditions. Meteorological data on temperature, wind speed, humidity, cloud cover and visibility are used as variables because such factors have an important bearing on heating and lighting demand. The art of load forecasting has been refined to such an extent that estimates are rarely in error by more than ±3% and on average in the UK system they are accurate to within ±1.3%. Demand prediction techniques are constantly being refined but there will always be occasions when unforeseen circumstances increase or depress the load. The average daily errors in demand in a typical month on the English system are shown in Figure 3.8; during this period the maximum error in prediction was just under 4% and on average it was less than 1.53%. The standard error during this month was 1.6% which, as the average demand was about 32 GW, corresponds to about 300 MW. The figure shows that on 11 November the forecast was adrift by 3.5%, representing a maximum error of over 1 GW. Scheduling error, % 4 3 2 1 0 –1 –2 –3 –4 0 5 10 15 20 Day of month (November 1995) 25 30 Source: Electricity Pool Standard deviation: 1.6% Figure 3.8 Typical scheduling errors on the network in England and Wales. (Reproduced from Milborrow, D., ‘Wind power on the grid’, in: Boyle, G. (ed.), Renewable electricity and the grid – the challenge of variability, with permission of Earthscan, 2007)
Slide 85: Power Balance / Frequency Control 67 52.0 Generators start tripping 50.5 50.2 Upper statutory limit Upper operational limit Lower operational limit Lower statutory limit 50.0 49.8 49.5 48.8 47.0 Load shedding starts Load shedding complete Figure 3.9 Operational and statutory limits of system frequency for the UK power network. (Source: National Grid plc) In conclusion, even if all generation were 100% reliable, a substantial reserve would still be required because it is simply not possible to predict the demand on a power system exactly. 3.4.3 Frequency Limits Although perfect balance between supply and demand is not achievable, the balance has to be good enough so that under normal operating conditions the frequency does not drift outside prescribed operational limits. Figure 3.9 shows that for the UK, statutory limits on permissible frequency variation are ±1% (95% of week) with absolute limits in emergencies of +4% and −6%. However, the National Grid’s usual operational limits are ±0.4% which are tighter than the legal requirements. Under unusual circumstances e.g. substantial loss of demand, the frequency may rise uncontrollably and generating sets are fitted with overfrequency protection that will trip them off the system if the frequency reaches 52 Hz. At the other extreme, due to loss of generation or less frequently, unexpected demand increase, the frequency may fall below its statutory limit. Large groups of noncritical loads (i.e. excluding hospitals, emergency services etc.) are fitted with staggered under-frequency protection so that switching off such loads, i.e. load shedding begins at about 48.8 Hz and is completed by 47 Hz. Further drops in frequency could trigger cascade tripping of power stations. Such cascade tripping incidents are not unknown, several having taken place in developed countries during the last decade. Complete shutdown of the electricity supply or blackout is an occasional, but traumatic event for millions of affected citizens trapped in lifts, underground trains, etc. Full restoration of supply may take as long as several days.
Slide 86: 68 Renewable Energy in Power Systems The topic of frequency control is of particular importance in systems with substantial renewable energy penetration. 3.4.4 Generation Scheduling and Reserve Demand forecasting provides a fairly accurate picture of the expected load over the following 24 hours so that enough generators are scheduled to provide the expected demand plus a reserve, a concept to be discussed in the following section. The process is complicated by the disparate characteristics of plant on the system. In Chapter 7, it will be shown how generators are loaded on economic grounds subject to various constraints. In anticipation of an increase in demand a choice has to be made by a utility on which uncommitted generating sets from a number of available generating units (some sets may not be available as they are undergoing repairs or maintenance) have to be prepared (heated, synchronized and loaded) and eventually shut down when the demand declines. This is a complicated economic and technical choice. In a privatized energy market one company, usually the one in charge of the transmission system, has the role of plant selection, but here decisions are taken in the context of the contractual relations between the different participants in the electricity market. Chapter 7 describes how regulatory tools, tariffs and bidding systems are used in these circumstances to ensure that supply tracks demand. 3.4.5 Frequency Control at Different Timescales [2, 3] The maintenance of frequency involves a response from generators over several timescales ranging from seconds to days. This response is not only required to follow the demand’s fast variability from second to second and its slow variability over the day, but also sudden substantial mismatches between generation and demand, for example during system faults. Figure 3.10 illustrates a typical system frequency trajectory plotted on a nonlinear time scale. From the origin of the graph until about 8 seconds, the frequency exhibits the usual noise associated with minor mismatch between the continuously varying demand and the efforts of generation to match it through governor action. At 8 seconds something unpredicted and serious takes place. It could be that a large power station trips because the overhead line connecting it to the transmission system suffers a mechanical failure due to high winds and accumulated snow. This contingency results in an instantaneous large shortfall of generation. The trace describes a typical time history of the frequency and the measures taken to constrain the frequency excursion within the statutory and operational limits. Such measures are taken by all utilities but are given a variety of labels. Here the labels shown in Figure 3.10 are the ones adopted in the UK. A power system has at its disposal a number of generators with diverse characteristics. These are arranged in a hierarchy of plant appropriate for operation at different timescales as described below. A continuous or frequency response is provided by generators equipped with appropriate governing systems that control their outputs to counteract the frequency fluctuations that arise from relatively modest changes in demand or generation [2]. Large generators on the grid
Slide 87: Power Balance / Frequency Control 69 Continuous Service 50.2 Frequency (Hz) 50.0 49.8 Primary Secondary (to 30 mins) Reserve 49.5 10 s 30 s 60 s Time 10 mins 49.2 Occasional Service Figure 3.10 Frequency response requirements. (Source: National Grid plc) are selected on technical and economic merits and instructed by the system operator to operate in frequency-sensitive mode (i.e. under active governor control) to provide this service. For this to be achieved, some generators are held below maximum output. An occasional service or reserve is available to contain significant and abnormal frequency excursions caused by sudden mismatches in the generation/demand balance (e.g. loss of generation) [2]. Part-loaded large synchronized generators as well as deferrable loads fitted with frequency sensitive relays provide such services. The reserves are subdivided into primary and secondary categories. In Figure 3.10 the initial rate at which the frequency drops after the incident at 8 seconds is controlled and limited instantly by the inertial energy release from all the decelerating generators (and consumer drives) on the system. This provides a breathing time before the fast governors of some generators begin to act. Massive steam valves have to be opened hydraulically and increased steam flow has to be transported from the boiler to the turbines. It therefore takes a finite time for the substantial stored energy in the boiler to be exploited. Primary reserves require the most rapid generator response. The key requirement for generators allocated this task is that they should be capable of increasing their active power output within 10 seconds of predefined system frequency excursions and be capable of maintaining this response for a further 20 seconds [2]. Secondary reserves require a slower initial response but maintained for longer periods of time. This requires the capability of increasing the active power output within 30 seconds and maintaining the response for a further 30 minutes [2]. A fast response capacity is also provided by partly loaded hydro or pumped storage (when available), which are not bedevilled by the constraints imposed on thermal plant. Water driven plant can respond in a few minutes and be started up automatically when the frequency falls below a critical value. A fundamental feature of generators providing frequency response and reserves, collectively known as the operating margin, is the requirement for generators to have headroom in order to increase output. The operating margin is the difference between available generation and actual demand. Generators providing such services will therefore be part-loaded.
Slide 88: 70 Renewable Energy in Power Systems Besides hydro and pumped storage plant (if available) the main providers of such services are flexible large coal fired power plant. Beyond primary and secondary reserves, power systems have further robust tertiary defences known as standing reserves. These are sourced from unsynchronized standby generators capable of mains connection and generation of the instructed level of output within 20 minutes [3]. Typically, standing reserve is provided from generators driven by open cycle gas turbines and reciprocating internal combustion engines. High frequency response services are required in the event of excessive system frequency events when large loads are suddenly lost. Such services happen rarely and are initiated through governor action requiring either to reduce output or to cease generation altogether. Finally, utilities have some control over the demand by implementing voltage control. System voltage, like system frequency is rarely exactly at its nominal value but is allowed to vary within controlled limits. One response to a loss of generation, which may occur due to manual intervention or automatically, is a reduction of system voltage. Total system load will fall with voltage depending on the nature of the loads. Voltage and frequency reductions can cope with serious credible demand or generation changes although only in exceptional circumstances would both be allowed to reach their minimum or maximum values. 3.4.6 Meeting Demand and Ensuring Reliability In the early power systems, a single generator provided power to a local network that serviced a number of loads. In such a system the generator would have to be exceedingly flexible and capable of supplying the lowest to the highest demand. In the case of failure of this generator the supply would be interrupted unless a similar generator was kept in reserve providing a 100% back-up. By interconnecting such isolated systems and allowing exchange of power, the risk of generator failure can be shared so that lower back-up capacity is required to provide the same level of system reliability. As mentioned earlier, system balancing requires a number of reserves to deal with unexpected demand fluctuations and loss of generation. These reserves are determined on a statistical basis that takes into account the margin of error in demand prediction and the probability of system failures. The aim is to meet specific levels of reliability that historically have proved to be acceptable to customers. To maintain near 100% reliability would be prohibitively expensive. Critical loads such as hospital operation theatres and major computing networks are further backed up by uninterruptible power supply (UPS) systems based on batteries or local engine generators. To maintain frequency stability the operating margin or balancing reserve provided must be sized in relation to three factors: 1. The maximum deficit that may suddenly occur in a power system due to loss of generation. 2. The expected availability of all plant connected to the system. 3. The demand prediction error.
Slide 89: Power Balance / Frequency Control 71 In the UK system the transmission system operator (TSO) currently carries a balancing reserve of 2.5 GW [4]. Within limits, the larger the available headroom of the frequency response plant the greater the reliability of the power system; however, the presence of these additional part-loaded generating sets on the system significantly increases the system fuel costs. Any part-loaded machine has a poor efficiency due to standing losses: fixed losses incurred by a generator when running unloaded. The risk of demand being unmet is measured by a statistical quantity known as the lossof-load probability (LOLP). In the UK pre-privatized system a LOLP of 9% was considered acceptable, i.e. expected loss of supply incidents for each customer in nine winters per century. The LOLP gives no information on how much load will be shed or for how long. It is understood that reliabilities of the same order are expected in present privatized systems of developed countries although figures are difficult to obtain. For such a reliability to be maintained in addition to the short term reserves discussed earlier, a system or plant margin must be maintained. This is the difference between installed capacity, including imports and exports, and maximum annual peak demand. This margin is necessary because at any time there are bound to be plants that have broken down or are out for maintenance. For reliable operation, a power system must have a plant margin that is substantially greater than the operating margin. Taking as an example a system with a 70 GW winter peak demand, to guarantee that the demand will be met to a typical reliability expected in an industrialized nation about 84 GW of conventional capacity (20% above peak demand) should be available on the network. In privatized or deregulated power systems all these activities are managed through the complex financial instruments that define the contributions and obligations of several independent generation providers all feeding into the same network. For example, in the UK and in other similarly privatized networks, system operators pay considerable premiums to power generators who provide response and reserves. 3.4.7 Capacity Factor and Capacity Credit [4] When renewables displace significant amounts of conventional generation plant, an extra conventional plant capacity margin is required to maintain system supply reliability. It is important to recognize that an additional plant margin will only be required at times of low electricity demand and high input from variable sources. At other times an additional margin will not be required as there will be sufficient conventional plant available to meet demand irrespective of the contribution from variable sources. The capacity factor of a generator, as mentioned in Section 2.4.3, is usually defined as the ratio of its yearly energy output to the output it would have produced if it operated continuously at its nameplate rating. Due to unavailability caused by maintenance schedules, breakdowns, etc., no plant achieves a capacity factor of unity. Base load thermal generators have a capacity factor of 85–90% when new, declining over the years until they are decommissioned. Wind turbines achieve capacity factors of 20–40% depending on the windiness of the site. In the UK the average capacity factor of onshore wind farms is ~30% with offshore schemes achieving higher values. Often, uninformed commentators quote the figure of 30% when they claim that wind turbines require back-up for 70% of the time. In fact wind
Slide 90: 72 Renewable Energy in Power Systems turbines provide some power between cut-in and cut-out wind speeds for 80% of the time. The capacity factor does not determine back-up requirements, which must be assessed statistically. Statistical analysis confirms that, for limited penetration, the capacity factor is a good guide of the probability that the generator in question will be available to contribute towards meeting demand [5]. On this basis variable sources are incapable of providing the same level of reliable or firm power during demand peaks as conventional generators, but they are still capable of providing a contribution to this. This firmness is known as the capacity credit and is a measure of the amount of load that can be provided by variable plant with no change in the LOLP. This can be illustrated by an example. A 1 GW dispersed installation of wind farms with a 30% load factor will provide the same yearly energy output as a 350 MW gas fuelled plant of 85% load factor. Both plant statistically have a capacity credit of 300 MW. The calculations above assume that the availability of generated power is completely uncorrelated to the demand. Although this assumption is correct for the gas fuelled plant this is certainly not so in the case of the wind plant. As will be shown in the next section, due to aggregation of wind farms and the characteristics of the wind resource, there is a positive correlation between the availability of wind power and the demand. 3.5 Impact of Renewable Generation on Frequency Control and Reliability 3.5.1 Introduction The introduction of variable RE generation into a network will have an impact and incur associated costs in two main categories [4]. The first can be labelled as the balancing impact and relates to the management of demand fluctuations from seconds to hours. The second, referred to as the reliability impact relates to the requirement that there is enough generation to meet the peak demand. Both balancing and reliability involve statistical calculations. The introduction of variable RE generation introduces additional uncertainties that can be quantified in terms of operational penalties that have to be taken into account when the value of electricity from RE sources is calculated. There is a widespread, but mistaken, belief that operation of an electricity system with renewables causes serious problems. A common misconception is that significant additional plant must be held in readiness, to come on-line when the output from the wind plant ceases. This would indeed be true in an island situation, with, for example, wind the principal source of supply. Modest amounts of variable renewables within an integrated electricity system pose, however, no threat whatsoever to system operation. The reason for this is that these amounts do not add significantly to the uncertainties in predicting the generation to ensure a balance between supply and demand. Therefore the risk of changes in the output from variable renewable sources has only a small influence on the needs for reserves. In the following sections the discussion will be limited mainly to the integration issues of wind power because this is currently the nonschedulable renewable energy source making the largest impact, and is likely to remain so for the foreseeable future. In a later section the impact from other renewables will also be briefly considered. The discussion will review
Slide 91: Power Balance / Frequency Control 73 Variation in average monthly wind power output 12% Proportion of annual wind power 10% 8% 6% 4% 2% 0% Jan Feb Mar Apr May Jun Jul Month Aug Sep Oct Nov Dec Figure 3.11 Variation in monthly wind power output. (Reproduced from Sinden, G.E., 2007, D Phil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment) the temporal availability and operational aspects of wind power and the penalties incurred for different penetration levels so that the system reliability is maintained at the desired level. 3.5.2 Aggregation of Sources In Section 3.2.2 it was shown how integrated electricity systems benefit immeasurably from the aggregation of consumer demand. Fluctuating sources can benefit in the same way. Figure 2.9 in chapter 2 confirms the benefits of adding the output of geographically dispersed wind sites in the UK [6]. Aggregation here has provided its usual benefits by smoothing the output over short and medium timescales. The Monthly Distribution of Energy The seasonal wind power availability from dispersed sites in the UK shown in Figure 3.11 indicates limited production during summer and greater than average production during winter [6]. On average twice as much electricity is generated during the winter compared to the summer months. This pattern matches the seasonal demand pattern in the UK. The Daily Distribution of Energy Figure 3.12 shows that, on average, wind power availability is higher during the daytime than at night in the UK [6]. This trend is present irrespective of the time of year and is of benefit in a system where the demand peaks during the afternoon period when the wind power availability is near its maximum.
