Copyright Q 2001 by John Wiley & Sons Ltd Daffins Lane, Chichester, West Sussex, PO 19 IUD, England National 012.43 779777 Lntmabonal (+44) 1243 779777 e-mail (for orders and customer sewicc enquines): [email protected] Rzprmted April 2002

Visit our Home Page on http:/h.wiley.co.uk

or http:/lwv.wiley.com All Rights Reserved. No part of this publication may be reproduced, stored in a rettieval systern, or hdnsmitted, in any form or by my means, etectxonic,niechariical, photocopyng, recording, scmrnng or otheMise, except under the terms of the Copyright Designs and Patents Act 1988 or under the tenns of a licence tssiied by the Copyright Licensing Agency, 90 Totienhirm Court Road, London, WlP 9HE, UK, without the permission in writing of the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the publication.

Neither the nuthors nor John Wiley & Sons Ltd accept any responsibility OT liability for loss or damage occasioned to any person or property though using the material, instructions, methods or ideas contained herein, or acting or r&iuning from acting as a result of such use. The authors and Publisher expressly disclaim all implied warranties, including merchantabilityof fitness for any particular purpose. Designations used by companies to distinguish theu products arc often claimed as aadetnarks. In all instances where John Wiley & Sons is aware of a claim, the product names appear in initial capital 01 capital letters, Readers, however, should contact the appropriate companies for more complete informationregarding trademarks and registration. Other W h y Editorial Ofices

John Wiley & Sons, Inc ,605 Thwd Avenue, New York, NY 10158-0U12,USA Wiley-VCH Verlag GmbH Pappelalloe 3, D-69469 Weinheim, G m y Jacdranda Wiley Ltd, 33 Park Road, Milton, Queensland 4064, Australia John Wiley & Sons (Canada} Ltd, 22 Worcester Road Rexdale, Ontario, M9W lLI, Canadta

John Wiley & Sons (Am) Pte Ltd, 2 Clement1 Loop #02-01, Jin Xing Distnpark, Singapore 129809 Libraq of Congress C a t a l o ~ t n g - i ~ ~ u b l&&U i~~on

Power system reshvctunng and daegulatian: trading, perfomance. and inlomrion technology/ edited by L.L. h i . p. em Includes bibliograpliical rsfemces and index. fS%N0 471 49500X 1. BIcciriml power systems - Control. 2. Electric utilities - Cost control. 3. Elcch’ic Utilities Deregulation. 4. Elecmc utilities . Technological innovation%1. Lai, h i Lei ~

TK1007. P68.2001 333.793’2- dc21

2001045404

British Libraty C Q t a l o ~ uin~ ~ Pu~li~aCion g Dafa A catalogue record for this book is avatlahle from the Bntish Library

TSBN 0 47149500 X Produced from Word files supplied by the Editor Pnnted and bound in Great Britam by Antony Rowe Ltd ,Chippenham, Wtlts This h w k is pnnted on acid-free paper responsibly manufactured from sustainable forestry, in which at least two trees arc planted for each onc used for paper production

............................................................................................................................. reface ............................................................................................................................... e~~~............................................................................................................ ................................................................................................................. ent ..........................................................

xv 5

xxi

............. 1

1.2 1.3 1.4 1.4.1 1.4.2 1.5 1.5.1 1.5.2 1.5.3

Competitive Market for Generation ............. .................................................. The Advantages of Competitive Generation .................................................. ....................... The Role of the Existing Power lndustry ............. Reconfiguring the Electricity System. ............................................ Trends in Conventional Electricity G Electricity Demand Operation and Reliability ............................................. Power Plant Operation .................................. Reliability Assessment ...................................................... Availability of Fuel .......................................

2 4

7

s ...........................................

1,6.5

Solar.....................................................

1.9.1

Capital Costs for New Plants ........................

....................................................................

17

.............................. .................................................

1.10.4

Coimectioii and Use of System Charges .

1.11.1

Introduction ..........

25

Contents

vi

1. I J .2

Circuit Connection and P 1.11.3 Performance Analysis..... 1.11.4 Solution Technique......... 1.1 1.5 Results and Discussion ........................... 1.1 1.6 S i ~ p l i ~ Phase-balanci ed 1 .11.7 Appendix ......................... ..................................................... Case Study 2: Controlling a S 1.12 1.12.1 1.12.2 The Solar Power Plant .................................................................................. 1.12.3 Control Structure of the Plant ....... ................................ 1 A2.4 CA Formulation............................................................................................ 1 .12.5 Experimental Results ........... .................................................................. 1 .I3 Conclusions ........................................................................ 1.14 References ...... ............................................................... .........................

38 40

42 46

....................................................................................

2.1 2.2 2.3 2.4 2.4.1

2.5 2.5.1 2.5.2 2.5.3

2.6 2.6. I 2.6.2

2.7 2.8 2.8.1 2.8.2 2.8.3 2.8.4 2.8.5 2.8.6

2.9

l a ~ ~ oofnE ~ ~ c~~ ~~ ~i c~ i t ~ e s Introduction ............................ .......................... Traditional Central Utility MO Reform Motivations...................... ................... Separation of Ownership and Central Dispatch Versus Competition and Direct Access in the Electricity Market.................................... Competition in the Energy Market ..... .............................................. Competition and Auction Mechanism ..................................... Direct A c c e s s ~ ~..e e ~ ~ ~ ~ Independent System Operator .............................................................................. Pricing and Market Clearing .................... Risk Taking............. Retail Electric Providers.,. ............................................................ Different Experiences........................................................ England and Wales .......... ........................................................ Norway ................... California................ ........................................ Scotland ..................................................................... New Zealand........... ........................................ The European Union and Gennany ..............................................................

......................................

.............................

.................................................................

3 CO a ~ ~ e ~ ~ 3.1 Introdtiction ................................................................. .................................. 3.2 The Independent System Operator ................... ............................................. 3.3 Wholesale Electricity Market Characteristics............ .................... 3.3.1 Small Test System ............ .......................................... ............................................... .................... 3.3.2

50

54 54

60

61 63 64

71 72 73 74

76 76 79 80 81 82

Contents

vii

3.3.3 3.3.4 3.3.5 3.3.6 3.3.7 3.3.8

Bidding ..................................................................... ........................... Market Clearing and Pricing ........ ............................................ Market Timing ........................................... Sequential and Simultaneous Markets ............................................... Bilateral Trading............................................... .............................. Scheduling..................

3.3.11

Physical and Financial Markets......................................

....................................................

83

89

97

stem Capacity .......................................................................... 3.5.4

Technical 'Issues......................................................................

4.2. I

Competition in Supply..

4.2.4

Key Issues €or Distribution Businesses ..................

4.2.6

Use of System Billing.................................................

4.2.8

Competition in Metering ..........................

4.3

4.3.5 4.3.6 4.3.7 4.3.8 4.3.9 4.3.10

Maintaining Distribution Planning ...................

Network Planning Tools............................................................................. 124 Asset Replacement Planning ................... . 125 Risk Assessment ......................................................................................... 125 Skills and Resources ................................. .......... 125 Neiwork Design ....... ............................................................................. 126 ~ i s ~ ~ on ~Automation u t i .......................................................

viii

Contents

4.3.1 1 Automation Case Study .Remote Control in London Electricity ............. 129 4.4 Future Devclopmeiit .............................. 4.5 Appendix: Distribution Automation i 4.5. I Introduction ................................ 4.5.2 Remote Terminal Units .................................................. 4.5.3 SCADA Master Station . ....................................................................... 134 4.5.4 S o h a r e Functionality .................................... .... 136 4.5.5 Operations and Maintenance (O&M)......................................................... 136 4.5.6 System Integration, Design and Management................... ............... 137 4S.l Coi~~inunication Systems . ............................................................ 140 4.6 References ............................................................... 5.1 5.2 5.2.1 5.2.2 5.2.3 5.3 5.3.1 5.3.2 5.3.3 5.4

5.5

....................................................1

Introduction ........................................................................................................ Role of the TP ....... .................................................

New Market Organisation .

Priority Insurance Scheme........................................................ Transmission Expansion ........................... Conclusions ........................................................................................................ References .............................................. ....................

...........................................................

6.1 6.1.1 6.1.2 6.1.3 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.4 6.2.7 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.4

153 155

169 170 171

................

pen Access ...*..........= 17 Introduction .................................................................... The Traditional Power Industry Motivations for Restructuring the Power Industry.. Unbundling Cencration, Transmission and Distribution ........................... 174 Components of Restructured Systems........................................ ...... 175 Gencos ................................. ............ ..................................... 175 BOT Plant Operators and Contracted IPPs .. 175 Discos and Retailers ..................................................... 175 Independent System Power Exchange (P

....................................................

176

....................................................

176

....................................

176

PX and ISO: Functio California Power Exchange ........................................... IS0 Functions and Responsibil~ties........... ..................................... Classification of IS0 types ..................................................... Trading A~angements ................................ ..........................

178

183

6.4. I 6.4.2 6.4.3 6.5 6.5.1 6.5.2 6.5.3 6.5.4 6.5.5 6.6 6.6.1 6.6.2 6.6.3 6.6.4 6.7 6.7.1 6.7.2 6.7.3 6.7.4 6.7.5 6.8 6.8.1 6.8.2 6.5.3 6.8.4 6.8.5 6.9 6.10 6.1 1 6.22

The Pool .......................

.........................................

ulti 1atera1 Trades ......... T r ~ s ~ i s s i Pricing on in

olled-in Pricing Methods ..............

..........................

184

..........................

186

.........................

187

System Control ............... hicillary Sewice Provision............................................. Congestion ~ a n a g e ~ eii?n Open-access t Transmissioo Systems...................... Congestion Management in Nomial Operation.......... Integrated Transmission patch Strategy................ wer System ........................ ~ ~ ~ u s t r a Using ~ i o i i a Sma Sfatic Security-constrainedRescheduling ...................................... Dynamic Secmity-constrainedRescheduling ................. ........................... Open-access ~oordi~iatioii Strategies..... Price Elasticity as a Me Relieving ~ongestion ~ o o r d i n a ~ ~betwce on ~l~ustrat~on o€Traii~actionCoordiiiatioii ...... ~ n t e ~ r ~Coordinati ted Conclusions ..........................