Slide 92: 74 Renewable Energy in Power Systems Average daily variation in wind power output 6% Percentage of average daily electricity 5% 4% 3% 2% 1% 0% 12 AM 2 AM 4 AM 6 AM 8 AM 10 AM 12 AM 2 AM 4 AM 6 AM 8 AM 10 AM Hour Figure 3.12 Average daily variation in wind power output. (Reproduced from Sinden, G.E., 2007, D Phil Thesis with permission of Environmental Change Institute, Oxford University Centre for the Environment) Short Term Variability The variability of wind power will cause changes in the power generated from one hour to the next. The maximum expected rate of change from hour to hour provides an indication of the reserves required to deal with shortfalls in supplying demand. Wind speed variations within the 15–25 m/s band will result in no change of power as the wind turbine will be operating at full output for winds in this range. However variations within the band 4–15 m/s will result in substantial power changes. The degree of dispersion of the resource will again be of advantage as increments of wind at one site will be compensated by decrements at another. Figure 3.13 shows the benefit of geographical dispersion from a German study [7]. In the graph, the frequency of occurrence of positive and negative changes observed in the hourly values of wind power output from actual wind farm measurements is shown for a single wind farm and for wind power aggregated over the whole of Germany. Whereas a single wind farm can exhibit hour to hour power swings of up to 60% of installed capacity this figure is less than 20% for aggregated wind farms. These maximum changes are likely to occur about once a year. These figures are significant because they indicate the requirement of fast response part loaded thermal plant to ramp-up or ramp-down and this would have economic implications as discussed later. The Capacity Factor Figure 3.14 shows the results of a UK study on the relative capacity credit between seasons and between onshore and offshore resources as a function of wind power penetration [7, 8]. Due to the windiness of the UK offshore areas the credit is higher than 60% during winter, decreasing to less than 25% in the summer for small penetrations. The onshore picture is
Slide 93: Power Balance / Frequency Control 75 10000 Occurrence per year 1000 100 10 1 –100 –80 –60 –40 –20 0 P installed 20 40 60 80 100 Wind power aggregated over Germany Single wind farm, inland Germany Figure 3.13 Frequency of occurrence of hourly wind power variations. (Reproduced from Reference [7], EWEA Report, 2005) 60 Relative Capacity Credit (%) 50 40 30 20 10 0 0-10 10-20 20-30 30-40 40-50 50-60 60-70 Wind power % of peak load 70-80 80-90 90-100 % Offshore capacity credit (%) winter Offshore capacity credit (%) summer Onshore capacity credit (%) winter Onshore capacity credit (%) summer Figure 3.14 Seasonal capacity credit as a function of penetration for onshore and offshore UK wind resource. (Reproduced from Reference [7], EWEA Report, 2005)
Slide 94: 76 Renewable Energy in Power Systems similar but scaled down. As expected, the capacity credit decreases with increasing penetration. It may be concluded that, in the UK, because of the favourable weather conditions, a statistical correlation exists between the availability of wind power and demand. The capacity credit associated with wind power and therefore the contribution of wind towards reliability could be substantially higher than that calculated from the annual average wind capacity factor. 3.5.3 Value of Energy from the Wind The financial benefit of wind power can be calculated by determining the cost of supplying the total demand and then subtracting the cost of supplying the residual demand (after wind power is added) from the existing or a reduced mix of generation. In practice this calculation is extremely difficult to do and so, for simplicity, many analysts concentrate on fuel savings which are more straightforward to estimate. Ideal fuel savings are simply those calculated from the cost of displaced generation and do not take account of any changes of operation forced on conventional plant by the time varying characteristic of the wind plant. To estimate the fuel savings more realistically the following operating issues should be considered [9]. • • • If wind generation is subtracted from the gross demand (wind power can be viewed as negative demand) the residual demand will have more variability than the gross demand. This means that the output of fossil fuel plant providing a continuous or frequency response service will have to be adjusted more frequently and to a greater extent. Additionally, with wind generation there is a need to ensure that the system can respond adequately to unpredicted changes over longer time periods. Extra balancing reserves provide greater headroom, but the lower loading level of thermal plant that this requires would result in lower operational efficiency of thermal plant. Finally, at higher penetrations (above 20%), some thermal plant may have to be shut down and started up to maintain adequate reserves. This will incur what are known as cycling costs. All of the above effects make up the balancing impact. Wind generated electricity may increase the size of the system margin required to maintain the required level of reliability. The reason for this is that wind plants are less likely than thermal plants to be available to contribute towards times of peak demand. This effect makes up the reliability impact. At high penetrations, energy may need to be spilled, discarded or curtailed because for operational reasons this energy cannot be safely absorbed while maintaining adequate reserves. The issues that define these impacts and the associated costs and penalties that have to be assigned to wind generated electricity are discussed in what follows. 3.5.4 Impact on Balancing Thermal plant incur some costs when they are run in the frequency control mode. It can be shown, [5, 9–11] that if variations in a variable source and in demand occur roughly
Slide 95: Power Balance / Frequency Control 77 independently, the total resulting variation in the net load to be met by the thermal plant is approximately a ‘sum-of-squares’ addition of the components: ⎛ Total variability ⎞ ⎛ Total variability⎞ ⎛ Total variability⎞ ⎜ of load on thermal⎟ = ⎜ of electricity ⎟ + ⎜ of variable ⎟ ⎜ ⎟⎜ ⎟⎜ ⎟ ⎝ units ⎠ ⎝ demand ⎠ ⎝ source ⎠ 2 2 2 (3.2) Thus, for example, when the average power variation of the added source equals that of the demand itself, the total variability is not doubled but increased by 40%. This has some important implications. The impact of fluctuations in variable sources at low penetrations can be taken to be practically zero; in other words this impact is just noise added to demand fluctuations. Over longer timescales, the level of operating reserve required at any given time depends on two key factors: uncertainties in demand prediction and the probability of loss of the largest generation plant on the network. When wind power plant is introduced into the system, an additional source of variation is added to the already variable nature of demand. To analyse the additional variation caused by the wind plant it is important to appreciate that the requirement is that the entire system must be balanced instead of balancing each individual load or resource. The operator has to ensure that the average system reliability is maintained at the same level it would have been without the wind resource. However, the crucial question is by how much does wind generation increase the balancing uncertainties? Intuitively it is known that minute-to-minute fluctuations in wind output are largely uncorrelated to load. This implies that the additional uncertainty introduced by wind power does not add linearly to the uncertainty of predicting the load. As for the issue of variability dealt with above, it can be shown that when errors in predicting the output from variable sources occurs independently of those in predicting demand, the combined error is again a sum-of-squares addition [9–12]: ⎛ Average error in ⎞ ⎛ Average error in ⎞ ⎛ Average error in ⎞ ⎜ predicting net load⎟ = ⎜ predicting electricity⎟ + ⎜ predicting variable⎟ ⎜ ⎟ ⎟⎜ ⎝ on thermal units ⎟ ⎜ demand ⎠⎝ ⎠ ⎠ ⎝ input 2 2 2 (3.3) Demand prediction techniques are constantly being refined but there will always be occasions when unforeseen circumstances push up or depress the load. Equation (3.3) indicates that for small penetrations of variable sources the prediction errors are lost among load fluctuations. However, since demand is fairly predictable, forecasting errors from substantial penetration of wind will incur some penalty. Analysis of the combined uncertainties of wind, demand and conventional generation based on the sum-of-squares calculation of Equation (3.2) make use of the standard error in predicting the generation/demand balance. On typical developed country networks, one hour ahead, this averages at around 1% of the demand. For four hours ahead, this figure rises to 3%.
Slide 96: 78 % of installed wind power Renewable Energy in Power Systems 10 8 6 4 2 0 0 Persistence Perfect 5 10 15 20 Wind capacity/peak demand % 25 Figure 3.15 Additional balancing power for wind (Published in Windpower Monthly News Magazine, December 2003 [12]) The need to schedule reserve to cover for possible trips of conventional thermal plant emphazises the point that no generation is 100% reliable. While the loss of a typical 1000 MW of thermal plant is a real risk, it is almost inconceivable that 1000 MW of wind plant would be suddenly lost. It is also assumed that dispersed wind plant is not sensitive to a common mode disturbance; i.e. the plant rides through voltage dips caused by faults on the transmission system. The more wind that is installed, the more widely it is spread, and sudden changes of wind output across a whole country simply will not in practice occur [13]. Calculations can be made over various timescales to determine the need for extra reserve. Figure 3.15 shows the estimated additional balancing power needed (expressed as a percentage of installed wind power) as a function of wind power penetration [12]. At 20% penetration, 7% of extra operating reserves are required if persistence1 in the wind is assumed. For perfect forecasting only 2% of additional capacity would be needed. Clearly these back-up figures are modest, but what about the associated costs? In the UK, the TNO has estimated that 10% of wind penetration would increase balancing costs by £40 m a year, which is equivalent to £0.002/kW h. A 20% penetration will increase the cost to £0.003/kW h. These figures should be viewed in terms of the retail cost of electricity, which at the time of writing is in the region of £0.10/kW h [14]. Hence a 20% penetration in the UK would incur a 3% additional cost on electricity at present prices. For relatively little expenditure the predictability of wind could be greatly improved. This could be accomplished partly through the installation of extra weather data monitoring stations (e.g. anemometry towers a few tens of kilometres from major wind farms) and partly through sophisticated computational techniques. Programmes to provide enhanced predictability are being developed in several countries [15]. Figure 3.16 illustrates this by showing a typical one hour wind forecast using sophisticated techniques against actual output for one wind farm over a period of a week. Such techniques could provide considerable cost benefits in operating reserve. 1 ‘Persistence’ wind forecasting assumes that the wind power output one hour ahead is the same as at time zero. ‘Perfect’ forecasting as in Figure 3.15 assumes that wind power can be predicted with total precision.
Slide 97: Power Balance / Frequency Control 79 100 90 80 Power (% of capacity) 70 60 50 40 30 20 10 0 1 May 2004 2 May 2004 3 May 2004 4 May 2004 5 May 2004 6 May 2004 7 May 2004 Figure 3.16 Wind farm forecast (+1 hour) versus actual output, Ireland, 2004. (Reproduced with permission of Garrad Hassan and Partners Ltd) 3.5.5 Impact on Reliability [6] In addition to the operating reserve, some system margin in excess of the system peak demand is required and this will be affected by the level of wind penetration. The example of a power system with a peak demand of 70 GW will be given, [6]. To guarantee long term security to an accepted level of reliability utilities in general would require that about 20% (14 GW) of that peak must be additionally available on the network as the system margin. The system in question has a yearly energy demand of 350 TW h. If 10% of the system energy were to be generated by wind power it is possible to assess what the new plant margin should be. To generate 10% of the energy, i.e. 35 TW h, a wind power capacity of 35 000/(8736 × 0.3) = 13 GW, where 8736 is the number of hours in the year and 0.3 is the average wind farm capacity factor. The 35 TW h would be generated by 35 000/(8736 × 0.75) = 5.3 GW of conventional generation with a yearly load factor of 0.75. Ideally the wind plant would directly substitute for this conventional plant, but the variability of wind power means that not all of this conventional plant can be removed from the network. Statistical calculations by utilities indicate that 81 GW of conventional plant would be required to meet the same level of reliability. Therefore only 3 GW (84 − 81 GW) can be removed from the network; i.e. 2.3 GW (5.3 − 3 GW) is retained because of wind power variability, which is 17% of the wind power installed capacity. In the UK context of a planned 20% penetration from wind and to maintain system reliability the associated cost lies within 0.03 and 0.05 p/kW h. 3.5.6 Discarded/Curtailed Energy As the capacity of the variable sources injected into a system increases, there might be occasions when the available power from such sources cannot be used. If the penetration is sub-
Slide 98: 80 Renewable Energy in Power Systems stantial, there might be periods when the available power from renewables exceeds demand, or cannot be accommodated by the transmission or distribution system. However, even before this stage is reached, energy from variable sources will have to be shed because the power system would need to keep a minimum level of thermal plant generation in order to maintain adequate operating reserve. Discarding energy from variable sources poses no particular operational difficulties. Output from wind turbines can be controlled through blade pitch variation, from photovoltaics through inverter control and from hydro, wave and tidal schemes through similar control techniques. However, this discarding or curtailment of energy results in an economic penalty on variable sources, which becomes increasingly important at high penetrations. This penalty is difficult to assess as it depends heavily on the flexibility of the base load units, i.e. the extent to which they could be operated in a stable regime at low power and upon how rapidly their output could be increased if required. For the level of penetration expected over the next decade or so, the penalties due to discarded energy are unlikely to be of major significance. It should also be noted that curtailment may also be required due to distribution or transmission system constraints. Reference [4] states that 5 out of 6 studies show that at a penetration of 20%, curtailed energy ranges from 0 to 7%. Most studies show that, with sensible design, curtailment due to local network capacity limitations would be rarely required. 3.5.7 Overall Penalties Due to Increasing Penetration In previous sections, operational penalties due to increases in variable source penetration were reviewed. A number of studies have been carried out to provide estimates of these penalties as the penetration of renewables increases. The majority of these studies relate to wind power, as this is the variable source with the largest installed world capacity to date. Table 3.1 gives some indicative figures based on EU-funded studies and on Danish and UK thresholds linked to operational experience of wind farms. The additional costs are tabulated in terms of the level of penetration. Table 3.1 Implications of increasing wind energy supply on the UK network Wind power penetration Up to 5% 5–10% Measures required Cost penalties 10–20% 20–50% None Occasional instances when some energy from the wind is discarded and more part loading of the thermal plant required As above, plus more use of pumped storage or hydro to balance wind power May be necessary to build more storage, or peaking plant, or retain old coal plant, depending on relative costs (extra storage will benefit the system as a whole) Negligible 0.1–0.2 p/kW h 0.2–0.8 p/kW h >0.8 p/kW h
Slide 99: Power Balance / Frequency Control 81 3.5.8 Combining Different Renewable Sources The benefits of combining different variable sources were mentioned in Section 2.10. As the capacity of a nondispatchable source increases, its marginal value declines, primarily because successive increments of capacity are correlated with those already on the system. In contrast, combining capacity from renewables with uncorrelated or complementary outputs can therefore be of considerable benefit [9, 10, 17]. Typically, a combination of wind and solar could be beneficial. In some circumstances, thermally driven winds can be strongest after sunset, so that the combination of wind and solar usefully covers periods of high demand. Other studies indicate that a combination of wind and tidal (two sources having statistical independence) increases their value compared with the case of having more of the same [9]. A recent study [14] indicates that a contribution from a mix of PV solar and wind plus domestic combined heat and power has the potential to reduce significantly the overall variability that would have been experienced if only one renewable technology were to provide the total contribution. The potential synergies among different renewable sources are clearly much too important to ignore, and they may often make the combined exploitable potential larger than the sum of the parts considered in isolation. 3.5.9 Differences Between Electricity Systems [4] It is appropriate here to stress that results from studies on one particular network do not necessarily apply elsewhere. The operational viability and costs of integrating renewable energy depend on a number of factors that characterize the local resource as well as the structure of the electricity network. These factors include: • • • • • • • • • the strength and temporal variability of the resource; the possibility of geographical dispersion over a large area to gain the advantages of aggregation; the possible complementarity between different types of renewable resources; the correlation, if any, between availability of the resource and demand variation; the extent to which the magnitude of the resources can be forecast, where some weather patterns are more predictable than others; the robustness of the electricity network and the proximity of transmission lines to the areas of maximum resource availability; the transmission links, if any, to adjacent networks; the operating practices of the network, in particular how far in advance the system balancing reserve is planned; the type of conventional plant in the network, for example, smaller and more modern thermal plants are more flexible than large base load plant such as nuclear. 3.5.10 Limits of Penetration from Nondispatchable Sources Early on in the development of renewables, the UK’s Central Electricity Generating Board (CEGB) carried out a number of extensive simulation studies to estimate the impact of large