............................................................................. ...... .................................................................................... a

............................................

*.**

195

202

209

2 16 216 217

a***...

7.1 Development of Electric Power Industry 7.2 S ~ c c e ~ Growth ~ ~ v e of Power Produ 7.2.1 Further Expansion of Power Nehvo 1.2.2 Continuo~I~ increase of Electricity C 7.2.3 7.3 ~ a n a ~ e i System n e ~ ~of Electric 7.3.1 The State Power Corporation ......... ..................225 hilosophy aiid Strategy o f tlie SP ............................................................. 23 1 7.3.2 Market in China... 7'4 7.4. I ~ o ~ i v a t ifor o ~Reformation s ...................................................................... 234 7.4.2 efonn PLaii of tile SP ........ 235

Contents

x

Obstacles in Establishing the Power Market in China 7.4.3 7.5 Electricity Pricing .................................................... Basic Theory of Predicting Electricity Costs .. 7.5.1 Electricity Cost Derivation.............................. 7.5.2 Elcctricity Pricing of Inter-provincial Power 7.5.3 7.6 Traiismission pricing ...........,....,.......................,.,,. Current Decomposition Axioms 7.6. I ~athematicalModels .,.................................. 7.6.2 ................................... '......'.....~..,..252 Methodology of Graph The0ry..... 7.6.3 Algorithms and Case Studies............................ .... ...... 253 7.6.4 ......................................... ......................... 254 7.7 7.8 Acknowledgenients............... ............... 7.9 ~

8.1.1 Benefits o f FACTS Technology....................... 8.2 Transmission System Limitations ........ 8.2.1 System Stability.................................... s..............._...261

8.2.4 8.3

Thermal Limits ......

FACTS TeGhnology.............

..........................................................

262

Control Methods and DSPiMicroprocessor Technology ........................... 264 8.3.2 8.3.3 Present Status on FACTS Solution Options with FACTS ............ 8.4 8.4.1 Fundamental Concepts of 8.4.2 Skuiit Controllers.. ............................................................................. 266 8.4.3 Series Controllers .................................................. Combined ScriedShuiit Controllers ...,............,.............. 8.4.4 8.4.5 Phase Angle Controllers.................... HVDC Transmission Controllers ............................................................... 278 8.4.6 Other Controllers........."............. 8.4.7 8.5 ........................ '........ 281 8.5.1 85.2 8.5.3 8.5.4 UPFC .................................. Concluding Remarks ....... 8.6 Ackaowledgements ........................... 8.7 ...."_........_._. ..... 284 .........l............................ 8.8 ~

~

Contents

Xi

.........................................................................................................

anagenment

...............................................................

9.2 Pre-privatisation (1 990): Th 9.3 Post-privatisation(1990): F 9.4 Early-inid 1990s: Getting t 9.5 1994/5+: Getting More for Less .................... 9.6 Late 1990s: Capital Effici August 1999 Interim Report: All Change?.... 9.7 9.8 The 1990/2000 Regulato 9.9 Asset Ownership............................. 9.10 Asset Governance. 9.11 Asset Management ...................................... 9.12 Asset Information and t 9. I3 Condition Monitoring.................................. 9.13.1 Transforniers.....

.......................

Switchgear..

9.13.5 9.13.6 9.13.7

~ n d e r s ~ a i i dLong-term in~ Asset Costs ......................................... Underground Cables ..................................................... HV Cables.................... .............................

9.13.9

Zero Sequence Impedance

...................

...............

9.17

9.17.5

Common Mode Failure.................. Asset Infoimation Acquisition ...........

Data Cleaning ...........................................

289

...............

9.13.3

9.16.4

287

298

Xii

Contcnts

--

I -

9.18 Conclusions .......................................... ........................ .322 Appendix: Fuzzy DGA for Diagnosis of Multiple Incipient Faults...................323 9.19 9.19.1 The IEC DGA Codes ...... ......................................... The Fuzzy IEC Code 9.19.2 9. i9.3 Fuzzy C)iagnosis Results... ............................. 9.19.4 Trend Analysis of Individual Faults .... ........................................ 327 9.1 9.5 Comments........................... ............................. 9.20 Refesences . y

10.1.1 10.1.2

10.2.1 10.2.2 102.3

10.3.4

10.6

11.2 11.2.2

..............................................................................................................

..................................................................................... A General Overview .................................... PQIS

The Wavelet Transform. ........................................... Wavelet Analysis ..................................................... Application to PQ [25] . .....................................

336

...........................................

342

........................................................

353

~ p c s i o ~Distortion ic

~~f~rc~ces

S o h a r e Agents .....................................

Genesal Issues and the Future of Agents.

....................................

.4

339

362

Evolutionasy Programming-based Optimal Power Flow Algorithln

.............................. 11.4.2

EP.. ...................................

..............................................

11.4.4 11.4.5

Load flow Solution.......... Gradient Acceleration........................................................

...........................

373

379

'.*

Contents

U

Complex Artificial Neural Networks for Load Flow Analysis ..... 11.5 Conventional A" for Real Numbers .. 11.5.1 New ANN for Complex Numbers ......... 1 I .5.2 Comparison of the two ANNs by Coinputer Sirnulatiou ........................... 11.5.3 113.4 Applicati 11.6 Virtual Reali Types of' VR systems................... 11.6.1 1 1.6.2 Non-immersive (Desktop) Systems. .............................................

11.6.6

Cave ....................................................

11.6.8

Augmented ...............................

11.7.1 11.7.2

The Hardware ................. ................................................................... The Correspondence......................................

388

396

401

X 1.7.4 Iinp~eiiientationExample........................ 11.8 Coiiclusioris ........................... 11.9 Acknowledgements..

................. 12.2 The Internet................................ ...................... 12.2.1 What Is the Internet? .................................................................................. 416 12.2.2 oes the Internet Work .... ............417 12.2.3 What Would Happm Without the Intcrnet? ............................................... 417 12.2.4 Wow Can the Power lndustry Benefit from the Internet?. 12.2.5 ow Can I Find the Inromiation I Need?.................................................. 419 12.3 Usability of the Interne 12.3.1 12.3.2 12.3.3 Internet Products.......... 12.3.4 12.3.5 iltimedia Access .............. 12.3.6 0x1-line Setvices ......................................................................................... 42 I 12.3.7 Support for Professionals ........... 422 12.3~8 The Power Industry and the liitemet .......................................................... 422 12.3.9 Recent Improvements on the Inteilnet ....... 424

Contents

xiv

12.4 Internct Technology............. .... ....... .................424 12.4.1 Access to the Internet ................................................................................. 425 12.4.2 Operating Platforms on the Internet ..

12.4.7

Internet Sccurity ......................

......................

433

Interpreled Versus Compiled Iaiguages .... 12.5.3 What Is JavaScript? .................................................................................... 434 12.5.4 What Is Java? .............. .......... ........ ........ 435 12.6 Web Pages. ................................................ 436 12.6.1 .. 437 12.6.2 Difference Between a Static and a Dynamic Web Page ............................ 437 12.6.3 Displayiiig Database Content ................................... 438 12.6.4 Web Pages with Fuuctionality.................................................................... 440 12.6.5 Web Pages with Integrated Applications ...... 12.7.1

Why the Need for XML

12.75

ation of Content and I .ayout.. .................... Layout Validation with DTD .............................. Styleshects .......

.....................

445

Monitoring Power Station Equipment........................................................ 454 12.8.3 .457 12.5) Case Study 2: Power Trading Application .... Trading Platform Architecture ................................................................... 458 12.9.1 12.10 Conclusions ................................ 12.11 Acknowlc~~ements ...... ................................................................... 460 12.12 Refercnces .................... ex

..................................................................................................................................

The electricity power utilities in many countries have been, or are being, rest~c~ured. There are many reasons for restructur~ng,In some countries restruc~uringhas been driven by the desire of gove~mentto meet ~ncreasingdemands for electricity by encouraging independent power production, which relieves government of a financial obligat~on.In countries where ownership of assets is in private hands, restructuring has been driven by mergers and ac~Liisi~ion~, as companies seek to gain competiti~eadvantage. In the most a ~ v a n countries, c~ restruciuring is being driven by the desire to allow consu~ersto choose their electricity supplier on the basis of price and service provided~ These drarna~~c changes in the organisation of electricity power utilities bring with them new challenges and opportunities, as the previous centrally designed and operated systems are di~mantledand replace by a new competitive framework. ~ o m p a n i ~operating s in a competitive market need more s o ~ ~ ~ i s t i c acontrol ~ e d and management systems to ensure that their business objectives can be achieved. The development and application of new technologies is also accelerated in this new environmen~~ as companies seek to improve their effecliveness and efficiency. This book is con~ributedby a group of world authorities. It explains in depth the reason ring, without including superfluous detail. Examples are given from tails are provide^ on new s~rate~ie$ and tec~nologie~ which are being f ~ e ~ ~ e ~ a~ransmission tion~ and supply. The implications for the ~ n v i r o n ~ e ~ ~ are also reviewed. Tools being ~ t i for ~ asset ~ s an~age men^ and fo management of ~nfrastruc~ur~ are i~l~strated with practical examples. m o d e ~ l i nand ~ general analysis of ~ o m p e t i ~ power ~ v e markets are also illus This book provides a com~rehensivereview of all the many facers place in a d y n a ~ i c~ ~ d u s X~t iys .c o ~ p u ~ s o rreading y for graduates and e n g i ~ e ~and ~s, other pro~essiona~s, who are entering or involved in the electricity power industry.

avid G. Jefferies CBE, B;