Slide 100: 82 Renewable Energy in Power Systems Figure 3.17 CEGB simulation study of large penetration. (From Grubb, M.J., IEE Proceedings C, V. 138(2), March 1991, reproduced with permission of IET) penetrations [9, 10]. To illustrate the challenges, an extreme example from one such study is reproduced here. Figure 3.17 shows the output that would have been expected from 25 GW of dispersed wind capacity (middle graph) and 10 GW of tidal (lower graph) alongside the demand over a period of one month. In this example 30 and 13% of the consumer energy would be supplied from wind and tidal respectively. The wind power varies less rapidly than demand, tidal more rapidly. Figure 3.18 shows the effect of subtracting the output from wind and tidal power from the demand, leaving a residual load (dashed curve) to be met by a conventional thermal plant. More recently [18, 19], studies have considered the possibility of meeting the entire demand from a mix of renewables. Figure 3.19 taken from Streater ’s work [18] shows hour by hour variations in the time variable renewables. Penetration levels of such a high magnitude result in periods when the available power from the RE sources exceeds demand. Even before this stage is reached, for reasons of system reliability, the RE sources would have to be curtailed. Table 3.2 shows the capacities of the different renewables installed and their penetration, defined as the output divided by the total load (i.e. ignoring any curtailment.) It is also apparent that for the year in question a deficit occurs between weeks 15 to 18 and weeks 44 to 52. In principal the shortfall could be made up from biomass based generation. Denmark and less so, Germany generate a high percentage of their total electricity needs from wind power and are planning further capacity. A recent study from Elkraft, a Danish system operator, asserts that a wind penetration up to 50% is technically and economically feasible. This is based on an increase of installed capacity of 3.1 GW in 2005 to 5 GW in 2025 and a substantial expansion of the Danish grid. The study claims that even with this
Slide 101: Power Balance / Frequency Control 83 Figure 3.18 Residual demand to be met by conventional generation. (From Grubb, M.J., IEE Proceedings, 138(2), March 1991, reproduced with permission of IET) 50000 solar PV 40000 Average Weekly Energy Production and Requirement (GWh) aggregated energy supply (variable sources) wave source-load 'power' energy requirement 30000 offshore wind 'heating' energy requirement aggregated energy requirement onshore wind 20000 10000 0 0 5 10 15 20 25 30 35 40 45 50 -10000 -20000 Week Number Figure 3.19 Hour by hour variation in renewable energy generation over one year, compared with variations in energy requirements
Slide 102: 84 Renewable Energy in Power Systems Table 3.2 Assumed renewable energy capacities and production in the Streater study [18] Annual energy produced (TW h) Onshore wind Offshore wind Solar photovoltaic Wave 72 000 144 000 36 000 108 000 Percentage of total variable load (%) 20 40 10 30 Mean power output (GW) 8.2 16.4 4.1 12.3 penetration, it will be necessary to shut down wind turbines only for a few hours in a year. Similar studies by DENA in Germany [20] indicate that up to 20% penetration by wind is possible with minor extensions of the grid and no need for construction of additional power stations. 3.6 Frequency Response Services from Renewables [2] With the anticipated rise in the penetration of variable renewables, power systems will be required to accommodate increasing second to second imbalances between generation and demand requiring enhanced frequency control balancing services. Some renewable generation in principle may contribute to frequency regulation services, but this would require headroom in the form of part-loading. Technologies that could potentially provide such services are biomass, water power, photovoltaics and variable speed wind turbines. In Chapter 7 it will be shown that economics dictate that energy from renewable sources should generally be used as fully as possible whenever available. Although this seems to contradict the idea of part loading such plant, there are some occasions when priorities may dictate otherwise. With large penetrations from renewables there will be occasions, for instance during low demand days over summer, when the number of conventional generators needed to supply the residual load will be so few that an adequate level of response and reserve may be difficult to maintain. Under such conditions renewable generators could be unloaded and instructed to take part in frequency regulation. Such a provision has been made, for example in the Irish (ESB) code, for the connection of wind turbines. In a privatized system the opportunity benefits of running in this mode must more than compensate the loss of revenue from generating at less than the maximum potential. 3.6.1 Wind power Early wind power technology was mainly based on simple fixed one or two speed stallregulated wind turbines with little control over the dynamic performance of the generator. However, over recent years active stall and pitch regulated variable speed wind turbines have been developed that are capable of increased conversion efficiencies but also of substantial control capabilities. In principle, modern wind turbines are capable of providing a continuous
Slide 103: Power Balance / Frequency Control 85 response by fast increase in power from part loading through blade pitch control in response to drops in frequency and through the same mechanism provide high frequency response through fast reduction in power in response to increases in frequency. As wind power capacity has increased, to the extent that at times it is the dominant form of generation in parts of Denmark and Northern Germany, there is an increasing demand for wind capacity to be dispatchable and to behave more like conventional generation. Very large wind farms are now expected to conform to connection standards that limit ramp rates for increase in power and also to contribute to frequency regulation under times of high network stress. These requirements are increasingly included in the national grid codes that regulate access to the public networks. In these early days it is unclear to what extent this will result in wind power being curtailed, for example to comply with given ramp rates, and to what extent such constraints add value to the system operator. Conventional steam generation plant assist the network frequency stability at the onset of a sudden imbalance of demand over supply by slowing down. Wind turbines respond differently. The stored energy is in the rotor inertia and fixed speed turbines will provide a limited benefit from their inertia provided that the voltage and frequency remain within their operating limits. Variable-speed wind turbines will not normally provide this benefit as their speed is controlled to maximize the energy production from the prevailing wind. Large wind turbines are now almost always of the variable speed type and as they increasingly displace conventional generation the total system inertia from such generation will decrease. Consequently the rate of change of frequency and the depth of the frequency dip caused by a sudden loss of generation will both increase. However, variable speed wind turbines could be controlled in principle to provide a proportionately greater inertial energy to the system than conventional plant of the same rating. Such sophisticated control arrangements to support system functions are likely to be requested by utilities as wind penetration increases. Finally, grid codes require wind turbines to maintain power infeeds to the system even under transient local voltage reductions. Such reductions are usually due to fault conditions in the vicinity of the wind farm. It can be shown that maintenance of power infeed from all generators is essential to ensure system recovery after a fault clearance. 3.6.2 Biofuels [9] Traditional thermal plant could be described as capacity limited, i.e. capable of theoretically generating its rated output continuously, as gas, coal, oil or fissionable material is abundantly available on demand. In contrast, an energy crop based plant could be described as energy limited because the locally harvested fuel is limited in nature and may or may not be capable of sustaining all year round continuous plant generation at full capacity. Transporting biomass fuel from remote areas would not be economical. A biomass plant would be expected to operate as a base-load generator running as far as possible at full output. Such plant would be able to contribute to continuous low or high frequency response services similarly to a conventional plant. For a low frequency response the plant would need to run part-loaded, a convenient strategy providing extra income if, say, due to a low crop yield year the stored fuel would not be capable of servicing continuous full output.
Slide 104: 86 Renewable Energy in Power Systems The land filled gas plant size is in the range of 0.5–1.5 MW and because of their small size they would not be suitable for the provision of frequency regulation services. 3.6.3 Water Power Small and medium sized hydro schemes without significant storage capacity are characterized by substantial variability of output depending on rainfall. Because of this and their small size they are not suitable for frequency regulation duties. Tidal schemes could be very large indeed and their output would be highly predictable. Such plant would incur exceptionally high upfront capital costs and long payback periods. Operation revenue is vital to service the large loans and it is unlikely that frequency response revenue based on part-loading would be attractive enough. As such schemes are not yet in the planning stages, the jury is still out on their frequency control capabilities. The comments made above on wind power generally apply with reservations to future wave power schemes. As the technology is still in its infancy and commercial schemes are not yet in existence, it is not known how their dynamics may be capable of responding to signals derived from frequency deviations. 3.6.4 Photovoltaics Here a distinction should be made between large concentrated PV installations and numerous roof top systems. For large installations, the comments on wind power apply albeit with some reservations. As the PV systems are interfaced to the grids through power electronic converters and as no mechanical inertia is involved, the speed of response in increments or decrements in power flow can be very fast indeed. On the other hand, solar radiation tends to vary more slowly than wind in the short term, and is fairly predictable. As with other renewables, in the future, PV systems may be required to operate at part load, thus providing ‘headroom’ for continuous or ‘occasional’ frequency control. At this stage of PV technology, roof installations are not yet numerous enough to provide a credible frequency response service. However, in years to come if, as predicted, costs plummet and installations are numbered in millions, it is conceivable that the local inverters are fitted with sophisticated controllers to assist system frequency stability. 3.7 Frequency Control Modelling [21] 3.7.1 Background Simple lumped parameter modelling is able to provide a rough estimate of the effect on frequency control of feeding relatively large amounts of renewable energy into a power system. A system model of this sort is outlined below, together with results showing the dynamic impact of high penetration of wind energy, and also the way in which dynamic control of demand can be used to good effect.
Slide 105: Power Balance / Frequency Control 87 Modelling a Generator The dynamics of real generator sets are highly complex and differ considerably between sets. However, it has been shown that a governor with a droop characteristic can be usefully modelled as a proportional controller [22]. The first step is to calculate the generator ’s target power output, PTAR, using the 4% droop characteristic: ⎛ fSP − f ⎞ PTAR = ⎜ P ⎝ 0.04 × fNOM ⎟ MAX ⎠ (3.4) where fSP is the generator ’s set point in Hz, f is the current grid frequency, fNOM is the nominal grid frequency and PMAX is the generator ’s capacity in MW. The next step is to reduce the error proportionally over time between PTAR and the actual output P at time t, using P(t + dT ) = P(t ) + ( PTAR − P(t ) ) GdT (3.5) where G is the governor gain. It can be shown that an appropriate value for G is about 0.3 as this results in a realistic settling time in frequency of the order of 15 to 20 seconds after a step-change in load. To get approximate but useful results it has been shown [22] that the total amount of primary response on the system can be modelled by a single governor-controlled generator of sufficient size to represent all generators with headroom. Also, the total amount of base generation can be modelled by an additional very large generator, but on fixed full output. Modelling Released Demand Many loads on the grid consist of rotating machines. As mentioned in Section 3.3.5 there is a built-in frequency dependence caused by the fact that these machines slow down as the frequency drops, and thus consume less power. It has been found empirically that for the UK the total active power demand decreases by 1–2% for a 1% fall in frequency depending on the load damping constant, D [22]. This change in power is the released demand. It is treated in the simulation as an injection of active power, PR, given by ⎛ f − fNOM ⎞ PR = − DPL ⎜ ⎝ fNOM ⎟ ⎠ (3.6) where D for the UK is assumed here to be 1.0 and PL is the total load if no built-in frequency dependence exists. Modelling the Grid’s Inertial Energy Store As already stated, the grid frequency falls as all the spinning machines on the system begin to slow down. In effect, the demand deficit is being met by extracting energy from the rota-
Slide 106: 88 Renewable Energy in Power Systems tional inertia of all the generators (and spinning loads). The fall in frequency will continue until the demand deficit is met by a combination of released demand and increased generation due to the governor response. In the simulation, all the inertia is assumed to be stored in an equivalent single flywheel of moment of inertia, I, rotating at grid frequency, ω = 2πf rad/s. The total kinetic energy stored is therefore KE = 12 Iω 2 (3.7) The inertial storage capacity of a power system is measured by an inertia constant, H, which is the number of full-output seconds of energy stored (assuming nominal frequency). Therefore, 1 HPGMAX = KE = I ω 2 2 where PGMAX is the total generation capacity. H is typically within the range of 2–8 seconds [22]. For this study, H is assumed to be 4. The inertia I for the system is calculated once at the start of the run: I= 2 PGMAX H 2 ω NOM (3.8) For each step of the simulation, the total power surplus, PS is then calculated: PS = PG + PR − PL (3.9) where PG is the total generation, PR is the released demand and PL is the load. Clearly, PS is the power going into the inertial energy store. Given that for each simulation time slice, dT, energy must be conserved, then KE (t + dT ) = KE (t ) + PS dT Hence 12 1 I ω (t + dT ) = I ω (2t ) + PS dT 2 2 (3.11) (3.10) which provides a difference equation for calculating the new frequency for each step of the simulation: ω (t + dT ) = ω (2t ) + 2 PS dT I (3.12) The simulation is then carried through the following steps for each time slice, dT:
Slide 107: Power Balance / Frequency Control 89 1. 2. 3. 4. 5. Calculate PL by summing the connected loads. Calculate PG by summing the total generation. Calculate PR using Equation (3.6). Calculate PS using Equation (3.9). Calculate the new ω (and f ) using Equation (3.12). 3.7.2 A Modelling Example The model described above was used to assess the effect that a large wind power input would have on the frequency stability of a power system. The example simulated is extreme and has been chosen because it illustrates key issues. Wind speed data from 23 UK sites were used in the simulation. A 50 hour period containing exceptional wind variability was chosen so as to provide a major challenge to integration. It was assumed that the variation in wind speed and physical separation of the wind turbines in each site would smooth second to second variations in power and thus the power system could be adequately modelled on a minute to minute basis. For each site, the power output was calculated on the assumption that a wind farm comprising 150 4 MW variable speed wind turbines was present at each site. A purely cubic power–wind speed relationship was assumed with a cut-in wind speed of 2 m/s, a rated wind speed of 15 m/s and a cut-out speed of 25 m/s. The output power from the 23 sites were added together to give a total maximum generation capacity of 13.8 GW. This represents a level of penetration of 25% as a fraction of peak demand. The simulation results are shown in Figure 3.20. The maximum power reached in the 50 hours chosen was 4.6 GW and the minimum was 550 MW. During the period, the largest sustained drop in wind power occurred during the 37th hour when 5.5 GW of wind was lost in 4.5 hours. Figure 3.20 Simulated wind power using measured wind speed data from 23 UK sites
Slide 108: 90 Renewable Energy in Power Systems 50.8 without DDC 50.6 50.4 Frequency (Hz) 50.2 50 49.8 49.6 49.4 0 10 20 Time (h) 30 with DDC 40 50 Figure 3.21 Simulation of frequency with wind on the system A simulation was set up with the above wind generation connected, along with 3170 MW reserve (providing both primary and secondary response), and a frequency dependent load of 36 000 MW at 50 Hz. The level of reserve was chosen arbitrarily so that the maximum frequency excursion would remain approximately within the statutory limits. Enough base generation was provided such that the reserve was half-loaded when the wind power was at its average for the 50 hour period – approximately 36 000 − 2500 − (3170/2) ≈ 32 000 MW. Wind predictions were not used and the reserve requirements are large enough to compensate for substantial but untypical large positive or negative variations of the wind resource over several hours. The resulting grid frequency–time relationship for this simulation is shown by the curve ‘without DDC (dynamic demand control, see Section 3.8.3)’ in Figure 3.21. Between 20 and 25 hours, the wind speed increased substantially to the extent that at 24 hours the base load plus the input from the wind exceeded the nominal demand. The consequence was that the frequency increased excessively to a level determined by the frequency dependence of the load. This undesirable situation could have been easily avoided by fitting wind turbines with a frequency dependent power regulator that reduces power generated at higher frequencies. At the opposite extreme between 40 and 42 hours, the output from the wind farms was so low that the base load plus the total available reserve (the remaining half of 3170 MW) was insufficient to balance the demand and the frequency had to drop to a level determined again by the frequency dependence of the load. This undesirable situation would have been prevented if at 37 hours use was made of meteorological data predicting a power decline from the wind. Standby generating plant would then have been commissioned to provide the expected shortfall in generation.