This book was written as a result of the ongoing stimulating worldwide dere of the power industry. This move away from the ~aditionalmo towards greater competition, in the form of increased numbers of indepe producers and an u ~ b u of~ the~ main ~ ~service, n ~ starred in the United King and this change was driven by the large differences in electricity tariffs across regions, by adva~icemen~s in technologies which &low small producers to c o ~ p e t ewith large ones, and by a strong belief that competition will produce an all-win situat~on. group of experts to produce a broad and The book was contributed by an ititernat~ona~ of the main issues. The intent has been to provide the reader with an in etail ut without excessive specialisation, to avoid a purely ~ualitativetrea~meIi~ epth by ~ ~some a ~~ a ~ y and ~~ c numerical ~ d to offer9 ~ whenever ~ possible, ~ a ~l methods, and real case studies, worked examples and project discussions. Since each power utility is unique, it will not be possible to present the best path to fotlow in the restructuring exercise. The market models, regulation and tariffs used by orks, and the r r ~ e c h ~ ~for ~ s~m~ i n t a a~high ~ ~ level n g of r e ~ ~ a ~will ~~i~y, use of the advancement of communications technology and increased compu~iii~ power, it is possible to consider different market structures. a d v ~ n c ~ ~no e n~ t ~ n f ~ ~ could a ~ be ~ availabtbte o n in time for the business o ~ ~ r a t ~ ~ n . Different markets have been considered in the book. In brief, they could be § u ~ m a r ~ s e d types. In the complete1 ~ a r ~ e t - d r i v eenv~~onment n rket ~orcesseek to the b e h a v i ~ of ~ rvarious layers in the market, e.g. the regulators. In the kransiti 1 markec there is a process o r ~ ~ u i a et n~ vdi r o n ~ ~ to n t a d ~ r e g u l a ~~nvironment. e~ In the embry~nicfree m a ~ ~ e ~ , state retains own~rshipof the generators and some of the ~ a n s ~ ~ s sinfrast~uct~re, ion opens up the market to ~ ~ m competition ~ ~ e d at the distribution level. As there is much u i i c e ~ ~ n in t ythese environments, due to the s t ~ c t ~of r ethe E p i ~ n i n gover a long-term horizon is p e r c e i v ~as very difficult at present. Yet, long-term planning, it is likely that the electricity power industry would be at great risk, as it ~~~$~~ not be able to supply the growing d ~ or to~~~~~a~~ n the ~ ~ service as it is c ~ r r e ~ t providing ly to its consu~ers.The recent chaos in his could have very serious con~~quences to the lon ~ndus~~. This book shows how new ~ e c ~ i n o ~will o ~allow y us to cha market structure to one that relies on co~~petition to set the t e c ~ ~ o ~ o g iwe e s ,can use less energy, result in^ in lower ene avoid OX defer a d d i ~ ~ o nexpensive a~ plant c o n ~ ~ rThe ~ ~a ~~~ do OF ~~new . ~ p~ ~o ~~ ~ ~ ~ such as independent power producers, power marketers and brokers, has a ~ d e da new task of maintai~inga reliab~eelectric system. This book will detail into accou$itsome of these issues. In the new market e n v ~ r o n generation ~ ~ ~ ~ ~ represents ~, most of the CO r e p o ~on~ the deve~opmentof new strategies and compares ~ ~ f f e ~tec ent e l e ~ t r ~ c i~~ey n e r a ~ iwith o ~ ~ n v i r o n m ~and n ~ ~political considerat~ons. This i ~ c i ~ d e s t

xviii

Preface

decen~a~ised power supplies, renewables, regulatory constraints, new technical challenges ifferent mechanisms, such as the pool, have been set up for the operation of the new emerging electrical market. The market should dictate when new generation is needed and where it is located. ince there i s a large number of players in the m ~ k ~it ti s, i m p o ~ n to t WO type of bidding, or negotiation strategies that each player can use. It is especially ~ ~ p o ~ a n t to work out the information content of the bidding strategies. Chapter 2 covers expe~ence tools for from various countries on power utility res~cturingand deregulation. An~lyt~cal the ~ o d e l ~ i nand g analysis of c o m ~ t ~ t i vpower e markets are presented. Chapter 3 also discusses several wholesale electricity markets around the world and most of these are in a continuous process of change. This evolutionary process is being d ~ v e Kby ~ the need to address some of the outs~andingissues in the design and implementation of these markets. Some challenges, such as reliability, market power evaluation and mitigation, are outlined. hapter 4 reports on the change in ~~s~nbutio13 business in a dere~ula~e Various issues such as planning, control, load forecasting, metering, customer services and risk assess men^ have been considered. A case study on the remote control of London ~ l ~ ~ r iiscincluded, ~ly Chapter 5 deals with transmission expansion. Following develop men^ of the market, the transmission provider transforms into the independent transmission company (TTG) so as to adnilmir a highly sophisticated market. The ITC is required to make c~mplexbusiness 5 over a wide range of time scales, such as the long-term, short-term and near realis chapter discusses future directions and ~ o d i ~ c a t i o ntos the ~ g u l a ~ o policies ry r. 0th a market maker and a service ssion open access. The economic issues associaked with scussion of some ~rnpor~an~ opera~oiia~ issues in the e ~ e r g ~ n mal dispatch, congestion mana~ementand the e ~ f e of c ~~ e c ~ i n en discussed with examples from the open-access viewpo~nt. Chapt~r7 deals with the Chinese market. A tailed back~roundon the industry is given. It also explains why the approaches a opted by the d e v ~ ~ o p are not suitable. The chapter also proposes a new app ch to c a l c u trans ~~~~ power systems with better e f ~ c i ~ n can y ,ac~urate To operate the ever i transmission loss m ironment, reactive power control to assure v o ~ ~ ~ Row control to avoid line overloading o p ~ r a t i o Flexi ~. ctronics technology presents the applica ms. The impact of entrants is discussed. Chap~er9 deals with asset management. A comprehensive awe ~i~y required to support business in the deregulated e ~ e c ~ i market. characteristics 06 the model components are descdbcd in detail. It wit1 benefit all internal and external users in the open-access environ~ent,resulting in realistic and traflspa~ent open-access charges, and bring long-term ecoi~omicbenefits to all pa anagemene in power industry r e s ~ ~ ~ c t utire n ni ~ ~ u ~ ~with a t epractical d ex~~pl~s.

Preface

xix

Elechicity industry restructuring has had a dramatic impact on the energy market. To gain a conipetitive advantage, toclay’s energy providers need to focus on value-aclded products and services, such as power quality. Powcr quality is a critical issue for industrial customers, especially in the high-tech sector. In order to understand power quality, many customers or energy providers have installed power quality monitoring systems to record electrical system perfo~ianceandor facility equipmcnt reactions, and the analysis of the monitored data has become a challenge. Chapter 10 reports on the techniques, methods and standards used or proposed for power quality issues. The explosion in thc use of information technology has seen the introduction of computer-based work management systems, asset management systems, and control systems to manage system operation. Information teclinology is rnalcing markets more efficient, resource production less speculative and costly, and the delivery and monitoring of energy more etficctive, while enfi-anchising customers to make more intelligent choices. Improvements in infomation technology will continue to allow economical aiid reliable solutions to problems facing tlie power industry. Chapter 11 introduces intelligent agents, genetic algorithms, evolutionary programming, artificial neural networks and virtual reality technology, and reports on their applications to load flow, valuing electrical options and power equipment diagnosis. Tlic chapter highlights the technology behind the new market brought about by deregulation. Energy service companies will continue to make iucreasing demands for more sophisticated software and equipment to monitor and control various aspects of power delivery. In just a few years, Java has taken the networked world by storm. Java comnbiries powerful, object-oriented programming with the ability to run on any computer platform without the need for recompiling or translating. Java promises to play a yet more kndaiental role in the future of on-line computing, including electronic commerce, for it can allow anyone to make use of powerful applications anywhere. One result of its i s that a scrap of code called a Java applet can be embedded in a platform iI~~lepe~idence World Wide Web page. Chapter 12 deals with the application of the Intcmet to power station monitoring and discusses its use for energy trading. It also presents an introduction to Web technology and i t s applications. This book addresses the most up-to-date problems and their solutions in the arm of power system restructuring aid deregulation in a cohesive manner. It will provide invaluable information for power engineers, educators, system operators, managers, planners and researchers.

i

The editor wishes to thank Mr Peter Mitchell of Wiley and his team in supporting this project. The editor also wishes to thank all the contributors, without whose siipport this book could not have been coiiipleted. In particular, the editor thanks Harald Brawn in maiiagiiig to complete the m a n ~ ~ s c rdespite ip~ great diffkulties caused by software ~iico~patibility. The editor also wishes to thank rs Vinay Sood and Professor Sood for their creation of the iuitial manuscript. The editor i s very grateful to Dr D a d Jefferies for writing the ~ o r $ w ( ? rThe ~ . permission to reproduce copyright materials by the IEEE and IEE for a number of papers mentioned in some of the chapters i s most helpful. The arrange~ento f the index by Miss Qi Ling Eai and Chun Sing Lai is imch appreciated. Last but not least, we all thank Wiley for supporting the prcparat~~n oftbis book and for the extremely pleasant co-operation.

ei Eai was appointed Senior Lecturer at Staffordshire Polytechnic (now Staffordshire University) in 1984. From 1986 to 1987, he was a Royal Academy of Engineering Industrial Fellow to both GEC Alsthom Turbine Generators Ltd and its Engineering Research Ceutre. He is currently Head of Energy Systems Croup and Reader in Electrical Engineering at City University, London. He is also an I-lonorary Professor at the North China Electric Power University, Beijing. Dr Lai is a Senior Member of the IEEE and a Corporate Member ofthe TEE. We has authoredlco-authoredover 100 technical papers. Tn 1998, lie also wrote a book entitled Ivrtelligenf System Applications in Power Engineering - Evolutionary P r o ~ a ~ m and i n ~Neural Networks published by Wiley. Recently, he was awarded the IEEE Third Milleiiiiium Medal and 2000 IEEE Power Engineering Society UKRl Chapter Outstanding Engineer Award. In 1995, he received a high-quality paper prize from the International Association of Desalination, USA. Among his professional activities are his contributions to the organisation of several ~nternat~ona~ conferences in power engineering and evolutionary computing, and be was the Conference Chairman of the International Conference on Power Utility Dercgulation, ~ e s ~ c ~ f l n g and Power Technologies 2000. Recently, he was invited by the Hong Kong Institution of Engineers to be the Chairman of an Accreditation Visit fo accredit the University (IIons) degree in electrical engineering. Dr Lai is also Student Recruitment Officer, IEEE UI(R1 Section. In 1999, he was included in The Dictionury of Contemporary Celebrities qf Worldwide Chinese. In 2000, his biography was included in the 18th Edition of J%zo ’5 !4%0 in the FVorld, Marquis, 1JSA. His b i ~ ~ g r has a ~ also ~ y been selected €or inclusion in the 2001 Who I;yho in Science and Engineering, Marquis, USA. Sc, PhD and DSc from UEUIIST, ~ a n c h e s ~ eIJK. r,

of the Royal Society of New Zealand. From 19’70 to 1975, he was Head ofthe Power Systems and High Voltage Groups, UMIST. From 1975 to 1999 he was Professor of Electrical Engineering, University of Canterbury, ~ h i ~ s t c ~ ~ uNew r c h ,Zealand. From 1982 to 1995, he was also the Director of Systems Software & Instrumentation (a Christchurch-based consulting conipany established in 1982). From 1985 to 1990, hc was Head of Department, Electrical and Electronic ~ n g ~ n e e rUniversity i ~ ~ ~ , of Canterbury. From 1988 to 1995, he was a ~ e i n b of ~ rthe CIGRE-I4 Working Group on HVdc harmonics (14-03). From 1989 to 19525, he was Convenor of GIGRE Task Force 36-05114-03-03 on AC System Harmonic ~ o d e l ~ i nfor g AC Filter Design. From 1990 to 1996, he was a Member of CIGRE JWG 11/14-09 on Unit Connection. From 1996 to 1999, he was Convenor of CIGRE Task Force 14.25 on Wannonic Cross-inodulation in HVdc Traiisniission. Since 1990 and 1995 respectively, he has been Dircctor of CHART Instniments, Clx-istchurch and Director o Consulting, a Christchurch-based consulting conipany. Professor Arrillaga h many awards, such as John Hopkins Premium of the IEE, UM, 1975; the Premium, IEEE Conference on Harmonics and Quality of Power, ~ ~ ~ Q P 9 Electrotechnical Paper, IPENZ Annual Conference, 1996; Uno Lamm Hig Current Award, IEEE, 199’7; John Munganest International Power Quali Power Industry, 199’7; President’s (Gold Medal) Award, Annual Meeting xxiii

~ogra~hy

XXIV

~otechn~cal paper, I P Annual ~ ~ o~n f e r e~n ~1999; e , Silver a1 S o c i e ~€or Innovatio~in S~ienceand ~ e c h n Q ~ o2~ y , ~ e c h n i e aCommittees ~ Award, 2000, trained in the area of ~ o ~ e~ec~ronics e r with ~emen$,~ r a n k f u ~ , Germany, from 1985 to 1989. He obtained his Diploma in ~ e ~ e c o m ~ u n i c a t iato n F r i e d ~ ~ r ~ - ~ ~University, essen G ~ ~ a n in y , 1994. He was a ~ a ~ - t i ml eec ~ r e rat City ni~~ ~ ~ ~ ~ r ~s Q i tny~, ofrom n , 1994 to 1996 ~ e a c ~oi bnj ~ c ~ - ~ r i e~n tr ~od g ~ a ~inr C++. e was a Senior Programmer at A M . EST I n t ~ ~ a t i o nEtd, a ~ ~ o ~ ~ from o n 1996 , to esent, he is a Senior Software ~ n ~ i n e e r A L T I ~London, , deve~op~ng new nology s o f ~ a r eHe . is w ~ r k i nfor~ his at City ~ n i v e r s on i ~a p a ~ - ~ i m e ects to achieve it in July 2001. His arch interest is the e~tractionof i n f o ~ from a ~ data ~ ~using ~ neural nehvor ~0~~~~~~A. is Chair Professor and ~ngin~ering, The Hong Kong Polytechni ity. His BE degree is from the U n ~ ~ ~ofrGeylon s i ~ and PhD from ~ ~ p e r i a l Londo~.He has ~ ~ e v i o u worked s~y in Sri Lanka, USA, ~imbabweand Sweden a search interests are in power system r e s ~ ~ ~ ~pricing, i n g , control, MVDC, ~ a n s ~ estability, n~ ~ro~ection and ~ e l i a b i ~ i ~ . ~ r ~ € e s sDavid o r was elected an IEEE FeItow in 2000 for his ou~s~and~ng ~ o ~ ~ i b ~to~ i o n trans~i~$sion acces He is the regional editor electricity supply in dust^ reform an ~o~~l tric ~ o wSystems ~ r for Asia of the ~ n ~ e r n a tJournal

is r~$p~nsible for skate s the dis~ibutioneonip ~ ~ n of c ethe pian~inga n systems in the UK and abroa businesses, and helped deve~opthe d~s~ibutjon artered ~ n ~ ~ nand eer lie has been at MIT since 1984 as a Senior Research Scientist in the here she conducts research and teache§ ~ r a d ~ acourses te in the area of systeni§~ Since September 1999

Young Investigator Award for ~ s ~ i n ~ ~Li esc~~ r~e er .Pro~ess d ale ~ i e c t ~ power c sys

Biography

xxv

has been Chairman of the National Grid Company plc since 1990, when the Company was formed as part of the privatisation of the UM electricity sector. His bold and far-sighted leadership has been a key ingre~ientin its success of the National Grid Group plc from the performance of the transmission system during a decade of major change in the industxy, though the conception and development of Energis, to the growth of the group internationally. He retired as the Chairman of the National Grid Group plc in July 1999. Dr Jefferies was previously Chairman of the London Electricity and of Viridian plc. He was the 1997/98 TEE President. Owing to his huge contrihution made to thc institution, he i s an Honorary Fellow of the IEE. He i s also a Fellow of the Royal Academy of Enginecring. He was a pioneer in the restnicturing and deregulation of the UK electric power utility. ia received his PhD from the University of California, Berkeley, in 1983. Since then, he has been at the University of Washington, Seatile. He is currently Professor of Electrical Engineering and Associate Dean of Engineering at the University. Dr Liu is a Fellow of the IEEE and the US representative on CTGRE Study Committee 38. His areas of interest include power system economics, intelligent system applications and vulnerability assessment.

o obtained his MSc and PhD from the University of Manchester Institute of Science and Technology. He is currently the Head of the Power Systems Research Group at the University of Strathclyde. His group specialises in energy management systems, issues concerning the electricity market and deregulation, simulation, analysis, monitoring and control of powcr networks. Professor Lo has been an international advisor and member of many organising committees of international conferences, consultant/visiting professor to over 12 educational institutions, and has lectwed extensively in the Far East, Europe and America. He is the author of over 260 technical pL~blications.He is a Fellow of the TEE and a Fellow of the Royal Society of Edinburg~~. is a member of London Electricity’s Executive and is currently the Managing Director of both London Power Networks (LPN), which i s the distribution business of London Electricity, and London Electricity Services (LES), which is the private networks business of London Electricity. As Head of the Public Distribution Business he led the work during 1999 which culminated in the formation of 24sevei1, the joint venture network management services provider formed by LE and TXU Europe (Eastern Electricity). He has been in the electricity supply industry for 25 years in a variety of both operational and strategic roles within the distribution business. He has a practical engineering background having worked in a number of operational, project manager and leadership roles i n utility power distribution. Mr Morton is a Chartered Electrical Engineer and a Fellow of the IEE. He also represents the UK in the business area of distribution at ~ U ~ the pan-European E ~ association ~ ~ of electricity ~ companies. ~ C ~ SS ~~~yreceived his BE and PhD degrees from the National University of Ireland, Dublin, in 1983 and 1989 respectively. He is currently a Professor at the National University of Ireland, Dublin, with research interests in power systems, control theory and biomedical engineering.

XXVi

Biography

ebllk is a Professor of Electrical and Coinputer Engineering, Iowa State University, Ames, Iowa. He received his BS and MS degrees in electrical engineering from Purdue University and his PhD in electrical engineering from Virginia Tech. His industrial experience includes over 15 years with a public utility (Commonwealth Edison), with a research and development film (Systems Control), with a computer vendor (Control Data Corporation) and with a consulting firin (Energy and Control Consultants). He has participated in the functional definition, analysis and design of power system applications for several energy management systems since 1971. Dr Sheble also designed the optimisation package in use at over 50 electric utilities to schedule electrical production. He has consulted since entering the academic world with companies in North America and Europe on electric industry deregulation as well as expert witness testimony on the National Electric Code and lntellectual Property Rights. His consulting experience includes significant projects with over 40 companies. He developed and implemented one of the first electric market simulators for the Electric Power Research Institute using genetic algorithms to simulate the competing players. He conducts approximately 24 seminars each year on optimisation, artificial neural networks, genetic algorithl~~s and genetic programming, and electric power deregulation around the world. His primay expertise is in power system optimisation, scheduling and control. Dr Sheblt? has been awarded over 1 million dollars of research support over tlie last 10 years, primarily in the application of adaptive agents to market bidding. He has authored a review of adaptivc agent market-playing algorithms for the Kluwer press release P o w r Systems Restmcluring: Engineerin'q and Economics edited by Ilk, Galiana m d Fink. He has written a monograph on tools and techniques for energy deregulation entitled Conjputational Auction Mechanisms.for Restructured Power Industry. He has also been an invited guest on radio talk shows and a resource for several news articles on electric power deregulation and industrial trends. His research interests include power system optimisation, scheduling and control. Professor SheblC is an IEEE Fellow. ~ o ~ Vijay e ~So0s obtained ~ ~ his BSc ftom University College, Nairobi, and his MSc degree fi-om Strathclyde University, Glasgow, in 1969. He obtained his PhD degree in power electronics from the University of Bradford, England, in 1977. From 1969 to 1976, DKSood was einployed at the Railway Technical Centre, Derby. Since 1976, he has been employed as a Researcher at IREQ (Hydro-Qukbec) in Montreal. Dr Sood also has held Adjunct Professorship at Concordia University, Montreal, since 1979. Dr Sood is a Member of the Ordre dcs ingknieurs du Quebec, a Senior Member of the IEEE, a member of the IEE and a Fellow of the Engineering Institute of Canada. He is the recipient of the 1998 Outstanding Service Award f%omIEEE Canada, the 1999 Meritas Award from the Ordre des IngBnieurs du Quebec, and the IEEE Third Millennium Medal. Dr Sood is presently the Managing Editor of the IEEE Canadian Review (a quarterly journal for IEEE Canada). He is a Director and Treasurer of IEEE Montreal Conferences Inc. He has worked on the analog and digital modelling of electrical power systems and their controllers for over 25 years. His research interests are focused on the monitoring, control and protection of power systems using artificial intelligence techniques. Recently, Dr Sood has been interested in the Internet and its applications for teaching purposes and was mandated by IEEE Canada to publish the journal IEEE Camdian Review on the Internet (www.ieee.ca). Dr Sood has published over 70 articles and written two book chapters. He has supervised 14 postgraduate students and examined 13 PhD candidates frotn universities all over the world. He is well known amongst thc electrical engineering community in Canada.