Slide 109: Power Balance / Frequency Control 91 3.8 Energy Storage 3.8.1 Introduction Energy storage devices capable of being topped up during periods of low demand and drained during periods of high demand would aid enormously the frequency control problem. Pumped storage schemes are classical but very expensive examples of such devices. Storage devices, if suitably cheap and efficient could be of benefit to the integration of high levels of RE source penetration, although if such devices were to be present in a power system they would be of operational benefit to traditional energy sources as well. Income may be derived from an energy store by charging it when the local electricity value is low and discharging it when the value is high, but the level of income critically depends on the round trip efficiency and how this compares with the system electricity cost differential. If, at times, the grid at the point of connection of the embedded renewable generation cannot absorb the entire output of the generator a local storage device could prevent curtailment. Whether this is an economically viable strategy depends, among other things, on the capital cost, the round trip efficiency and the operation and maintenance costs of the device. Another source of income from such devices results from the supply of ancillary services, for example reactive power, voltage and frequency control, and emergency power during a power outage. It is clear that assessing the economic benefit of storage is not a simple matter and that if renewable energy sources have low capital costs it may be more effective to curtail output whenever a surplus exists, rather than attempt to store the surplus. 3.8.2 Storage Devices As mentioned above, storage technologies depending on their characteristics may aid the integration of renewables, and they could assist the operation/control of a power system over the range of timescales discussed in Section 3.4.6. Conventional technologies include: 1. The large hydro is an ‘old’ renewable but whenever available could facilitate the take-up of variable renewable sources by suitable timing of water release. 2. Compressed air storage has been talked about for some time and involves the storage of compressed air in disused underground cavities, e.g. exhausted salt mines. At present it is uncertain how economically and technologically viable this technology is. 3. Pumped hydro is an excellent energy storage technique as far as the maturity of technology is concerned, but few attractive sites exist and upfront capital costs are very high. Less conventional technologies include: 4. Hydrogen can be produced by the electrolysis of water using energy from a renewable resource. It can then be ‘burnt’ 2 as fuel to generate electricity. Alternatively it can be piped as a gas or liquid to consumers to be used locally providing both electricity and heating in a total energy scheme, or it can be used for transport. The combustion of 2 Literally burnt or oxidized in a fuel cell.
Slide 110: 92 Renewable Energy in Power Systems 5. 6. 7. 8. 9. 10. hydrogen results in energy plus pure water with no harmful emissions or by-products. This may not be attractive if electricity is the final product since the round trip efficiency is very low (generally below 50%). For this reason there is much interest in using hydrogen for transport, but this depends on much improved on board storage systems. Flow cells operate in a mode similar to that of a car battery but without involving the electrodes. Instead, when the flow cell is used as a ‘sink’, electrical energy is converted into chemical energy by ‘charging’ two liquid electrolyte solutions. The stored energy can be released on discharge. In common with all DC systems connected to the AC network, a bidirectional power electronic converter is required. Batteries. The lead–acid battery is one of the most developed battery technologies. It is a low cost and popular choice of electrical energy storage but has disadvantages in terms of energy density per unit weight, a short cycle life and the dependence of the delivered energy on the rate of discharge. As a consequence, a large variety of other batteries have been under intensive development to provide high round trip efficiency, low life-cycle cost, high reliability, capability of deep discharge and a large number of charge–discharge cycles, low maintenance, high power energy density per unit weight and low capital cost. Although they are generally perceived as too expensive for general inclusion in large power systems, high temperature batteries such as the Zebra cell [23], are of increasing commercial interest at the substation level. Flywheels. There has been a number of projects carried out to store energy in very fast rotating flywheels. The major problems that had to be overcome were the maintenance issues and losses relating to bearings, the low specific strength of standard materials such as steel and the associated severe failure management problems at high speeds. Contemporary flywheels are made of fibre-reinforced composites which have powdered magnetic material introduced into the composite which when magnetized form either the rotor of a high speed motor/generator or the rotating element of passive magnetic bearings. The flywheel motor/generator is interfaced to the mains through a power electronic converter. Currently the technology is expensive and only used for niche applications. Super capacitors, alternatively known as ultracapacitors, consist of a pair of metal foil electrodes, each of which has an activated carbon material deposited on one side. The activated carbon sides are separated by a paper membrane and then rolled into a package. Ultracapacitor operation relies on an electrostatic effect whereby charging and discharging takes place with the purely physical (not chemical) and reversible movement of ions. As a result there are some fundamental property differences between ultracapacitors and battery technologies including long shelf and operating life as well as large charge– discharge cycles of up to 500 000. Superconducting magnetic energy storage (SMES) stores energy within a magnetic field created by the flow of direct current in a coil of superconducting material. Typically, the coil is maintained in its superconducting state through immersion in liquid helium at 4.2 K within a vacuum-insulated cryostat. A power electronic converter interfaces the SMES to the grid and controls the energy flow bidirectionally. With the recent development of materials that exhibit superconductivity closer to room temperatures this technology may become economically viable. Heat or cold store. There is a long tradition of using thermal storage to assist in power system operation. The UK’s old white metre off-peak tariff and the more recent Economy 7 tariff were both primarily for charging storage heaters. More recently, sophisticated
Slide 111: Power Balance / Frequency Control 93 ways to use heat and cold storage in the context of power systems has been explored. This technology involves modulation of the energy absorbed by individual consumer electric heating elements and refrigeration systems for the benefit of overall system power balance. An example of this is described in the next section. 3.8.3 Dynamic Demand Control A scheme is being investigated that uses the already existing stored energy in millions of consumer appliances and requires the installation of dynamic demand control (DDC). These monitor system frequency and switch the appliance on or off, striking a compromise between the needs of the appliance and the grid. Initially fridges and freezer applications have been investigated. Refrigerators are ‘on’ in all seasons, throughout the day and night, and are therefore available to participate in frequency control at all times. The total energy demand on the UK grid from domestic (excluding industrial and commercial) refrigeration has been estimated as 16.7 TW h per year which amounts to an average load of 1.9 GW. The refrigeration load is dependent on ambient temperature, winter load being approximately two-thirds that in summer. Daytime load is also slightly higher than that at night. Refrigerators are designed to handle considerable switching as they typically have a switching cycle of the order of 5 minutes to 1 hour depending on characteristics and contents. Any additional switching caused by frequency control should not therefore present a problem. As an example, the power system model of Section 3.7.2 was used with the addition of DDC refrigerators. For this purpose, a refrigerator with a dynamic demand controller was modelled. The aggregation of 24.9 million such appliances (one per UK household) with statistically uncorrelated behaviour and equivalent to 1320 MW of deferrable load was investigated for its response to fluctuating wind power with a reserve of 2000 MW. Figure 3.21 shows that the system with the DDC (black trace) considerably reduced the variation in frequency even though the system was operating with substantially less reserve. This is because the simulated controllers reacted more quickly than the generator governor to changes in frequency. The graph also shows that the system frequency with and without DDC fell below the operational limit of 49.8 Hz. However, the system with DDC provided a considerable breathing time. Frequency did not fall below the operational limit until nearly 2 hours after the non-DDC system. A team of engineers operating the power network would therefore be given a wider choice of generation with which to balance the system including slower acting (and therefore possibly cheaper and more efficient) options. Also the delay may allow generation to be scheduled more cost effectively through the electricity market, which may operate a ‘gate closure’ (see Chapter 7) time of half an hour in advance of any particular generation slot. It may be concluded that an aggregation of a large number of dynamically controlled loads has the potential of providing added frequency stability and smoothing to power networks, both at times of sudden increase in demand (or loss of generation) and during times of fluctuating wind or other renewable power. The devices, if incorporated on a real system, could displace some reserve and result in a significant reduction in governor activity of the remaining generators with assigned headroom. The amount of reserve displaced will depend on the
Slide 112: 94 Renewable Energy in Power Systems extent to which low magnitude but long term frequency excursions can be tolerated, and on the amount of slower-acting back-up generation available, but it could be of the order of the total amount of dynamically controlled loads connected. The potential demand that could be operated under dynamic control is considerable. Deepfreeze units, industrial and commercial refrigeration, air conditioning as well as water heating systems could provide DDC. In principle, the potential available in a developed country would be several GW. The future will tell whether this potentially useful, but not essential, companion to variable renewable energy sources will attract the attention it deserves. References [1] Laithwaite, E.R. and Freris, L.L. Electric Energy: Its Generation, Transmission and Use, McGraw-Hill, Maidenhead, 1980. [2] ‘Ancillary service provision from distributed generation’, DTI report, Contract No. DG/CG/00030/00/00, URN No. 04/1738, UK. [3] ‘National Grid Seven Year Statement’, available from www.nationalgrid.com/uk/library/documents. [4] ‘The costs and impacts of intermittency’, Report by UKERC, Imperial College, London, March 2006. [5] Farmer, E.D., Newman, V.G. and Ashmole, P.H. ‘Economic and operational implications of a complex of wind-driven power generators on a power system’, IEE Proceedings A, 1980, 127(5). [6] Sinden, G. ‘Wind power and the UK wind resource’, Environmental Change Institute, University of Oxford, 2005. [7] Van Hulle, F. ‘Large scale integration of wind energy in the European power supply: analysis, issues and recommendations’, Report by EWEA, December 2005. [8] ‘ILEX Energy Consulting and UMIST at quantifying the system costs of additional renewables in 2020’, A Report of DTI and UMIST, October 2002. [9] Grubb, M.J. ‘Value of variable sources on power systems’, IEE Proceedings C, March 1991, 138(2). [10] Grubb, M.J. ‘The integration of renewable energy sources’, Energy Policy, September 1991. [11]Parsons, B. and Milligan, M. ‘Grid impacts of wind power: a summary of recent studies in the United States’, in European Wind Energy Conference, Madrid, June 2003. [12] Milborrow, D. ‘Forecasting for scheduled delivery’, Wind Power Monthly, December 2003, 19(12), p.37. [13] Milborrow, D. ‘False alarm’, IEE Power Engineering, April/May 2005. [14] Sinden, G. ‘The practicalities of developing renewable energy stand-by capacity and intermittency’, Submission to the Science and Technology Select Committee of the House of Lords, Environmental Change Institute, University of Oxford. [15] Kariniotakis, G., Marti, I., et al. ‘What performance can be expected by short-term wind power prediction models depending on site characteristics?’ in Eropean Wind Energy Conference (EWEC 2004), London, 2004. [16] ‘Wind power in the UK’, Sustainable Development Commission, UK, 2006. [17] Sinden, G. ‘The practicalities of developing renewable energy. Stand-by capacity and intermittency’, Submission to the Science and Technology Committee of the House of Lords’ Environmental Change Institute, University of Oxford. [18] Streater, C.J.M. ‘Scenarios for supply of 100% of UK energy requirements from renewable sources’, REST MSc Thesis, Department of Electronic and Electrical Engineering, Loughborough University, UK. [19] Barret, ‘Integrated systems modelling’, Chapter in Renewable Electricity and the Grid (ed. G. Boyle), Earthscan, 2007. [20] ‘Planning of the grid integration of wind energy in Germany onshore and off-shore up to the year 2020’, in International Conference on The Integration of Wind Power into the German Electricity Supply, Berlin, May 2005. [21] Short, J.A., Infield, D.G., Freris L.L. ‘Stabilization of Grid Frequency through Dynamic Demand Control’ IEEE Transactions on Power Systems, 22, 3, August 2007. [22] Kundur, P. Power System Stability and Control, McGraw-Hill, Maidenhead.
Slide 113: Power Balance / Frequency Control 95 [23] Tilley, A.R. and Bull, R.N. ‘The Zebra electric vehicle battery – recent advances’, in Proceedings of the Autotech ’97 Conference, NEC, Birmingham, UK, 4–6 November 1997. Other Useful Reading Barton, J.P. and Infield, D.G. ‘Energy Storage and its Use with Intermittent Renewable Energy’ IEEE? Hartnell, G. ‘Wind on the system-grid integration of wind power ’, Renewable Energy World, March–April 2000, 3(2). Holt, J.S., Milborrow, D.J. and Thorpe, A. ‘Assessment of the impact of wind energy on the CEGB system’, CEC Brussels, Contract No. EN3W-0058-UK. Patterson, W. Transforming Electricity, Earthscan, 1999. This book gives a readable, informative and entertaining account of the development of power systems over the past century. ‘Wind power in the UK’, Report by the Sustainable Development Commission, UK, May 2005.