Biography

xxvii

is Technical 2% Regulation Manager of London Power Networks (LPN). r Cliff LPN is the distribution business of the London Electricity Group. In his current position, he is responsible for all technical and regulatory matters regarding the public cleclricity distribution system in London and particularly the quality of supply and reliability performance that sets London apart. He has previously been Strategy Manager, Asset Manager and Planning IvIaager for London Electricity's Public Networks Group. In his recent roles he has championed the development of an integrated technology strategy, strategic asset management, fault causation analysis, incipient fault detection and location techniques, as well as creating the strategies behind the implementation of one of the largest distribution remote control, telemetry and automation pro-jects. Mr Walton joined LPN whea it was established in April 2000; his career in electricity distribution spans 29 years. IIe has worked with a number of overseas utilities and has written and presented many papers on a wide variety of technical and asset governance and m ~ a g e ~ issues. en~ He is a Chartered Electrical Engineer and a Member of both the IEE and IEEE.

r ~ ~ ~ s s ~ r was born in May 1936. He graduated from Xi'an Jiaotong University in 1957. He has since been with the School of Electrical Engineering of the university, where he now holds the rank of Professor and is the Dircctor of the Electric Power System Department. He is a Senior Member of the IEEE. From September 1981 to September 1983, he worked iii the School of Electrical Engineering at Cornell University in Ithaca, New York, USA as a Visiting Scientist. From September 1991 to September 1993 he worked at the Kyushu Institute of Technology in Kitakyushu, Japan, as a Visiting Professor. Prof Wang has a 40-year experience of researching and teaching in electric power system analysis and planning. His main research fields include reliability evaluation, generation and transmission network planning, operation planning, system contingency analysis, dynamic and transient stability, short-circuit current calculation, optimal load flow, and probabilistic load flow. He is especially proficient in constructing mathematical models and developing application software in the above areas. He also took part in many research and planning tasks of key electric power projects in China, such as the Three Gorges Hydro-Power Station. He proposed a new transmission system, namely the fractional frequency transmission system (FFTS) which uses a lower frequency to reduce the reactance of AC h-ansniission systems. I n recent years, he has been researching the electric power market.

~ t received s ~his BE ~ (€Ions) and PhD degrees fiom the University o f Canterbury (New Zealand), where he is now a Senior Lecturer. Dr Watson has authored and co-authored approximately 100 technical papers and 3 books. Paper awards received include; Best Paper Award (The Sixth International Conference on Harmonics in Power Systems, 1994), the William Perry Award (TPENZ) and Finalist for the Carter Holt Harvey Packing Award for Innovative Technology (IPENZ). He has also given a nuinbcr of invited lectwes in Singapore, Australia and Canada, ail Wen received his BEng and h4Xng degrees from Tiarijin University, China, in 1985 and 1988, respectively, and his PIiD from Zhejiang University, China, in 1991, all in electrical engineering. He was a Postdoctoral Fellow at Zhejiang University Eroin 1991 to 1993. He joined the faculty of Zhejiang University in 1993, and has been a Professor of Electrical Engineering since 1997. We held a visiting position at the National

xxviii

~io~apliy

apore from 1995 to 1997. e is on leave from Zh ong Polytechnic Universi as a research fellow. ral Science Award of China, Zhejiang Provin~~aI Top Young ~cientist and several other awards from the Ministry of Education (China), Zhejiang ial gove~ment,Zhejiang University and the National University of Si~gapore.He is a ~ e ~ bofe the r editorial board of the JournQzo ~ ~ ~ ? o mofEleclric u ~ ~ o n Power ~ y s ~ e ~ s ese) and was a guest editor of a special issue on ‘Artificial intelligence ~pplicat~ons r systems’. His research interests are in power system r e s t ~ c ~ r i nand g artificial lications in power systems. obtained MSc, PhD and DEng from University of ~ a n c h e s ~ e r echnology in 1971, 1974 and 2001 res~eceive~y. C u ~ e he n is ~ ~ ~ Electrical En~ineeringat the University of Western Australia. system dynamics, protection, electromagnetic transient evaluatio n, artificial intelligence and c o ~ p u ~ a t i o nin~elligence a~ in power system operation and planning. Professor Wong has published over 140 research papers and has been awarded the Sir John Madsen Medal of the Knstitution of Engineers Australia. He was the Founding Chairman of the Western Australia Chapter o f the IEEE Power Eiigineer~ng Society and was the Chairman of the Western Australia Section of the IEEE from 1999 to 2000. He h a member of numerous technical committees for intema~i~nal 2000 E ~ co~~erences~ r Wong was the General C h a i ~ a nof the IEEE ~ ~ S / C S Inte~ational nce on Power Systems Technology powerc con 2000). We is an editorial board member of the interna~ionaljournal Electric Power Systems Research and Jour~~ul of ~ ~ t e f l i g~en~f of r ~ u t i oProcessing n Syst~ms.In 1999, he was uts~ndingEngineer Award of the lEEE Power Enginee~ngSociety WA Chapter. He was a recipient of the IEEE Third Miflennium Medal in 2000. Professor Wong is a Fellow of the Hong Mong ~iisti~ut~on of Engineers, Fellow of ~nstitu~ion o f ~ngjneers Aus~alia,Fellow of the IEEE, and Fellow of the TEE. cquired her degree of Bachelor of Engineering in Electrical ee S The University of Hong Kong in 1996. In the same year SYS Miss Yuen was awarded The China Light & Power Company Prize in Electrical Energy, ‘The applic~~ion of ~ i ~ c i a l because of the dis~inctionof her final year project en~~tled neural n e ~ o r k son the detection o f high i~pedancefaults’. During 1994 to 1998 Miss Yuen pursued the degree o f Master of Philosophy with a thesis entitled ‘Fault detection and oven tection in low voltage power systems’. In 1998 she was awarded the China Light Co. Led. Electricai Energy Postgraduate Scholarship. In the same year she was awarded John Swire & Sons Ltd. James Henry Scott ~cholarshipfor ~ngineering Studies at the U n i v e r ~ oi ~f S~athclyde ~ which enabled her to pursue the degree o f Doctor of ~hilosophyin Scotland. Miss Yuen is also an Associatc Member of the IEE. Her current research interests include the analysis of international energy markets, congestion m a n a g e ~ e ntransmission ~~ piicing and the application of i n f o ~ ~ ~tcc~nology ion in energy markets. received SB degrees in applied m a t h ~ ~ a t i cand s in electrical engineering and com~uterscience and MEng degree in electrical engineer~ngand computer science from the Massachusetts Institute of Technology (MIT), Cambridge, in 1995 and in 1997, respectively. He completed his PhD degree in electrical e n g i n e e r i ~and ~ computer

IT, c ~ n c e n t ~ t i nong e l ~ ~power r ~ csystem economics engineer^ network economics: underlying p~inciple§ is entitled ‘Electric independent transmiss mpany (ITC) and designing a ~ c ~ ~ tfor e cre~ r ~ research i n t ~ r e s ~include s ~ o d e l l i ~ofg energy markets as stochastic dyn epts for the 1°C and ~esigningsoftware tools for various en~rgym a ~ ~ e ~ has a strong b a c ~ ~ r o uinn control, ~ estimation, m a ~ h e ~ aresearch ~~~s, design and r ~ g u l a t economics. o~

Dr Loi Lei Lai City University, London UM

Restructuring of the ele&icity supply industries is a very complex exercise bas na~~onal energy strategies and policies, macroe~onomic develo conditions, and its application varies from country to country. It is i m p o ~ to~ point t out that there is no single solution applicable to all countries and there is a broad range of diverse trends, ~ ~ b e r a ~ i s ~ ~eregulation tion, (or reregulation) and pr~vatisationare all processes under the general label of market reform. Liberalisation refers to the ~~troduction of a less restrictive regulatory framework for companies within a power sector. This could deregulation, which is the modification of existing regulation. It can reregulation is a more accurate term than deregulation since new laws are on the industry with reguIato~watchdogs appointed to protect c o n s ~ ~~nterests. er I~eally, then, a true liberalised energy market would work within a set r e g u l a t o ~framework, overseen by a regulator and with no external political influence upon the particip sation is the sale of g o v e ~ e n tassets to the p n v ~ t e sector, by itself, ~rivati~ation is not sufficient to introduce competition into a reformed sector. ~ompet~tion will be the result of careful regulation of the privatised entities to allow new e n ~ r access a ~ ~ ~ to the ~ a r ~ e~ompetition t. is ~ u n d a m e n ~to1most market reforms and it is introduced in order to reduce costs and increase efficiency. There is considerable variation in the extent tion which is introduced. For example, competition could be introduced just n of new gene~atingcapacity and referred to as competitive bid din^ where the existing gen~ratingcompany invites contractors to tender to build, operate and sell d Alternative~yall licensed g~nerator~ ower to the monopoly at a ~ p e c i ~ eprice.

e allowed to compete to supply wholesalers or retailers through a short-term market ~ k or via ~ longer t ~ term contracts; this is called compet~t~ve g e ~ ~ ~ tThe i onext ~. vel i s wholesale competi~ion~ i.e. competit~onin the sale of electricity to wholesale ies for resale to a retail level or directly to final customers. This usually allows the ion at final c o n s ~ ~ mlevel, ~r n s ~ ~ e to r s choose their own s This is us~allythe very last ~ouseholdconsumers, is calle step o f the reforms, as it requires a complex information technology system because of the of small users involved. Retail c o ~ p e ~ t i oisnusually i n t r ~ d u c ~ the larger i n d u s ~consumers, ~~~ then the medium cons~mers und the world are currently in ~ a n s i t towards i ~ ~ more arkets. The changes were initiated by: 8 r~alisat~on that generation and dis~ibution nctions need not be mono a feeling that public service obligations are lion potential of competition; availability and fuel supply s ~ ~ b i l iand ty~ the develop men^ of new technologies in power generation and i n f o ~ a t i o ntechnolo

erican electricity m The ~ontinu~ng growth of competition i the 1978 passage o f the Public Utility latory P o k e s Act ( conservation measure, PURPA establ ducers (IPFs) to sell electricity to local regulated investor-owne~utilities (IOUs). were broadened s ~ b s t ~ t i by a ~the ~ ypassage of the Energy Policy Act of 1992 which requires transmission line owners to wheel bulk power [l]. Thus, under current fe(iera1regu~ationsnon-utility power producers can sell electrici~to any utility on the grid, n a policy F u ~ h e r ~ o rin e , April 1994, the California Public Utility C o ~ m i s s ~ oadopted establishing complete open access to all power producers. By 1996 ~ n d e p e n d ~generators n~ could compete to sell electricity directly to large industrid customers, ef~ectiveiy ~ d i t i o n autilities. ~ By 2002, the policy will pennit all ele consu~ers, of size, to purchase electricity any utility or independe rator on the grid. No longer will the consumer be res to buying e l e c ~ c i t yfrom the local utility. A ~ o ~ p e t i ~market i v e for gene~tionwill have been es~abli§hed[2,3]. The system evolving in the USA provides i ing competition and div~rsityamong gen~r~tors. They vary from established utilities and co-generators to small producers that use renewable fuels and other non-utility genera~ors y 1990, a decade after reform movement got under way in the USA, co-gen enerating capacity than were the ~aditionaiutili C a ~ ~ f o r Edison n ~ a buys 30% of its power from NUGs. in M ~ c ~ ~ consists g a n of 12 gas turbines with a generati C ~ ~ p a nin y ,~ i z o n ais , an indepen~~nt power b~~~ custo~ers,n a ~ e l ythe Tucson Electric Power Edison 141. Compared with the deregulation of I 0 monopoly requires a complete and ~ d ~ chane n ~ ~

Energy Generation under thc New Environment

3

property rights in the electricity supply industry in order to obtain the benefits of increased efficiency and innovation. A shift from public to private ownership refocuses the goal O f the producer towards profits. Pursuit of the latter provides a strong econon~icincen~ive,in a competitive environment, to improve and maintain the quality of customer services, monitor costs more closely, and invest in productivity-enh~cingt e c ~ o l o ~ i eThese s~ incentives are blunted by state ownership. With respect:to privatisation, the since 1989 seems more germane than does the regulatory reform the USA has been undergoing since 1978. The European C o i ~ is ~addressing i ~ these same issues and has agreed to draft directives calling for open access in energy markets. As of January 1993, the E u r o p e ~ Commission seeks to let large users of electricity, those using 100 g i g a w a ~or more of power per annum (aluminium, steel, chemicals, glass and fertiliser producers), to purchase electricity from any supplier in the Community.

.3

Th

Competitive generation provides a market within which independent fimis compete on the basis of price to sell electricity directly to large industrial customers, and to supply electricity, via common carrier transmission, to distributors who in twn sell power to final users [5,6], Produc~rsmay specialise or diversify by load characteristic. For example, some may prefer to compete for long-term base-load contracts. These firms are likely to own hydro and nuclear power plants. On the other hand, f m s with fossil fuel plants might seek to supply base and cycling loads. Finally, producers with gas combustion turbines and co-generators could compete to meet peak loads. Other firms may diversify and be ready to compete for base, cycling and peak loads. Prices charged for each type of service (peak and off-peak load, daily to ~ e a s o n a l ~ could be established by contract, 24 hour advance notice, and in spot markets. Unit could vary by the amount of electricity purchased per period. As a result, customers face more service options and a more complex pricing scheme. There are a nu advantages to having a variety of types of generators linked to the transmission grid. The first major advantage involves cost savings. At any given moment^ power is supplied to the transmission grid by the firm with the lowest marginal costs. according to merit saves resources and reduces the cost of generating electricity. Because the different plants may have different load characteristics, peak and load duration curves, generating capacity can be more fully utilised and additional capital resou~cessaved. The second ~ d v a n of ~ competitive g~ generation is that a spot market for electricity will develop. The ability to sell electricity on the spot market increases the ge~erator’s ~exibilityin scheduling production. The presence of a spot market means that less idle capacity must be maintained in order to provide a given level of service re Shortfalls and emergencies can be met by purchasing power on the spot market. and supply are eq~libratedby flexible spot prices. The third advantage o f competitive generation is that the market will provide an anray of service standards that more closely match consumer preferences. Consumers could be offered priority service with a schedule of electricity rates increas~ngwith the level of reliability. According to reference [7], priority service offers significant efficiency gains over random ration~ngwith fixed electricity rates. A compet~tivemarket in elec

Power System Restructuring and ~ e ~ e ~ l a t i o n

generation would offer a much broader m a y of services than do state n i ~ ~ o p o l i eors r e ~ l a t e dgenerators. erhaps it is not surprising that 70% of USA private utilities, facing new c o ~ p e ~ ~ tpressure ive at the generation stage, now offer some form of voluntary inte~up~ible service 181. The fourth advantage of competit~vegeneration is innovation. C o m p e ~ i ~ not i ~ n only leads firms to be more responsive to consumer demands, monitor costs more closely, and compete:on the basis of price, but also provides an incentive to be i ~ ~ o v a t i vDevel e, a new c o n s u ~ e service, r a better method o f reducing costs, or a faster way of d e a l ~ nwith ~ pro~~em promises s the innova~ora competitive edge.

xis~i The nature of the existing generating plants will affect the speed of reforms. In countries where the coal industry has dominated the economy there has been opposition to r e s t ~ c ~ r i nthe g electricity industry, which usually includes a s u b s ~ t i a al ~ o ~ofn coalt fired capacity. Deregulation of the electricity sector meant loss of a secured market for coal w h i ~ hnow has to compete for its share in the market. The nuclear industry in the UK was initially excluded from competition and subsidised. The nuclear power s ions bid into the power pool and were electricity due to the n-Fossil Fuel Obligation (NFFO). The on the distribution companies to buy a set percentage of their electricity from stations using non-fossif.hels. In 1990, this was mainly nuclear power. A Fossil Fuel Levy was placed on the e l e c ~ i cbill i ~ of all electricity consumers (which ~ ~ u n t toe 10% d of the total bill) and over 90% of the money collected was given to Nuclear Electric to cover gen not recouped from sales of electricity to the pool [9]. In 1996, when British formed, the subsidy to the nuclear power industry was abolished. The levy and since then it has been used to support renewable energy projects. Prices tend to go down as competition is introduced and are expected to fall sign~~cantly in the long-term. For example, in the UK prices have fallen since the market open in^ and they are expected to fall even lower. In 1995 real prices, the price of elect~cityfor industry decreased by almost 13% and the price for households by 6.3% between 1991 and 1995. It is has been observed that i n d u s ~ prices a ~ have decreased more ousehold prices in most of the countries where reductions have occurred [IO]. ne of the conse~uencesof p~vatisationis the ~eve~opment of the i n t e ~ a ~ i energy on~~ c o ~ p a n yconcept - a company whose focus is becoming more global and more multile US electricity and gas companies have been ~ u r c h ~ electricity in~ Australian and UK companies have been heavily involved in setting r projects in developing countries. Another change with privatisation older value. Privately owned companies have to compete for funds in the capital market and it is important to show that they operate efficiently to do well in the business environment to attract investors. That means a comple~elynew organisational structure and strategies for companies from what were used in the highly r e ~ l a t e dpower industry. Goal is expected to retain a strong position in power generation worldwide in the future. In 1995 solid fuel, mainly coal, accounted for almost 40% of world electricity pro~uctionand is expected to retain this percen~geuntil 2020. In 1995,60% of total world

Energy Generation under the New ~ n v i r o ~ e n t consumption was for power generation and this is expected to grow to 65% in 2020. The emand for coal will increasing~ybe dominated by Asia. expect^ to increase from 25% in 1995 to 43% in 2020 E1 11. There are a number of issues that will affect future use of coal and in some cases the results are quite u n c e ~ i n .The Inte~ational Energy Agency (IEA) points projections of coal use are subject to the outcome of competition between coal urope, and to the policies adopted by governments to improve nnance and comply with greenhouse gas reduction c o ~ i ~ e [InI].~ s