Slide 115: 4 Electrical Power Generation and Conditioning 4.1 The Conversion of Renewable Energy into Electrical Form Renewable energy is available in a variety of forms. Uniquely, biomass is available in the form of combustible fuel and thus can play a similar role in generation as conventional fossil fuels. To generate electricity, all other renewables require a number of conversion stages that differ from those found in systems based on traditional fuels. • • • • • Wind energy is available in kinetic form. The function of a wind turbine is to extract energy from the intercepted wind by slowing it down and to convert this energy into a mechanical form that suits an electric generator. To improve the efficiency of conversion and for other operational reasons, the generator may be interfaced to the mains through a power electronic converter. Water power in the form of tides (potential) or water flow (kinetic) requires a turbine to transform this energy into rotational form for further conversion into electricity, again through a generator. The kinetic energy in water currents caused by tidal effects can alternatively be captured through an underwater ‘wind turbine’ that uses the same technology as those on land. Wave energy conversion requires specially designed devices that transform the low frequency energy in the waves into (usually) pressure energy in oil, which in turn drives an electrical generator. In other wave energy concepts the rise and fall of waves drives air through a turbine coupled to an electrical generator. Solar energy is available as radiation ranging from ultraviolet to infrared. Conversion into electricity can be implemented thermally by solar furnaces that raise steam to drive conventional turbines or through solid state photovoltaic devices that utilize the radiation to separate charges in semiconductor junctions. The efficient operation of PV based systems depends on interfacing the PV array to the grid through a power electronic converter. Renewable Energy in Power Systems Leon Freris and David Infield © 2008 John Wiley & Sons, Ltd
Slide 116: 98 Renewable Energy in Power Systems This summary indicates that all renewables, except PV systems, rely on electromechanical generators for the final stage of conversion from mechanical into electrical energy. This chapter introduces the principles of operation of two classes of electrical generators, the ‘synchronous’ and the ‘asynchronous’ types, both used extensively in RE applications. Additionally, this chapter deals briefly with the principle of operation of the transformer, a ubiquitous device in multivoltage level power systems. Understanding the operation of the transformer is a necessary prerequisite for the study of the ‘asynchronous’ type of generator. Power electronics plays a vital role in PV and an increasingly important role in the wind power area. A review of power electronic devices and the converters based on them is covered in the penultimate section. Finally, the chapter concludes with a description of how electromechanical and/or power electronic converters are used in PV and wind systems. In what follows a symbol written in regular type indicates that the parameter is a scalar while bold type is used if it is a vector, phasor or a complex number. 4.2 The Synchronous Generator 4.2.1 Construction and Mode of Operation In an electrical generator, mechanical input power is converted into electrical output power. To get an appreciation of how this energy conversion process is carried out it is useful to look briefly at the underpinning physics. Faraday’s law of electromagnetic induction [1] states that a conductor of length l (m) moving with a velocity u (m/s) through a magnetic field of uniform flux density B (Tesla), l, u and B being mutually perpendicular, will experience an induced voltage across it given by v = Blu ( volt ) (4.1) Nature is such that this mechanical (u) to electrical (v) conversion process described by Equation (4.1) is ‘mediated’ through the presence of a magnetic field (B). The equation shows that to generate a high and therefore useful voltage it is necessary to have a high magnetic flux density, a long conductor length and as high a conductor velocity as possible. All the above requirements are particularly well satisfied if the mechanical–magnetic–electrical structure of the generator is arranged as a rotating rather than a linear one. In practice it has been found that it is preferable to have the conductors stationary and move the source of the magnetic field. With a stationary set of conductors the problems of insulation and electrical connections are eased and centrifugal forces on the main windings are absent. Figure 4.1(a) shows in outline an AC generator also known as an alternator or synchronous generator. Here the conductors that form a winding, known as the stator, are stationary and the source of magnetic flux rotates. The source of flux is a rotor with poles marked north and south that carries a field winding as shown in the figure. The field winding is fed or excited from an external DC source through sliding contacts known as slip-rings. An additional reason why it is preferable to arrange the main windings to be stationary is that the DC power associated with the field is a small fraction of the power delivered by the stator winding. In fact, in some designs the field winding can be dispensed with completely and replaced by a
Slide 117: Electrical Power Generation and Conditioning 99 v N dc in Axis of rotation ac out (a) (b) 0 t S Figure 4.1 A two-pole synchronous generator. (Reproduced from Reference [1] with permission of John Wiley & Sons, Inc.) rotor built with permanent magnets. There are both advantages and disadvantages to each type of excitation which will be explored later. The flux density produced by the rotor in Figure 4.1(a) is maximum positive upwards along the pole axis, zero at 90 ° to the pole axis and maximum negative at 180 °. Equation (4.1) indicates that a variable voltage with polarity reversals is generated when the field winding rotates at constant angular velocity because the magnetic field cutting each conductor increases to a maximum, decreases and successively reverses in each revolution. By proper spatial distribution of the stator winding turns and shaping of the pole faces, the generated voltage across the stator terminals in Figure 4.1(a) can be made to approach the sinusoid waveform shown in Figure 4.1(b). It should be evident that a full rotation of the rotor will result in a full complete cycle of the sinusoidal waveform. Hence the frequency of the generated voltage in Hz (cycles per second) is identical to the angular velocity of the rotor in revolutions per second. Appendix A explains that three-phase AC systems are universally used for the generation, transmission and utilization of electrical energy. One of the reasons is that synchronous generators are particularly well suited for the generation of three-phase voltages. When wound for three phases, alternators make optimal use of the iron that carries the magnetic flux and of copper that carries the electric current. Figure 4.2(a) shows three separate one turn windings, where a, b and c indicate the beginnings and a′, b′ and c′ the ends of these windings. The winding axes are shifted in space with respect to each other by 120 °. As a consequence the voltages generated by the rotating field are also shifted in the time domain by one third of a period, thus forming the three-phase system of AC supply shown in Figure 4.2(b). A more effective arrangement for superior power output from a three-phase stator winding is shown in Figure 4.3. Here the winding is embedded and distributed in slots in an iron cylinder. The magnetic flux φ and therefore the flux density B generated by the rotating electromagnet is approximately proportional to the magnetomotive force (mmf) F (ampere turns) given by F = N If (4.2) where N is the number of turns of the field winding (shown as having one turn in the figure) and If is the field or excitation current. In a practical machine the air gap between the rotor pole faces and the internal surface of the cylinder is made small, and the cylinder itself is made of a ferromagnetic material so that minimum resistance is offered to the flow of magnetic flux. A large radial magnetic flux can therefore be produced from a moderate If.
Slide 118: 100 Renewable Energy in Power Systems c b´ v Va´a Vb´b Vc´c N a´ a 0 120° 240° 360° S b (a) c´ (b) Figure 4.2 A primitive three-phase synchronous generator a b´ N ω If c S b c´ a´ Figure 4.3 A practical three-phase synchronous generator The peripheral velocity with which the conductors in the embedded windings cut the magnetic flux is, of course, proportional to the angular velocity ω. Using Equations (4.1) and (4.2), the induced voltage V generated in a stator winding for a rotating generator is therefore given approximately by: V = kω I f (4.3) where k is a constant of proportionality that depends on the number of turns in each winding, the distribution of the conductors in the slots, the length of the air gap and the general geometry and magnetic properties of the iron that carries the magnetic flux. Electrical machine designers take all of these factors into account when designing synchronous generators so
Slide 119: Electrical Power Generation and Conditioning 101 that the mechanical input power is converted into electrical power with maximum efficiency and with minimum cost of materials. Up to this point the alternator in standalone mode has been considered. The rotor driven by a prime mover induces a set of balanced three-phase voltages in the stator windings, as shown in Figure 4.2(b). If a balanced three-phase load is connected across the windings, a balanced set of currents will transfer power from the prime mover to the load. This standalone operation is the exception rather than the rule in power generation. Rather, the alternator is most likely to be required to inject power into a grid that to all intents and purposes may be considered an infinite bus. This mode of operation is discussed in the following sections. 4.2.2 The Rotating Magnetic Field An examination will first be made of what happens when the primitive stator windings of Figure 4.2(a) are connected to a three-phase supply and the rotor is absent. Figure 4.4(a) shows the balanced currents that will be drawn by the three primitive windings which are redrawn in Figure 4.4(b) [1]. When stator winding aa′ carries the sinusoidal current ia, Equation (4.2) tells us that ia will generate a sinusoidally pulsating space vector field Fa and therefore a flux φa along the axis of coil aa′ in Figure 4.4(b). Similarly, the remaining two currents ib and ic in Figure 4.4(a) will generate pulsating fields and fluxes Fb, φb and Fc, φc along the axes of coils bb′ and cc′ respectively. Positive currents in the windings flowing into the ⊗ conductor and returning through the conductor in each (a) (b) (c) Figure 4.4 Production of a rotating magnetic field. (Reproduced from Reference [1] with permission of John Wiley & Sons, Inc.)
Slide 120: 102 Renewable Energy in Power Systems Figure 4.5 A one- and two-pole pair wound stators phase produce mmfs Fa, Fb and Fc in the directions indicated in Figure 4.4(b). At the instant ωt = 0 ° in Figure 4.4(a), ia is a positive maximum and ib and ic are negative and one-half maximum. At this instant, the flux contributions in space can now be added as shown in Figure 4.4(c) for ωt = 0 °. The stator flux Φs is the space vector sum of the three flux contributions. At ωt = 30 °, the relative magnitudes of the three currents have changed and the position of the resulting stator flux has shifted anticlockwise by 30 °. At ωt = 60 °, the current magnitudes have changed again and the resultant stator flux vector has shifted another 30 °. Over a complete cycle of mains frequency the vector Φs would have completed one revolution. It follows that for a 50 Hz supply the flux vector will complete one revolution in onefiftieth of a second; i.e. it will complete 50 revolutions per second. It is said that a rotating magnetic field (RMF) has been created by virtue of the space distribution of the three windings and of the time distribution of the currents in the three windings. This is a very important concept on which the operation of three-phase generators (and motors) depends. To summarize, if a stationary observer were to position him- or herself inside the stator cylindrical space of Figure 4.3, he or she will observe a north–south pole pair rotating at 50 × 60 = 3000 revolutions per minute for a 50 Hz supply and 60 × 60 = 3600 revolutions per minute for a 60 Hz mains. These are known as the synchronous speeds for 50 and 60 Hz systems for a one-pole pair winding. Stator windings can be arranged so that not just one pair but several pairs of poles can be created in the interior of the stator cavity. For example, a two-pole pair arrangement will have two separate windings for each phase, each winding side occupying 30 rather than the 60 ° shown in Figure 4.3. It can be shown that with such an arrangement in one-fiftieth of a second the rotating magnetic field in a 50 Hz system will advance by 180 rather than 360 °. The magnetic field produced by an alternator wound for one- and two-pole pairs is shown in Figure 4.5 (a) and (b) respectively. This property is used by designers of electrical machines to generate rotating magnetic fields that rotate at submultiples of the figures shown above. The reason why such machines are useful will be explained in the following section. The general relationship linking the synchronous speed in radians per second ωs to frequency f and number of pole pairs p is: ωs = 2π f p (4.4)
Slide 121: Electrical Power Generation and Conditioning 103 Table 4.1 Relationship between pole pairs and synchronous speed in rev/min p 1 2 3 4 200 Rev/min (50 Hz) 3000 1500 1000 750 15 Rev/min (60 Hz) 3600 1800 1200 900 18 Figure 4.6 Rotor configurations for a one- and two-pole pair wound stators Table 4.1 shows the relationship between the synchronous speed in rev/min and the number of pole pairs p for 50 and 60 Hz systems. 4.2.3 Synchronous Generator Operation when Grid-Connected Having established the presence of a rotating magnetic field within the cylindrical interior of the stator, the rotor is now ready to be incorporated. Figure 4.6 shows the salient rotor configurations for a one- and two-pole pair wound stators corresponding to the flux patterns in Figure 4.5. A one-pole pair rotor will be considered. The rotor field current is switched on so that a north–south pole pair is created. With the rotor stationary, the interaction between the rotating magnetic field (RMF) and the rotor field is to say the least unproductive. As the RMF sweeps by at 50 rev/s or 20 ms per revolution, the only effect experienced by the rotor body is a pulsating torque as the RMF pole pair approaches and then overtakes the rotor pole pair. Due to the rotor inertia it is just not possible for the rotor to accelerate and lock on to the RMF within the required milliseconds. Consider now that some external torque is applied to drive the rotor at increasing speed until the synchronous speed is very nearly reached. If the speed difference is close enough and decreasing, at some point the rotor N–S pair will lock on to the RMF S–N pair and the two magnet systems will rotate in synchronism. With the external torque now removed and
Slide 122: 104 Renewable Energy in Power Systems assuming a lossless environment with no friction and windage, the rotor will rotate at synchronous speed with no requirement that any power is fed to it either mechanically from the shaft or electrically from the power network to which the synchronous machine is connected. The system is idling and can be thought as being suspended between motoring and generating. In the synchronous machine the speed of the locked magnet pair is fixed because the infinite bus frequency and therefore the RMF speed is fixed. However, the relative angular disposition of the magnetic axes of the two locked pairs is not fixed and is the mechanism that regulates the direction of energy conversion. If an external braking torque Qt is applied to the shaft the rotor will keep rotating at the synchronous angular speed ωs, but its magnetic axis will fall back or spatially lag the magnetic axis of the RMF. The tangential magnetic tension forces caused by the misalignment of the magnet pair is at heart the mechanism of electromechanical energy conversion. The mechanical power Pm given by Pm = Qt w s (4.5) is extracted from the shaft; therefore an equal amount of electrical power must be supplied from the electrical system to which the machine is connected if the conservation of energy principle is to be satisfied. The synchronous machine is now motoring. If the external torque is accelerating rather than decelerating the magnetic axis of the rotor is advanced with respect to the RMF axis, mechanical power is now supplied to the shaft and the energy conservation principle demands that an equal amount of power is fed into the electrical grid system. The synchronous machine is now generating. In renewable energy applications, a high number of pole pairs may be selected for the generator as this requires a low rotational shaft speed to generate mains frequency voltages. Hydro turbines and wind turbines are in this class. The multipole arrangement is particularly desirable when a wind turbine is coupled directly to a synchronous machine without an intervening gearbox. In such cases, if a frequency of 50 Hz is to be generated from a wind turbine rotating at 15 rev/min, Table 4.1 indicates that two hundred pole pairs would be necessary! This power generation process will now be looked at from the perspective of the electrical network to which the generator is connected. 4.2.4 The Synchronous Generator Equivalent Circuit To analyse the power flows in electrical systems, component representations are required that can be incorporated into network or circuit models. To use the available circuit analysis tools described in Chapter 5, it is necessary to build up these representations from basic circuit elements, namely: resistors, inductors, capacitors and voltage or current sources (Appendix A). Electrical power engineers over the years have developed a range of what are known as equivalent circuits for network simulation of electrical generators. Here, for an approximate steady state analysis of power flows a description is required of a synchronous generator by the simplest possible equivalent circuit. The Thévenin principle explained in Appendix A can be used, for example, to describe the behaviour of a DC battery by a source voltage in series with a resistance. Amazingly, this principle can also be effectively used to describe through a simple circuit, and to a good approximation, the behaviour
Slide 123: Electrical Power Generation and Conditioning IB X o If VA=VA δ QB Qt S 105 A B PB VB=VB 0 Figure 4.7 Equivalent circuit of a synchronous machine of a generator as complex as an alternator. The steps in this transformation are not given here but can be found in any book on electrical machines (for example see Reference [2]). In what follows it is assumed that the reader is familiar with the use of phasors to represent AC quantities. Readers not familiar with this concept should, at this stage, study the material in the Appendix. In Figure 4.7 the electrical generator has been reduced into a single-phase (the relationship with 3-phase is dealt with later) Thevenin equivalent circuit consisting of a voltage source VA = VA∠δ (the generated or ‘internal’ voltage of Equation (4.3)) and a source impedance Xs, known as the synchronous reactance. The synchronous reactance represents in one lumped element all the internal complex interactions between the rotor and stator magnetic fields, which are not of concern here. To maximize conversion efficiency, synchronous machines are designed to have as low winding resistance as possible; hence the source resistance representing the ohmic value of the stator winding is omitted here with little loss in accuracy. The equivalent circuit is shown connected to an infinite bus, i.e. a network of fixed frequency f and of fixed voltage VB = VB∠0 ° where its 0 ° angle defines it as the reference voltage. An investigation will explore how the two available external control parameters, namely the field current If and the shaft torque Qt, influence the synchronous machine and consequently the equivalent circuit behaviour. Equation (4.3) shows that |VA| depends on the field current, which is the source of the magnetic flux. It is also known that the angular disposition of the rotor magnetic axis depends on the direction and magnitude of the torque applied to the shaft. Angle δ (the load angle) is defined as the angle by which the axis of the rotor flux space vector that induces VA leads the axis of the net flux space vector in the machine that induces VB. The load angle in the spatial disposition of rotating vector fields is the same as that in the phasorial disposition of voltages in the equivalent circuit. An accelerating or ‘generating’ torque will result in a positive δ and in VA leading VB. A decelerating or ‘motoring’ torque will result in a negative δ and in VA lagging VB. 4.2.5 Power Transfer Equations There is interest in exploring the mechanism by which power is injected into the grid by a synchronous generator. This can be done by means of the concept of complex power developed in Appendix A. The grid connected synchronous generator of Figure 4.7 will be considered. The complex power at end B of the line is given by * SB = VB IB (4.6)
Slide 124: 106 Renewable Energy in Power Systems Then IB is expressed as a function of the line voltages using Kirchoff ’s voltage law, IB = VB − VA XS (4.7) Substituting Equation (4.7) into Equation (4.6) gives ⎛ V −V e ⎛ V − VA ⎞ SB = VB I* = VB ⎜ B * = VB ⎜ B A B ⎝ Xs ⎟ ⎠ ⎝ − j Xs SB = − Hence PB = − QB = VAVB sind Xs (4.8a) (4.8b) − jδ 2 VV VB ⎞ − j A B e − jδ = ⎟=j ⎠ Xs Xs VAVB ⎛V ⎞ sin δ + j ⎜ B (VB − VA cos δ )⎟ = PB + jQB ⎝ Xs ⎠ Xs (4.7a) VB (VB − VA cosδ ) Xs If the above analysis were to be carried out for terminals A in Figure 4.7, the results would be PA = − PB and QA = VA (VA − VB cosδ ) Xs (4.9b) (4.9a) Equation (4.9a) confirms the trivial fact that as the system is lossless, power PB coming out of terminals B is equal to power PA fed into terminals A. The scalar equations (4.8) and (4.9) are important in power systems technology as they describe the flow of active and reactive power of grid-connected synchronous generators. 4.2.6 Three-phase Equations The synchronous machine equations were derived without any reference to its three-phase nature. Assuming that the voltages in Equations (4.8) and (4.9) are phase to neutral in volts, the equations will give the single-phase active and reactive powers in watts and VAR respectively. If the voltages are in kV then the active and reactive powers – both functions of voltage squared – will be in MW and MVAR. In a balanced three-phase system the three-phase P and Q will be three times the per-phase P and Q. Applying this to Equation (4.8), P3ϕ = 3VAVB sin δ = Xs 3VA 3VB VV sin δ = Al Bl sin δ Xs Xs where VAl and VBl are line voltages.