In the past, power systems were developed to transmit large amounts of power at hi voltage from remote generati~gstations and to diskibute power at lower voltage down millions of small consumers. This was the favoured pattern, allowing ever-l~gerpower stations, mostly coalfired, to be built and achieving economies of scale and high efficiency. The national grid evolved to ensure secure supplies to all consumers and centralised conkol and supe~isionwas essential. In the present privatised electricity supply ~ n d u s ~ based on free trading of electricity as a commodity, central control is unwelcome. er ever possible, electricity generation should be closely i ed with space and stems. newab able process heating in a diverse array of combined heat and power energy sources should be harnessed by large numbers of wind and wave machines, marine t i d a l ~ ~ u r r eor n ~s m ~ ~ - h y d rplant, o solar photovoltaic generators on roofs and small generating plant close to farms supplying wood fuel or to sources of combustible waste products. Generating plant will be small and dispersed and since CHP systems must be located close to their heat loads there will be a natural tendency for most e ~ e c ~ c i ~ generation capacity to lie close to the consumer. There will be little need to transmit large amounts of electric power over long distances. The h c t i o n of the power system will be to handle the f l u ~ ~ a t i o in n s load and in the output from the renewable power generators. ~ i g ~ - p o w long~distance er~ kansmission will be much less important, In the current energy structure, a central power plant is the key facility providing energy for houses, factories and offices. With decentralised co-gener power and the d ~ l o ~ eofn renewables, t this situation would change. would be less centralised and more dispersed. Network stability and frequency regulation would gain in importance and energy storage would become very ~mportant.Ele genera~ionis provided by a large number of small units rather than a small number units, Co-generation is the generation, on site, of your own power and at the same time taking advan~ageof the exhaust heat from your gas turbine or other engine to meet on-site heat needs. Heat can be used to heat buildings, heat dryers, generate steam ~ o u an~ h HRSG (heat recovery steam generator), or to provide air-conditio a b s o ~ t i o nchiller. Power and beat can be generated locally from na~ural using an efficient, reliable gas turbine. The uncertainty in the USA today is what will happen to electricity prices. The major c o ~ p e t i factors n~ are limited deregulation and lack of new generat~ngstations ~ ~ c u ~ a r l y large coal or nuclear stations). Estimates range from modest decreases in prices, to the levelling of local inequities, and significant increases driven by demand without supply. Our view is that prices over the long haul will increase slightly with some local equities

Power System Restructuring and ~eregulation

being eased. All this means that for many sites cogen (distributed power) will be a viable option for those willing to improve their competitive position through ~ e d ~ c e dn e r ~ costs. New enabling technologies have now improved transport of eleclkcity in ~ ~ g h - v o ~ t a g e C systems to the point where this may be cheaper, and use less energy, than ~ a n s p o ~ i n g fossil fuels, for distances o f 5000 km and above. This might make it possible to link lowCOzpower sources where demand is low to distant regions where demand is high.

1.4.2

Trends in Conventional Electricity Generation Tec~n~logies

Co~ventionalsources of electricity supply will m a i n ~ their i ~ central role in ~ r i ~ a r y energy supply for many years to come. Further advancement of fossil fuel generation technologies will increase the options for mitigating greenhouse gas (GBG) emissions. This is particularly important for some developing countries and transitiona~economies with abundant, low-cost fossil fuels, where electricity demand is increasin~rapidly. The large share of nuclear and hydro in the generating mix of some countries already makes a s i ~ i ~ c a~ontribution nt to mitigation of GHGs.

atio World electricity production is expected to grow by an annual rate of 3% in the period 1995 to 2020 according to IEA projections. Coal retains a strong position in world power generation and will continue so. However, gas is expected to grow faster at 6% than solid fuels at 2.9% (e.g. coal) [I 13. This is because, in countries where gas is available at competitive prices, gas-fired plants are cheaper to build and operate. D e r e ~ l a t ~ ohas n played a role in opening the way for gas to compete with other fuels. Coal is still the favoured fuel in locations close to low-cost coal production (e.g. p a t s of North America, Australia and South Africa), in areas where gas is unavailable or expensive (as in those deveIop~gcountries that have coal available, like China and India), and in areas where there are existing coal-fwed units. Prior to deregulation, utilities tried to predict the future energy demand in their area and build new capacity accordingly. In a deregulated energy market gener current demand is and try to fill as much of the demand as possible plants. The predicted growth in the demand for energy on a wor provide an incentive for generators to build new plant or extend their existing capacity to take advantage of this trend. Competition rules will determine the market players. However, the only players in practice who can invest in new capacity are those who feel they can achieve a competitive advantage. In deregulated markets this should not be market access or cost of capital but a genuine advantage such as feedstock, technology, captive market of heat, extension of existing plant to take advantage of existing assets, refurbishme~t,etc. The possibility of having stranded costs would seem to rule out new, ensive power plants. Most of the additional capacity is expected to come from incremental i n v e s ~ e n t in extensions done as part of general ~ p ~ v e ~ e or n t s ma~ntenance.New plants are likely to be smaller, more cost effective, and close to areas of demand that can compete effectively for local market share. This means that there could be a swing away from large fossil-fuel-fired plants in the ene y mix towards sma~ler,less

-

-

Energv ~ e n e r a ~ under ~ o n the New Envirolment

7

intrusive plants sited close to the area o f demand. The fact that industrial sites are now allowed to install their own genera~ngcapacity and export electricity to the grid could lead to an increase in smaller scale distributed g ~ e r a t i n gcapacity.

1.5.1

~ Q w ~ r

The operation o f power plants is also changing dramatically in dere Generating companies are no longer obliged to generate electricity; generate and sell their electricity when they think it is profitable for them. This means that most of the generators will want to operate their plants at base load where most profit can be made. There is little incen~vefor the generator to provide electric~tyfor more expens~ve intermediate and peak demand, which make up only a small portion of the market. As d e r e ~ l a ~ i proceeds on an increasing number of players enter the system which is no centrally controlled. This makes the quality and reliability issues more difficult to m Experience so far shows that deregulated markets can reliably meet demand and are expected to do so in the foreseeable future. The UK system’s re~iabilityand availa~i~ity actually increased between 1992 and 1997 when the transmission and dis~ibu~ion network was restructured [4]. It i s believed that the system will work without problems of security of supply for the next 5-10 years. Coal contracts are also affected by changes in power plant operation. There is a general move to shorter term fuel supply contracts to match the electri sales contracts in deregulated markets. ~ ~ e x i ~ini plant l i ~ operation i s an adv the competitive small-scale units market where conditions change quickly. Distributed gene could also give more flexibility to the system. An advantage of coal is the fact that it cm be easily stored in stock~iles,whereas storing gas is much more complicated and expensive and restricted to certain quantities. In deregulated markets demand and a v a i ~ ~ ~o fi l i ~ dictable and therefore the risk of disruption in fuel s u ~ p l yis more es can ensure security of supply for the generator.

Utilities are forced to operate in a more reliable, economic and efficient manner and plan their expansion investments more accurately. There are a number of reasons promoting int~rcon~ec~ions among utilities. These include economic interch~nge,Brm power and energy transactions, wheeling, improved operating reliability and ~ ~ x i b iand l i reduction ~ in installed generation reserves. Usually utilities construct new power plants to meet the increas~ngdemand or to rep~aceold plants, which need large investments, ~ o w e v e r ~ ~ t ~ r ~ o ~ ~utilities e c t emay d jointly install a generating unit in which the utilities may have different or similar shares or the interconnected utilities may buy a certain perce the output of a generating unit, which already exists in the other utility, Therefore, the failure of a jointly owned generating unit will cause a decrease in the available capacities of all the sharing utilities simu~taneous~~. Because of this correlation, the conventional model of a ~eneratingunit cannot be used to represent a jointly o ~ e generating d unit, The re~iability modelling and evaluatio~methods of composite ~efierationand transmission systems need to be extended when the system being analysed includes generating units that are jointly owned with other interconnected systems. This is because

Power System R

~

~

~and c~ ~ r~e g ur l ~it i o~n ~

the modelling of jointly owned units causes two major problems. The first problem is that they cannot be included in the area generation model in a conven~ionalmanner because a jointly owned generator contributes generating capacity to two or more areas. Consequently, a failure or derated state of a jointly owned generator affects all the sharing areas. This condition cannot be incorporated in the traditional generation model, which has an inherent assumpt~onof independence among generation models of various areas. The second prob~emis with the transmission model. In the absence of jointly owned units the transmission links are used only for emergency help and energy transaction^^ Since the ontracts and the transmission c ity states are fixed, emergency help th n e ~ g h b o ~ nareas g is fixed. when jointly owned units are includ reliability analysis of the system, common generation flows are present and vary depending on the states of jointly owned units. Consequently, the emergency help that can be given to neighbouring areas is dependent not only on the tr~nsmissioncapacity states and energy contracts, but also on the common generation flows which vary according to the states of the jointly owned generating units [12,13]. Further research on a detailed system representati~nis necessary to consider the particular operating features of jointly o w e d units so that their impact on the reliability performance of the respective power systems can be i n v e s ~ i g a t ~ . It is impo~antto ~nderstandthe market response to the increased risk associated with the introduction of competition into the market for generating electricity, Typically a v e ~ i c a l ~iyn t e ~ a t e dstate monopoly deals with fluc~ationsin demand and r ~ d o m equipment failure by carrying excess capacity, including redundant backup capacity. It may also address predictable fluctuations in demand by offering peak-load pricing schemes, although the incentive to do so is weakened by state ownership or regula~on. Competitive generation produces at least two additional sources of complex pricing structure, and loop flow problems when independent electricity into the transmission network. Moreover, electricity flows along the path of least resistance. Thus, for example, electricity sold by Generator A to Industrial Customer may not travel along the ‘contract path’ that is, the shortest line within the network tha directly links the buyer and seller, Depending on circumstances, electricity introduced into the network at any point may give rise to ‘loop flow’ affecting ail suppl~ersto the grid. Loop Bow can disrupt the quality and reliability of service to everybody taking electricity from the grid at the moment additional power is introduced. If decentralised markets introduce additional risk, they have to provide a bro ways of dealing with it. All of these sources of risk potentially influence the service to the final consumer of electricity. In general, the market offers methods to reduce risk and to price risk so that it can be spread or shared optimally. Consider how a generator faces the risk of uncertain prices for electricity. Firstly, the producer can sell power by long-term contract to large industrial customers and regional distributors. ~ o n ~ a cspecifL ts prices and adjustment clauses. Thus, only a small proportion of its output may even be exposed to unknown price fluctuations [ 141. Se the spot market on a regular basis offers normal returns because prices mean over a large number of sales. By selling regular~yon the spot market, the producer is reducing risk through diversification. Thirdly, the producer can hedge spot market sales in the futures market.