Slide 125: Electrical Power Generation and Conditioning 107 Worked Example 4.1 A synchronous generator of 10 ohms synchronous reactance is supplying 5 MW and 2 MVAR to an 11 kV network. Calculate the generator internal voltage. Rearranging Equations (4.8a) and (4.8b) gives VA sinδ = and VA cosδ = VB − Therefore tanδ = XP V − XQ 2 B XP VB XQ VB In this case P = −5, Q = −2, VB = 11, X = 10. Hence δ = 19.52 °, and VA = 13.6 kV 4.2.7 Four-Quadrant Operation Consider the case where a lossless synchronous machine is run up to synchronous speed by an external prime mover. The field current is then gradually increased until the terminal voltage of the machine (the same as the internal generated voltage, as no current is taken) is made to be equal to the voltage of the local bus to which the machine is to be connected. Precise adjustments to the speed of the prime mover are made so that through some external instrumentation it is possible to detect an instant at which the internal voltage VA exactly matches VB in magnitude and phase. The synchronous machine can now be safely connected to the bus. This process is known as synchronisation and must be carried out each time a synchronous machine is to be connected to the mains. Now arrange for the prime mover to apply an accelerating torque. This will result in a positive load angle δ and according to Equation (4.8a) a negative active power, i.e. active power injected into the power system. As expected, the machine is generating. With a braking torque on the shaft, i.e. with the machine motoring, the load angle is negative and active power is supplied to the machine from the power system. Returning to the idling state, consider now what happens if the field current is increased so that VA is made larger than VB, but no external torque is applied so that δ and the active power are zero. Equation (4.8b) shows that the reactive power is negative; i.e. the synchronous machine injects reactive power into the system, and acts as a generator of reactive power. In this state the machine is said to be overexcited. Conversely, if the excitation current is decreased so that VA < VB, the reactive power flow is reversed, the machine is a consumer of reactive power and it is said to be underexcited. The synchronous machine is capable of operating in each of the four quadrants of the quadrant diagram shown in Figure A18 in the Appendix.
Slide 126: 108 Renewable Energy in Power Systems –PB PBmax δ = 90° δ Figure 4.8 Power angle characteristic of an SG 4.2.8 Power – Load Angle Characteristic: Stability Equation (4.8a) is plotted in Figure 4.8 to illustrate the dependence of the generated active power on the load angle. Note that a synchronous generator (SG) connected to an infinite bus is capable of generating a maximum active power PBmax, at δ = 90 °. Any additional applied mechanical torque will increase the load angle beyond 90 °, with a consequential decrease in electrical power. Physically, the peripheral magnetic forces linking the two fields together is insufficient to maintain the locking effect. The power balance between mechanical and electrical powers has now been lost, the excess mechanical power accelerates the rotor beyond synchronous speed and the synchronous generator has lost its ability to act as a stable power converter. In this unstable regime the SG is described as having lost synchronism i.e. its rotor generates an internal voltage VA of a higher frequency than the infinite bus voltage VB. Operation under this condition, known as pole slipping, results in large overcurrents, is highly undesirable and protection equipment will be brought into action to disconnect the synchronous machine from the mains. During system contingencies violent transient changes may take place that cause swings in the load angles of synchronous generators. To ensure safe stability margins so that pole slipping is prevented, the steady state load angles of SGs are kept well below 30 °. Power system stability is a major topic in its own right involving the simultaneous solution of the differential equations characterizing all the network components. This is beyond the scope of this book, but it is worth noting that connection of large RE sources on to the grid will alter the system dynamics. Therefore studies may be required to assess the new system stability margins. 4.3 The Transformer 4.3.1 Transformer Basics The transformer is an indispensable part of any power system operating at a range of voltages. The transformer mode of operation is included here as it provides a useful aid to the under-
Slide 127: Electrical Power Generation and Conditioning 109 Figure 4.9 The Transformer. (Reproduced from Reference [1] with permission of John Wiley & Sons, Inc.) standing of the mode of operation of induction generators which are used extensively in wind turbines. The properties of an inductance are discussed in the Appendix. A current carrying inductor, usually wound in the form of a coil, generates a magnetic flux. This flux links with the turns of the coil or the winding and induces a voltage in these turns if the current and therefore the flux is increased or decreased. It stands to reason that if a changing magnetic field produced by one inductor were to ‘link’ with the turns of wire in another adjacent inductor, it would induce a voltage in the unpowered inductor. This phenomenon is called mutual inductance and the transformer is a device constructed to exploit this effect. By having two inductors coupled together by a common magnetic field path, it is possible to transfer energy from one inductor circuit to the other. In order for this to work, the magnetic field has to be constantly changing in strength, otherwise no voltage will be induced in the unpowered winding. Thus, the transformer is essentially an AC device. The powered winding of a transformer is called the primary, while the unpowered winding is called the secondary. Figure 4.9 shows a transformer in outline [1]. The primary winding is connected to an AC supply V1. The winding is wound round a substantial closed ferromagnetic core. This provides a very low ‘magnetic resistance’ to the flow of magnetic flux in comparison to the surrounding air. Because of the very low resistance to the flux flow, only a very low current, known as the magnetizing current, is required from the supply to set up an mmf that circulates a flux which induces a voltage in the primary that exactly balances V1. For the present the effect of this low current may be disregarded. Because of the low ‘magnetic resistance’ of the ferromagnetic core it can be assumed that practically all of the generated magnetic flux is constrained to flow within the core with only a small amount of leakage flux taking paths in the surrounding air, which will also be disregarded at this point. The consequence of this is that the secondary winding which is also wound round the transformer core, links exactly the same flux as the primary winding. Finally, if it is assumed that the windings consisted of large cross-section copper wire their ohmic resistance could be disregarded. All these simplifications lead to the concept of the ideal transformer. In an ideal transformer the voltage induced in an inductor of N turns linking a flux varying at dφ/dt is proportional to N. Hence for the transformer in Figure 4.9, we can write
Slide 128: 110 Renewable Energy in Power Systems V1 N1 = V2 N 2 (4.10) Suppose that the secondary is connected to a resistor that draws current I2 with the consequence that power V2I2 is extracted from the secondary. The energy conservation principle requires that this power is supplied from the source to which the primary is connected. It follows that a primary current I1 is established the value of which can be determined from: V1I1 = V2 I 2 and from Equation (4.10) I1 N 2 = I 2 N1 (4.12) (4.11) Note that an open circuit or a short circuit on the secondary winding of an ideal transformer are seen as an open or short circuit respectively on the primary side. An ideal transformer with identical primary and secondary windings would manifest equal voltage and current in both sets of windings. In a perfect world, transformers would transfer electrical power from primary to secondary as efficiently as though the load were directly connected to the primary power source, with no transformer there at all, but it will be found later that this ideal goal cannot be realized in practice. Nevertheless, transformers are highly efficient power transfer devices with no moving parts achieving efficiencies in the high nineties. The transformer has made long distance transmission of electric power a practical reality, as AC voltage can be ‘stepped up’ and current ‘stepped down’ for reduced ohmic resistance losses along power lines connecting generating stations with loads. A transformer that increases voltage from primary to secondary (more secondary winding turns than primary winding turns) is called a step-up transformer. Conversely, a transformer designed to do just the opposite is called a step-down transformer. 4.3.2 The Transformer Equivalent Circuit The next task is to develop an equivalent circuit capable of representing realistically the behaviour of a transformer in studies aimed at determining the flow of power in electrical systems. Figure 4.10 shows how such an equivalent circuit can be built up starting with the ideal transformer in the hatched box. The small but finite current required to set up the flux in the core is simulated by the presence of the shunt inductance Lm, the magnetizing inductance, which has a large value in henries. Next, the finite resistance of the two transformer windings can be simulated by the series resistors R1 and R2. To keep losses low, these have low numerical values. Finally, any magnetic flux not contained in the core is free to store and release energy rather than transfer it from one coil to the other. Any energy thus stored by this uncoupled flux manifests itself as an inductance in series with the relevant winding. This stray inductance is called leakage inductance and is represented in the equivalent circuit by L1 and L2 both having numerical values much smaller than Lm.
Slide 129: Electrical Power Generation and Conditioning 111 R1 L1 N1 N2 L2 R2 V1 Lm V2 Ideal transformer Figure 4.10 The transformer equivalent circuit Rt1 = R1 + R21 Xt1 = X1 + X21 V1 Xm Ideal transformer Figure 4.11 Simplified transformer equivalent circuit The equivalent circuit of Figure 4.10 can be further simplified if the R2 and I2 are transferred to the primary so that the primary and secondary resistances and inductances could be lumped into just two series components. This can be done through the conservation of energy principle. The transferred resistance from the secondary (2) to primary (1), which can be called R21, should have such a value that when it carries I1 should dissipate the same amount of 2 power as R2 when carrying I2, i.e. R2 I 2 = R21I1 . Substituting the number of turns for currents 2 from Equation (4.12) we get: N R21 = R2⎛ 1 ⎞ ⎜ ⎝ N2 ⎟ ⎠ 2 (4.13) Similar logic based on the reactive power conservation principle (Appendix A) can be applied for the transfer of the secondary winding reactance to the primary: N X 21 = X 2⎛ 1 ⎞ ⎜ ⎝ N2 ⎟ ⎠ 2 (4.14) A new equivalent circuit incorporating these changes is shown in Figure 4.11. Here Rt1 and Xt1 are the total winding resistance and reactance respectively referred to the primary winding. In this circuit the magnetizing reactance Xm is connected across the mains with little loss in accuracy because it can be shown that the voltage drops across Rt1 and Xt1 are small. This equivalent circuit is frequently used to calculate the effect of a transformer in a power network and will be used later to develop an equivalent circuit for the asynchronous generator.
Slide 130: 112 Renewable Energy in Power Systems 4.3.3 Further Details on Transformers The manufacturer of an electrical machine such as a transformer will indicate on the nameplate the normal operating conditions, e.g.: ‘11 000 : 415 V, 50 Hz, 500 kVA’. The rated output of 500 kVA can be maintained continuously without excessive heating and the consequential deterioration of the winding insulation. Because the heating is dependent on the square of the current, the output is rated in apparent power (kVA) rather than active power (kW). When supplying a zero power factor load, a transformer can be operating at rated temperature while delivering zero active power. On large transformers, taps on the windings allow small adjustments on the turns ratio. Often these taps are operated by an automatic tap changer that maintains the voltage, usually on the secondary, at a fixed value irrespective of the load on the transformer. It can be shown [1] that the size and therefore, weight and cost of a transformer are intimately related to the frequency of operation. The higher the frequency the lower are the weight and cost. For these reasons, in power electronic systems whenever a transformer is to be used, higher frequencies are employed, a topic to be revisited later. 4.4 The Asynchronous Generator 4.4.1 Construction and Properties Asynchronous or induction machines operating as motors are the most widely used electromechanical converters. In an induction machine the stator is identical to the one for synchronous machines shown in Figure 4.3 in which three-phase currents supplied to the stator produce a rotating magnetic field (RMF). The rotor, however, is radically different and it has neither an external magnetizing source nor permanent magnets. Instead, alternating currents are injected in the rotor from the stator through induction or transformer action – hence the useful parallel with the operation of a transformer. It is the interaction between these induced rotor currents and the stator RMF that results in torque production. In its most common form, the rotor consists of axial conductors shorted at the ends by circular rings to form a squirrel-cage or just cage, as shown in Figure 4.12. Although for the purposes of renewable energy sources there is interest in the generation mode, it is easier initially to understand the operation of the induction machine from the motoring perspective. As the stator RMF moves at ωs (given by Equation (4.4)) past the stationary rotor conductors, three-phase electromotive forces (EMFs) are induced in the spatially shifted rotor conductors by a flux cutting action. The resulting rotor currents, according to Lenz’s law [1], are of such magnitude and direction as to generate a torque that speeds up the rotor. If the rotor were to achieve speed ωs, there would be no change in flux linkage, no induced voltage, no current in the rotor conductors and therefore no torque. For EMFs to be induced in the rotor conductors they should possess some relative speed with respect to the stator RMF. For motoring, the rotor therefore turns at a lower speed ωr. It can be shown that the rotor currents produce an RMF whose speed depends on the frequency of these currents. For a constant torque interaction to take place, the rotor RMF must rotate in synchronism with the stator RMF, as in the case of the synchronous machine. How is this accomplished if the rotor rotates at a lower speed than ωs?
Slide 131: Electrical Power Generation and Conditioning 113 Figure 4.12 Induction machine with cage rotor (1, shaft; 2, cage rotor; 3, stator three-phase winding; 4, terminal box; 5, stator iron core; 6, cooling fan; 7, motor frame). (Reproduced with permission of Asea Brown Boveri Ltd) The difference between ωs and ωr is expressed as a ratio with respect to ωs and is known as slip s where s= Therefore (ω s − ω r ) ωs (4.15) ω r = (1 − s)ω s (4.16) The relative motion between the stator ’s and rotor ’s RMFs determines how frequently the stator RMF cuts the rotating rotor conductors, so the frequency of the rotor induced voltages and currents fr is fr = s f (4.17) where f is the mains frequency. The frequency of the rotor currents determines ωrr, the speed of the rotor RMF with respect to the rotor: ω rr = 2π fr p = 2π s f p = sω s (4.18) The speed of the rotor RMF with respect to the stationary stator is the rotor speed plus the rotor RMF’s speed with respect to the rotor: ω r + ω rr = (1 − s) ω s + s ω s = ω s It can be concluded that the rotor and stator RMFs rotate together at synchronous speed as required for a uniform torque to be developed as in the synchronous machine. However, in contrast to the synchronous machine, the rotor RMF is produced through induction from
Slide 132: 114 Renewable Energy in Power Systems the stator. The larger the applied braking torque the higher the slip, the larger are the induced EMFs and resulting rotor currents, and the stronger the interaction between the two RMFs to produce an electrical torque equal and opposite to the braking torque. The induction motor therefore exhibits a small decrease in speed with increments in braking torque. An ideal induction machine could be imagined to operate at zero slip. This is equivalent to the ‘idling’ state of the fixed speed synchronous machine. The vital difference, however, in the induction machine is that motoring or generating torques will be accompanied by a decrease or increase of speed below or above synchronous respectively. For generating, Equation (4.16) now gives a negative slip. The induction machine will move seamlessly from the motoring into the generation mode as the external torque changes from a decelerating to an accelerating type. Indeed, in small wind turbines, it is very common to find that the induction generator was originally designed as a motor and has been employed as a generator without any modification. Worked Example 4.2 A six-pole 50 Hz induction motor runs at 4% slip at a certain load. Calculate the synchronous speed, the rotor speed, the frequency of the rotor currents, the speed of the rotor RMF with respect to the rotor and the speed of the rotor RMF with respect to the stator. Model answer The synchronous speed from Equation (4.4) is Ns = f/p = 50/3 rev/s = 50 × 60/3 = 1000 rev/min The rotor speed from Equation (4.16) is (1 − s)Ns = (1 − 0.04) × 1000 = 960 rev/s The frequency of the rotor currents are: fr = sf = 0.04 × 50 = 2 Hz The speed of the rotor RMF with respect to the rotor: N rr = fr × 60 p = 2 × 60 3 = 40 rev min The speed of the rotor RMF with respect to the stator: N r + N rr = 960 + 40 = 1000 (i.e. the rotor and stator RMFs rotate together) 4.4.2 The Induction Machine Equivalent Circuit The induction machine can be viewed as a transformer with a rotating secondary. Imagine an induction machine with its rotor mechanically locked, i.e. at standstill. The stator RMF will be rotating at ωs with respect to the rotor and inducing in each phase the voltage E2 at mains frequency f. The current that flows in each phase will be I2 = E2 E2 = R2 + j 2π fL2 R2 + jX 2 (4.19) where R2 and L2 are the effective per-phase resistance and inductance of the rotor winding and X2 is the rotor reactance at mains frequency. At standstill the slip s = 1 and the rotor voltages and currents are of the stator frequency f. At any other rotor speed, the slip is s, the
Slide 133: Electrical Power Generation and Conditioning 115 I2 R2 jX2 E2 R2 1 – S ––––– S Figure 4.13 Equivalent circuit of induction machine rotor R1 X1 Air gap R2 X2 V1 E1 E2 R2 1 – S ––––– S Ideal transformer Figure 4.14 Induction machine stator-rotor equivalent circuit induced voltage is sE2, the rotor frequency is sf and the rotor reactance is sX2. For the rotor current at slip s we can write the more general expression: I2 = sE2 E2 E2 = = R2 + jsX 2 ( R2 s) + jX 2 R2 + R2[(1 − s) s ]+ jX 2 (4.20) since R/s = R2 + R2[(1 − s)/s]. Equation (4.20) provides the rational for the equivalent circuit shown in Figure 4.13. This circuit resembles that of the secondary winding of the transformer (Figure 4.10) but with a variable resistive load connected to its output. The energy conservation principle indicates * that the electrical power transferred to the rotor is the real part of E2 I2 . The power lost irre2 2 versibly in the rotor ohmic resistance is R2 I 2 and the remainder, i.e. I 2 R2[(1− s) s ], must be and indeed is the electrical power converted into mechanical power. The transformer equivalent circuit analogy is extended in Figure 4.14 to include the stator parameters. Here R1 and X1 represent the stator winding resistance and inductance while Xm represents the magnetizing reactance drawing the current necessary to establish the RMF. The dashed line corresponds to the air gap interface across which energy is transferred from the stator to the rotor. In a further simplification the ideal transformer can be omitted by transferring elements from secondary to primary using the transformation ratio. In Figure 4.15 Rs and Xs are the stator winding resistance and reactance respectively. The elements Rr and Xr represent the rotor resistance and reactance respectively referred to the stator using the rotor–stator transformation ratio. The product Rr[(1 − s) s ]I s2 = Rem represents the electrical power per-phase converted into mechanical power. This equivalent circuit tells us that when the rotor is locked, s = 1, Rr[(1 − s)/s] = 0, so all the input to the rotor is converted into heat in Rr. When s < 1, the energy into the rotor is partly converted into heat in the winding resistances and partly into mechanical form.
Slide 134: 116 Renewable Energy in Power Systems Air gap Rs jXs Is Rr jXr Vs jXm (1 – S)Rr/S Figure 4.15 Induction machine equivalent circuit referred to the stator With the induction machine generating, the slip is negative and the notional resistance Rr[(1 − s)/s] is also negative. This is perfectly consistent with circuit analysis. A positive RI2 implies irreversible conversion of electrical energy into thermal energy. A negative RI2 implies the conversion of some other type of energy (in this case mechanical) into electrical. Irrespective of whether the machine is motoring or generating the mains to which the machine is connected supplies the reactive voltamperes absorbed by all the inductive components of the equivalent circuit. This must be so as only positive or negative real power is associated with the mechanical/electrical energy conversion in the resistance Rr(1 − s)/s. The consequence is that induction generators always absorb reactive power from the mains. 4.4.3 The Induction Machine Efficiency If the total electrical input power per phase fed into the stator is Ps, the power crossing the motor’s air gap i.e. the power per phase transferred from the stator to the rotor is Pr = Ps − Rs I s2 . All of Pr is dissipated in { Rr + Rr[(1 − s) s ]} = R r s so, Pr = ( Rr s) I r2 . Hence the rotor copper loss is: Rr I r2 = sPr (4.21) Subtracting the rotor copper loss from Pr gives the average per-phase mechanical power Pm = Pr − Rr I r2 which through substitution from Equation (4.21) gives Pm = (1 − s) Pr (4.22) The developed torque Qm of the motor is its total mechanical power 3Pm divided by the motor shaft speed ωr. Therefore i Qm = 3Pm ωr (4.23) Substituting Equations (4.16) and (4.22) into (4.23) gives Qm = 3(1 − s) Pr Pag = (1 − s)ω s ω s (4.24) where Pag = 3Pr is the total three-phase power crossing the air gap. Neglecting the stator copper losses and rotating mechanical losses, the efficiency of an induction motor is given approximately by:
Slide 135: Electrical Power Generation and Conditioning 117 η= Pout Pm (1 − s) Pr = = = (1 − s) Pin Pr Pr (4.25) For an induction generator (s negative) the power flow is in the reverse direction hence approximately η= Pout Pr 1 = = Pin Pm (1 − s) (4.26) Equations (4.25) and (4.26) indicate that for the conversion efficiency to be high, s at full load must be as small as possible. Real induction generators have losses that have not been taken into account in this simplified analysis. The mechanical power available to produce electricity is reduced by windage and other mechanical frictional losses within the generator. Additionally, electrical and magnetic losses within the rotor reduce the power that is transferred from the rotor across the air gap to the stator. Finally, in the stator there are more losses associated with the winding resistance and the setting up of the magnetic excitation in the shunt branch of the equivalent circuit. As a consequence, large induction generators have efficiencies in the region of 90%. These extra losses will be referred to in a later section. 4.4.4 The Induction Machine Speed–Torque Characteristic An important characteristic of any electromechanical converter is its speed–torque relationship. For the induction machine the developed torque Qm from Equations (4.23), (4.22) and (4.16) is Qm = Substituting for Pr = Rr 2 Rr Vs2 Is = s sZ 2 3Pm 3(1 − s) Pr 3Pr = = (1 − s)ω s ω s ωr and using the equivalent circuit of Figure 4.15 we get ⎤ R I 2 3R ⎡ Vs2 Qm = 3⎛ r ⎞ s = r ⎢ ⎝ s ⎠ ω s sω s ⎣ ( Rs + Rr s)2 + ( Xs + X r )2 ⎥ ⎦ (4.27) This relationship is nonlinear and can be generalized to describe typical performances of an induction machine by normalizing it in terms of torque and speed. Taking as normal torque the rated torque and as normal speed the synchronous speed, the normalized relationship of Equation (4.27) for a typical induction generator is plotted in Figure 4.16 with Rr as a parameter. Note that, as for the synchronous machine, there is a maximum or pullout torque beyond which the generator will accelerate uncontrollably. However this condition is far away from the normal operating regime. The curve for rotor resistance ‘Rr’ represents the performance of a typical induction machine with low rotor resistance and shows that the variation in speed from zero input torque to rated torque varies by about 3–4%. For s small, Rr /s is large compared to Rr and Xs + Xr and to a good approximation Equation (4.27) can be written as
Slide 136: 118 ω — ωs Renewable Energy in Power Systems Slip 2.0 –1.0 1.8 Rated torque 1.6 Max torque –0.8 –0.6 1.4 Rr˝>Rr´ –0.4 Rr´>Rr 1.2 Rr 1 0 –0.2 0.5 1.0 1.5 2.0 0 2.5 Torque Q/QR Figure 4.16 Normalized speed and slip against torque for a generator Qm ≈ 3Vs2 s ω s Rr (4.28) Equation (4.28) indicates that in the normal operating range (zero to rated torque), torque is directly proportional to slip and therefore speed and is inversely proportional to Rr. Curves for Rr ′ > Rr and Rr ′′ > Rr are also plotted on Figure 4.16. By selecting the value of the rotor resistance, a designer has the ability to change the slope of the torque–speed characteristic. If a substantial variation of speed with torque is required, the rotor can be designed to have a large resistance. The downside of such an arrangement is the unacceptable reduction in efficiency. A method to access the rotor windings and therefore exploit the property of speed change is to arrange a rotor that has coils rather than short circuited bars, with the coil terminals connected to slip rings and brushes so that additional external resistance can be connected in series with the windings. In such a rotor wound induction machine, the rotor winding is similar to that on the stator. The disadvantages of the wound rotor induction generator include a higher capital cost and a higher maintenance cost. Worked Example 4.3 A wind turbine rated at 450 kW has the following induction generator parameters in ohms: Rs = 0.01, Xs = Xr = 0.15, Rr = 0.01 and Xm = 6. At a time when it is supplying its rated output the slip is 0.01. Calculate the mains voltage and the power factor at which the induction generator is supplying power to the grid using the simplified equivalent circuit.
Slide 137: Electrical Power Generation and Conditioning 119 A 0.01+0.01 0.15+0.15 B Rem = (1-s) Rr s (1+0.01) 0.01 = –0.01 ~ –1 6 Figure 4.17 Equivalent circuit for Worked Example 4.3 Model answer The circuit in Figure 4.17 brings to mind the synchronous generator equivalent circuit of Figure 4.7. The reactance to resistance ratio of the series impedance in figure 4.17 is 0.3/0.02=15; hence, to a good approximation, Equations (4.8) and (4.9) can be used. The mechanical power converted from the wind into electrical power per-phase is 450 000/3 = 150 000 W. This appears as a ‘negative’ power dissipation in Rem therefore the voltage VA across Rem is given by 150 000 = and VA = 387 V Using Equation 4.8(a) 150 000 = and VB sin δ = 116.28 The reactive power at end A is zero, hence, from Equation (4.8b). 0= and VB cosδ = 387 Hence tan δ = 116.28 387 and δ = 16.7 ° VA 387 (VA − VB cos δ ) = (387 − VB cos δ ) X 0.3 VAVB 387VB sin δ = sin δ X 0.3 2 VA Rem giving VB = 404 V, i.e. 404 3 ≈ 700 line volts
Slide 138: 120 Renewable Energy in Power Systems The reactive power associated with end B of the line is given by Equation (4.9b); hence QB = 404 VB (VB − VA cos δ ) = ( 404 − 387 cos16.7) = 44874 VAR 0.3 X The Q absorbed by the shunt reactance Xm = 6 Ω is given from Qm = 4042 = 27202 6 72 076 = 0.9 150 000 The total Q per phase is = 44 874 + 27 202 = 72 076. Hence the power factor is cos tan −1 = 4.4.5 Induction Generator Reactive Power Worked example 4.3 shows that the induction generator is a source of active power but a sink of reactive power. Even when the real power output from an induction generator is zero it will still draw considerable reactive power through Xm to magnetize its iron core (3 × 27.2 kVAR in the worked example). As increasing torque is applied and more real power is exported to the network, extra reactive power is absorbed due to the reactive power consumed by the series reactance (3 × 44.87 kVAR in the example). A typical relationship between active and reactive power for an induction generator is shown in Figure 4.18. The induction generator power factor will vary from zero at A to around 0.9 at B. In order to improve the power factor it is often necessary to fit local power factor correction (PFC) capacitors at the generator terminals (Appendix). These have the effect of shifting the overall characteristic downwards to A′ B′. The amount of reactive power ‘compensation’ required depends on a number of technical and economic factors. Q Import (MVAR) B B´ A A´ 0 Effect of PFC P export (MW) Figure 4.18 Relationship between active and reactive power for an induction generator
Slide 139: Electrical Power Generation and Conditioning 121 Table 4.2 Comparison between synchronous and asynchronous generators Induction generator Synchronous generator • • • • Features Efficient • Expensive • Requires maintenance • Reactive power flow can be controlled through • field current • Fixed speed hence very stiff • Moderately efficient Less expensive Rugged and robust, little maintenance Sink of reactive power • Responds in an oscillatory manner to sudden changes in torque • Can be built with permanent magnets for a large number of pole pairs and low rotational speed (‘ring’ form) • Suitable for variable speed operation through a power electronic interface • Suitable for connection to weak networks. Used in autonomous systems • Requires special synchronization equipment to connect to mains Small change in speed with torque, hence more compliant • Responds to sudden torque inputs in a nonoscillatory way • Cannot be built economically for low rotational speeds • Suitable for variable speed in its rotor wound form in conjunction with a power electronic converter • Suitable for week networks only in conjunction with power electronics • Can be simply synchronized to the mains Use in RE generation • Used in wind power mainly in its ‘ring’ form • Used extensively with a gear box in wind for gearless coupling to a wind turbine power • Variable speed provided through a DC link • Variable speed provided with power power electronic interface electronics in the rotor wound form • Used in water power when reactive power • Used in water power with gearbox control is required 4.4.6 Comparison between Synchronous and Asynchronous Generators Table 4.2 provides an overall comparison of the characteristics of induction and synchronous generators with particular reference to renewable energy applications. 4.5 Power Electronics 4.5.1 Introduction Power electronics is concerned with the application of electronic devices to control and condition electrical power. Power control involves the regulation of the power transfer from the renewable energy generator to the mains either to maximize this transfer continuously as the available resource changes or to limit the transfer for operational reasons. Power conditioning involves the transformation of power from one voltage/current/frequency/waveform to a different voltage/current/frequency/waveform.
Slide 140: 122 Renewable Energy in Power Systems A power electronic interface that controls and/or conditions power is referred to as a converter. Converters are interposed between the RE generator and the mains and often carry the total power transfer. They should be designed so that as little power as possible is lost in this transfer and that their capital cost is as slow as possible per watt transferred. Transformers can be viewed as power conditioning devices. Converters, not unlike transformers, are rated in volt-amps (VA). The devices (transistors, etc.) used in power electronic converters are made from semiconductor silicon, like those in a computer processor, but physically they are much larger, and usually they are discrete – one device per piece of silicon. A converter will typically contain between one and a hundred power semiconductor devices. Power electronic converters are found in an enormous range of applications and sizes, from the 2 GW UK-France Channel DC Link down to the 40 W energy-saving light bulbs in homes. In renewable energy, power electronic converters are already used in most PV systems and many wind turbines. It is also expected that they will be used in the future in practically all the emerging wave and tidal power technologies. Despite the enormous range of converter sizes and applications, they are mostly based on just a handful of device types and basic circuit configurations. This section will discuss the most common of these. The aim is not that the reader will be able to design converters, but that he or she will understand their specifications, potential uses and limitations. The following section will deal with the application of these converters in the conditioning of power from renewable energy sources. More details on solid state switching devices and converters can be found in many books on power electronics (see, for example, Reference. [3]). Some of the most common devices are reviewed below. 4.5.2 Power Semiconductor Devices Diodes The diode in Figure 4.19 conducts current from anode to cathode, (in the direction of the arrow) whenever the anode is made positive with respect to the cathode, i.e., positive bias voltage is applied across the device, and blocks current in the reverse direction when the device is negatively biased. The diode has no capability to control the current after it has been established. Diodes are readily available with ratings up to thousands of volts and thousands of amps. They are very reliable and very cheap (per VA) and very useful. I Anode Cathode V Figure 4.19 Diode
Slide 141: Electrical Power Generation and Conditioning 123 I Anode Gate Cathode Figure 4.20 Thyristor Collector Base Gate Emitter Drain Gate Source Collector Emitter BJT MOSFET IGBT Figure 4.21 Transistor types Thyristors In contrast to the diode, the thyristor (Figure 4.20) has a control capability through a third electrode known as the gate. Even if the thyristor is positively biased no current will flow through the main anode–cathode circuit until a small pulse has been applied to the gate. Thereafter, the thyrstor acts like the diode so that the current will continue to flow through the anode–cathode circuit until that current reduces to zero. The thyristor itself cannot reduce or switch off the current. Unlike the diode, however, after the current has reached zero, the thyristor regains its capability to block current in the forward direction. Thyristors were the first commercially available controllable power semiconductor devices and were used in converters of all sizes. They are still the cheapest controllable devices and are used in very high power converters (hundreds of MVA) and in smaller cost-sensitive applications. Gate turn-off thyristors (GTOs) and integrated-gate commutated thyristors (IGCTs) are devices based on the thyristor principle but capable of turning themselves both on and off. These derivatives are expensive and are being superseded by advances in transistor technology. Transistors Bipolar junction transistors (BJTs), metal oxide semiconductor field effect transistors (MOSFETs), insulated gate bipolar transistors (IGBTs) (Figure 4.21) are all significantly more expensive than simple thyristors, but have the big advantage that they can be turned off by a control signal. Practical and economic considerations currently favour the use of MOSFETs for small converters up to about 30 kVA, which includes most PV inverters. IGBTs are used in converters up to about 10 MVA, which includes wind turbine applications.
Slide 142: 124 Renewable Energy in Power Systems il is Supply vs vl Load Figure 4.22 Circuit of a single-phase diode bridge rectifier vs t vl il is Figure 4.23 Waveforms in a diode bridge rectifier connected to a purely resistive load 4.5.3 Diode Bridge Rectifier In Figure 4.22 alternating current power is fed from the supply to the diode junctions. It is rectified by the four diodes that form the bridge, and the resulting DC power is supplied to the load on the right. Thus, the circuit converts the AC power to DC. Assuming, for simplicity, that the load is purely resistive, the resulting waveforms are shown in Figure 4.23. The voltage supplied to the load is DC, in that it is always positive, but it is far from smooth because it contains a considerable ripple. Very few practical loads would tolerate a supply like this except those, e.g. heating resistors, that would be equally happy on an AC supply. For this reason, it is common to add a smoothing capacitor as shown in Figure 4.24. The waveforms in Figure 4.25 show that, as before, the diode bridge is rectifying AC power to DC. Now, however, the voltage supplied to the load vl is almost constant; a bigger capacitor would reduce the ripple even further. The capacitor acts as an energy store so that when the output voltage from the rectifier is lower than the voltage across the capacitor, energy is supplied to the load from the capacitor. When the output voltage from the rectifier is larger than the capacitor voltage the capacitor is charged from the supply. This charging effect is provided
Slide 143: Electrical Power Generation and Conditioning 125 Smoothing capacitor Supply Supply impedance ib is vl Load vs Other customers vc Figure 4.24 Circuit of a diode bridge rectifier with a smoothing capacitor and a supply impedance vs t vl ib is vc Figure 4.25 Waveforms of a diode bridge rectifier with a smoothing capacitor by a series of pulses shown as ib in the figure. Because the duration of the pulses is short, their amplitude must be high, if the same average power is to be supplied to the load as it was supplied before the capacitor insertion. The downside of this arrangement is that the current drawn from the AC source is is no longer a sine wave. Instead, it is a series of positive and negative pulses.
Slide 144: 126 Renewable Energy in Power Systems 4.5.4 Harmonics The nonsinusoidal current is in Figure 4.25 may cause problems for the AC supply network. In particular, transformers and cables in the network will experience additional heating. Despite this, the circuit shown in Figure 4.23 and variants with the same problem are very widely used, because they are cheap. TVs and computers are the biggest offenders, mainly because there are so many of them. Indeed, in areas where the load is dominated by TVs and computers, it is common for the customer ’s voltage waveform vc in Figure 4.25 to be flattopped, due to the nonsinusoidal voltage drop in the supply impedance. Plotting instantaneous supply current is against voltage vc would show a highly nonlinear relationship. Appliances such as TVs and computers are sometimes called nonlinear loads. Another way of describing the problem is by considering the harmonic content of the waveforms which can be obtained through Fourier analysis. For example, the supply current is in Figure 4.25 has a very large third harmonic component. This is a particular problem in three-phase systems (in just about every large power system in the world) because the third harmonics in the neutral conductor do not cancel (Appendix). Thus, the normal assumption that the neutral conductor carries zero current in a balanced three-phase system no longer holds. In general, all harmonic currents cause undesirable heating in the transformers and cables of the supply system. In a bid to prevent excessive harmonic currents in distribution networks, electricity utilities have instigated the introduction of regulatory standards that set maximum permissible levels for the harmonic currents that individual appliances may cause. In general, all power electronic converters cause some harmonic currents in the AC network to which they are connected. Converters, used in large high voltage DC (HVDC) transmission systems are particularly prone to this. Indeed, much of the engineering design of such converters concentrates on reducing the harmonic currents to an acceptable level. It is therefore understandable that electricity utilities have expressed concern regarding the suggestion that, in future, large numbers of inverters will be used to connect PV and other renewable energy sources to distribution networks. Fortunately, the converters used in modern renewable energy systems use techniques that reduce the low order harmonics to negligible levels. In the context of a network supplying a typical collection of TVs and computers, etc., the harmonic contribution from renewable energy systems can be expected to be negligible. 4.5.5 The Thyristor Bridge Converter Replacing the diodes in the bridge of Figure 4.22 with thyristors, as shown in Figure 4.26, allows the DC voltage to be controlled. In many applications, the anode–cathode current is that of a line in an AC circuit, and thus, switching-off occurs when the line current naturally reaches zero. The transfer of current from one conducting device to another previously nonconducting device is called commutation. Thyristors are solely suited in converter applications where the commutation process is carried out by the AC supply. Such circuits are known as line commutated converters. Note that the load now includes an inductance, which is typical of practical applications of thyristor bridges. This inductance is usually large enough to ensure that the current idc is
Slide 145: Electrical Power Generation and Conditioning 127 idc is Supply vs vdc Load Figure 4.26 Circuit of a single-phase thyristor bridge vs t Average vdc α is Figure 4.27 Waveforms in a thyristor bridge, while rectifying nearly constant, at least in the short-term from one cycle to the next. In the longer term, it will vary to match the average level of vdc. A control circuit (not shown in Figure 4.26) applies firing pulses to the thyristor gate connections. The thyristors are fired in diagonal pairs. The fired pair take over the conduction of the current, which reduces the current in the other pair to zero and switches them off. Control of the firing angle α provides a means of controlling the average DC voltage. Figure 4.27 shows the situation when α = 50 °. The shaded areas, above and below the average DC voltage, are equal (because, in the steady state, the average voltage across the inductance must be zero). At α = 90 °, the average DC voltage would be reduced to zero. If α were increased past 90 °, the average DC voltage would theoretically become negative. This is an impossible operating condition in the circuit of Figure 4.26, because the direct current would be required to reverse direction, i.e. flow from cathode to anode through the thyristors. However, if the load resistor were to be replaced by a DC source, whose voltage was larger than the average negative voltage of the converter, the DC current direction will be maintained, power will be transferred from the DC source to the AC side and the bridge will operate as an inverter. Rotating the circuit diagram in Figure 4.26 through 180 ° puts the positive DC rail back at the top and gives the normal way of drawing an inverter, with the power flowing from left to right as in Figure 4.28.
Slide 146: 128 Renewable Energy in Power Systems + DC power source Grid Direction of power Figure 4.28 Line commutated thyristor inverter Note that the inductance on the DC side is still required. It causes the direct current feeding the inverter bridge to be pretty much constant (from one cycle to the next), which leads to this circuit being called a current-fed or current-source inverter. The thyristor inverter shown in Figure 4.28 has several shortcomings: 1. The current waveform on the AC side is almost a square-wave (see Figure 4.27) and therefore has very high harmonic content. 2. The fundamental component of the square-wave alternating current is out of phase and lagging the AC voltage. The circuit consumes reactive power from the grid. 3. The circuit requires connection to the grid (or a synchronous generator) in order for the currents to commutate. Therefore, the circuit cannot be used as a general purpose standalone AC supply. Nonetheless, thyristor inverters and controlled rectifiers, based on the above concepts, were the foundation of the power electronics industry during the 1960s and 1970s. Thyristor inverters were used in early variable speed wind turbines and grid-connected PV systems. The shortcomings of the converter based on thyristors are all but removed by the use of transistor switching devices. 4.5.6 The Transistor Bridge During the 1980s, self-commutated power semiconductor devices (BJTs, MOSFETs, IGBTs, GTOs and IGCTs) started to become commercially viable, which led to the development of self-commutated inverters an example of which is shown in Figure 4.29. In a self-commutated inverter, the switching, both on and off, of the main power semiconductor devices is under control. Thus, a self-commutated inverter does not need a grid connection in order to operate and can be used to create a standalone AC power supply, the frequency of which is determined by the transistor firing control circuit. Note that the inductance on the DC side is no longer required. This inverter is fed with a voltage that is pretty much constant (from one cycle to the next), which leads to it being called a voltage-fed or voltage-source inverter. On the other hand, for reasons explained later, an inductance is required on the AC side if the inverter is to be connected to the grid.
Slide 147: Electrical Power Generation and Conditioning 129 T1 D1 T3 D3 + DC power source vx vdc T4 D4 T2 D2 io vo vi Figure 4.29 Transistor bridge vi T1&T2 Fundamental AC component t T3&T4 Figure 4.30 Voltage waveform of a basic square wave inverter The diodes, shown next to each of the transistors, make the circuit appear more complicated but, in practice, this is not an issue. They are necessary in order to provide a path for the current when the transistors are turned off. Basic Square Wave Applying a very simple switching pattern to the transistors in the bridge circuit shown in Figure 4.29 allows it to produce a basic square wave as in Figure 4.30. A basic square wave can be used to supply some non-critical loads such as incandescent light bulbs and simple universal motors, but it is not suitable for the majority of AC loads, which are designed to operate from a sinusoidal supply. The basic square wave has a sinusoidal component at the fundamental frequency, as shown in Figure 4.30, but it also has a very high harmonic content. These harmonics will typically cause excessive heating in induction motors and transformers. For the same reason, a basic square wave is not acceptable for grid connections. Quasi-Sine Wave (Modified Square Wave) A minor modification to the control system allows the bridge circuit to produce a quasi-sine wave. As shown in Figure 4.31, the quasi-sine wave is really just a modified square wave, but it is a much better approximation to a sine wave. Setting the switching angles to 60–120–
Slide 148: 130 Renewable Energy in Power Systems vi T1&T2 T1&D3 Fundamental AC component t D2&T4 T3&T4 Figure 4.31 Voltage waveform of a quasi-sine wave inverter vi Fundamental AC component of output waveform t Figure 4.32 Pulse-width-modulated (PWM) inverter output voltage 60–120 ° eliminates the third harmonic completely and makes the waveform good enough for many practical loads. The switching angles may be adjusted slightly to control the RMS voltage, bearing in mind that this adjustment increases harmonics. Pulse-width Modulation (PWM) For loads that are more harmonic-critical and for grid-connection, pulse-width modulation (PWM) allows the transistor bridge circuit shown in Figure 4.29 to produce an almost-pure sine wave (low order harmonics are virtually eliminated) and provides full control of its amplitude. In a pulse-width-modulated (PWM) inverter, the transistors are switched at a much higher frequency than that of the intended output waveform. The width (duration) of the high frequency pulses, having short width at the edges and increasingly longer width towards the centre of the waveform, as shown in Figure 4.32, is controlled so as to create a good approximation of a sine wave output. In order to provide a high quality sine wave, the switching frequency needs to be as high as possible. If it is too high, however, switching losses will become significant, making the inverter itself less efficient. Transistors when at their on or off state dissipate very little power. It is during the periods of transition between the two states that they are most lossy. For small MOSFET PWM inverters, switching frequencies up to 20 kHz are typical. While PWM virtually eliminates the low order harmonics, there can be significant harmonics around the switching frequency and its multiples, but these can readily be filtered out. Figure 4.33 illustrates one simple method that the inverter ’s internal control system may use to create the switching signals. The pulses that switch the transistors on and off are generated at the intersections of the reference sine wave with the carrier wave which is usually triangular. In a standalone inverter, the reference sine wave would be created by the internal control system, normally to provide a constant voltage and constant frequency output. Reference [3] provides more information on this rather complex topic.
Slide 149: Electrical Power Generation and Conditioning 131 Reference sine wave Carrier wave t Figure 4.33 PWM construction of a switching pattern Comparison of Switching Methods Quasi-sine wave inverters are widely used for small standalone applications. For larger applications, and for grid-connected inverters, PWM is normally employed. The above discussion has focused on inverters providing AC power at 50 or 60 Hz, but inverters are also used to provide power at much higher frequencies. This is very useful internally in converter systems because it allows the use of high frequency transformers, which, as mentioned earlier, are much smaller, lighter and cheaper than 50/60 Hz transformers of the same power rating. This is a topic that will be visited later when converters for renewable energy sources will be reviewed. Output Control in a Grid-Connected Inverter In a grid-connected inverter, the reference sine wave shown in Figure 4.33 is created so that it is of the same frequency as that of the grid but it can be phase-shifted with respect to the grid voltage. Referring to Figure 4.29 and using phasors rather than instantaneous values, it is clear that Vi = Vo + Vx where Vi is the fundamental component of the pulse-width modulated voltage output from the inverter bridge, Vo is the output voltage, which is the grid voltage once the inverter is grid-connected, and Vx is the voltage across the inductor. Furthermore, the voltage across the inductor must lead the output current Io by 90 °, ie: Vx = jIo X where X is the inductor reactance. Phasor diagram (a) in Figure 4.34 illustrates the general case and shows that control of the amplitude and phase of Vi with respect to the grid voltage Vo provides control of both amplitude and phase of the output current Io. This is essentially the mechanism for the control of the active and reactive power flows as described in Equations (4.8a) and (4.8b) in the case of the synchronous generator. Phasor diagram (b) shows the particular case where the inverter is operating at unity power factor (zero reactive power flow) where the output current Io is in phase with the grid voltage Vo. The required amplitude of Vi and its phase with respect to the grid voltage can readily be calculated for any required output current Io. Furthermore, neglecting losses in the circuit, we know that power-out must equal power-in, i.e.:
Slide 150: 132 Renewable Energy in Power Systems Vi Vx θ Vi Vx Vo Io Io Vo (a) (b) Figure 4.34 Phasor diagrams of the output stage of a grid-connected inverter; (a) general case and (b) at unity power factor DC AC Figure 4.35 Three-phase IGBT bridge Vdc I dc = Vo I o cosφ where Idc is the average value of the current drawn from the DC source. Thus, by using the PWM pattern to control the output, the input current has been controlled. The Three-phase Bridge All of the preceding circuits and discussions may readily be extended to three-phase systems. Usually this requires only an additional 50% of devices as shown in Figure 4.35 where a single-phase bridge with four transistors has becomes a three-phase bridge with six transistors. Inverters above 10 kVA are usually three-phase. Three-phase power electronic converters are extensively used in variable speed wind turbines to convert AC to DC and vice versa. The circuit shown in Figure 4.35 is known variously as a transistor bridge, an inverter, a voltage-source converter, a variable frequency drive or just about any combination of those words. It is widely used throughout industry to operate induction motors at variable speed. IGBT converters are expensive, typically several times the cost of the associated electrical machine.