Energy Generation under the New Environment

Fuels used to generate eleclricity are produced using the follow~ngfuel sources: namely, coal, nuclear, natural gas, ail, hydrogen and renewable resources. ~ e n e w a b ~resources e include hydro power, geothermal, biomass, wind, solar and p~otovolt~ics. Coal is the predominant fuel source. ~ u c l power e ~ is projected to decline her over the next 20 years owing to retirements of existing units, Generation from both natural gas and coal i s pro~ectedto increas~to o€fset these retire~entsand to meet the growing demand for e l e c t r ~ c iThe ~ . coal trade has been increasing and is expected to continue doing so in the future. It is expected to increase faster than coal production. Between 1992 and 2010 the coal trade is projected to grow by an annual 4.3% whereas coal product~0nwill 2.3% a n n ~ a l ~[15j. y Coal prices dropped during the 1990s in line with compet~t~on and with the fact that there is excess capacity for mining coal for the international market. Cheap coal i s seen as being readily availabie in the short and medium tern. The ~ollowing sections s ~ a r i s the e discussions of issues related to the markets for coal nuclear, natural gas, oil and renewable fuels, followed by electric power industry res fuel markets. Goal Power generators will attempt to pass on market risks to coal producers and carriers wherever they can. As a result^ coal purchase contracts will ~ i k e ~become y s ~ ~ in~ e r duration and lower in price. The existing capacity of the power industry in each country will play an important role in its ~ t u r fuel e mix. In the EU, 17% of the conventional thermal capacity is over 30 years old, indicating that much of the plant is in need of refurbishment or replacement [16]. Where coal-fired plants already exist it is usually more economic to operate them rather than build new gas-fired capacity. Refurbishing or repowering an existing coal-fired plant can reduce costs as the entire i n f r a s ~ c ~ uremains re in place. Retrofit of pollution ~ n t r o ~ e ~ ~ i p m e nmay t be necessary to meet environmental standards. In cases where h y d ~ e l e ~ ~a ~n ~ oi nuclear tr y power dominate base-load generation other fuels notab~y coal and gas wiIl compete more strongly for position in the mid-merit market for electricity,

-

~

wer plants are expected to become uneconomical. ~ o m p e ~ i t i ev el e c ~ i c ~ ~ prices may be so low that nuclear power plant operators will not ee enough income to enable them to recover the costs of operating and maintaining the ants and the costs of capital ~~provenients, such as steam generator replacements. In the immediate f u ~ e , some nuclear power units will be at risk of early retirement as a result of r e s t ~ c ~ ~The ng. additional inability of plant operators to cover a plant’s full costs, ~ n c ~ u d ~capital n g costs, under restructuring produces ‘stranded costs’. For nuclear plants, operating costs after deregulatio~will be driven mainly by plant size, age, capacity factors, and requirements for new c a p i ~ limprovements. Average fuel costs make up only about 0 n e - f o ~ hof the operating costs for nuclear power plants, but the competitive environment created by a r e s ~ c ~ electric e d power industry will encourage nuclear power plant operators to ~ d u c ~ all o ~ e r a t i ncosts, ~ inc~udingthe costs of purchasing and managing nuclear fuel. ore over, if early retirements of nuclear power plants result from competition in electricity markets, the deniand for nuclear fuel will be reduced. To compete, suppliers in the n u c l e ~fuel

Power System ~ e s ~ c ~ rand i n~ge r e ~ l a t i o n

0

~ d u s t r ywill be forced to reduce prices or improve efficiency, In 1996, 434 nuclear reaclors in operation in 32 countries produced 2400 TWh of electricity avoiding an estimated 10%of global human-made emissions of carbon dioxide. S

gas is primarily used during peak demand periods and is the prefe~edenergy source for new generating capacity. The electric power and natural gas industries are both network industries, in which energy sources are connected to energy users through ~ s m i s s i o nand distribution networks. As the restructuring of electric^^ m a r k e ~ proceeds, the develop~entof htures contract markets and electronic auction markets could lead to greater integration of the electricity and natural gas industries and the em~rgenceof competitive energy markets. The availability of market information and public markets for natural gas and electricity will be a key to the development of an integrate for those commodities. The use of natural gas in electricity generation has been growing rapidly. According to the IEA World Energy Outlook, gas-fired e l e c ~ c i t youtput will almos~double ~etween 1993 and 2010, even under an energy savings scenario. Low capital cost, short construction time and competitive fuel price make natural gas generation attractive, especially in deregulated markets. Technologies being in current c o ~ e r c i a l operation are gas turbines and gas engines. The rapid devel o f gas turbines in recent years - bringing higher efficiency, lower cost, reduced NO, emissions and increased ope~ationalflexibility . puts natural gas electricity generation tec~ologiesin a position to make a large contribution to GHG mitigation. For large gas turbines, complex cycles (Le. reheat, intercoo~edcycles, etc.) may hrther improve efficiency. Gombined-cycle power plants attained thermal efficiencies of 40% in 1970, and are now close to 60% ~ ~ i c i e n t , Gas turbines and gas engines for small-scale generation need firther to improve their e ~ c i e n c yprice , and e n v i r o ~ e n t a performance l to gain wider application in the market, Conver~iontechnology using electrochemical reactions, namely he1 cells, should become competitive in the near future. Natural gas-fuelled fbel cells can attain 50% e f ~ c i e n ~ y (under very h i g h ~ t e ~ p e r operation), a~e which would be further i ~ ~ r o v to e d70% if used in combined cycle.

a Oil prices have ranged between US$l0 and 20 per barrel during the 1990s and &ere is no sign of any shortage in the short or medium term. Owing to assumptions about electricity industry restrucbxing prompting the construction of Iess capita-intensive and more efficient natural gas generation technologies, the share of coal generation will e ~ e n ~ a l l y decline while the natural gas share will continue to increase. With the d e r e ~ ~ a t i oofn electricity generation and the resulting incentive for power generators to lower fuel costs, the use of relatively expensive residual he1 oil for electricity production is likely to decline even fisther. As a result, petroleum refiners may be faced with a growing ~roblem:that is, how to dispose o f leftover residual fuel and petroleum coke. Among other options, two po$s~bilitiesare related to electricity markets: (1) selling petroleum coke to e l e c ~ c i t y generators for use as a fuel component, and (2) gas~~cation at the refine^ by using integrated gasification combined-cycle (IGCC) technology to produce steam for process heat and for electricity production.

11

Energy ~ e n ~ r a t i under o n the New ~ n v i ~ o ~ e n t

~ e c a u s ee l e c ~ c i t ygenera~ionfrom renewable sources generally is more conventiona~sources, constrained competition in electri result in a reduced role for renewables. As a result, a variety of propos~~s, schemes and policies incIude specific ~rovisionswhich are used to s~pportthe c ~ n t i n ~ e system ment and use of renewable energy. Renewable portfolio standar ing and charges are among the programmes being considered. Green pric~ngprog~ammes,already being im~lementedby electric utilities, may also provide a se c o n s u ~ demand ~r for electricity from renewable fuels. The role of y sources in competitive electricity markets will also depend an the cost of the indiv~dualrenewable fuels. In addition, because renewab~ee ~ e r g y generat~g~acilitiesgenerally depend on the availability of energy resources at s p e c ~ ~ c sites, often at sites remote from major electricity grids, transmission issues will affect the pene~ationof renewable fuels in the electricity ~enerationmarket. e an essential element of the climate change p r o g ~ ~ ssions and ~ ~ ~ i ~ clower a n levels ~ l y of other ~ ~ l ~ u t a n ~ s ~ ort for renew~bles,policies and prog stry to become comp~titive. supply a proportion of renewable powcr ren~wable~eneratorsc ~ n ~ ~ e that n c ethere will be a market for their pro ren e~~ctricity genera~ionschemes, using established te Pro power at prices which are more or less competitive mains~eamcoal and gas. Figures 1.1 1.2 show the changes in the arke et shares and the geneTa~ionmix ~ ~ s p e c t i in v ~the ~y

Links First Hydro Others 1pp* 7%

7

1%

,/-1%

lPPs

-. .

National

.

Energy 18%

1.1 C ~ a ~in~the e market s shares

\-Mission 4%

12

Power System Restructuring and Deregulation

8%

33%

igure 1.2 The generation mix

Althou~ha number of the technologies are inherently small-scale c o m p ~ with e ~ central station power generation, this has some distinct advantages suc d operation. As electricity markets are restnuc is likely to expand and renewables will become There is a more diverse range oE techno~ogies, 0th their technical and economic devel energy crops, p h ~ t o v o l ~ i cfuel s , cells, ass residues, wave power and geothetrnal energy. The world is c h a n ~ i nand ~ es are driven by the use of energy.

n operates on a small to m d combined cycles can also a d v ~ ~ of g er e m o v ~ ~allg p ~ i c u l a t e sfrom the co iency of over 85%. This teclmology is close Eurther ~evelopmentis the fuel cell, where a version at conve~ingche i authorities, difficult f trade or promotional orga~isatio~s w tariffs for sale of bi ackup electric supplie costs also consti~teserious barriers.

Energy Generation under the New Environment'

6.6.2

13

Fuel Cell

A fuel cell consists of two e~ectrodessandwiched around an e ~ e c t r o l ~ e ~ over one electrode and hydrogen over the other, generating electricity, water and heat. Fuel cell systems will compete with other distributed generation technologies, inc~uding micro~rbinesand reciprocating e n ~ ~ iavailable ~s, at prices competitive with e x i s t ~ g forms o f power ~eneration.Fuel cell systems will have a competitive advan~agein that they can be more easily scaled to residential size and will be more efficient in handling the load profile of residential customers. They will be quieter, e n v ~ o n m e n ~cleaner, l~y more efficient, and less expensive to install, service and maintain. Fuel cell systems will also te with solar and w~d- ower red systems. enerative fuel cell technology, National Power rece developed a new electricity storage technolo change the way power systems o f the future are planned and opera the world's most ad~ancedr e ~ e n e r a t ~fuel v ~cell t e c ~ o ~ o gRe y be attractive as a closed-loop form of power generation. Water is separated into and oxygen by a solar-powere~el water. The water i cell, which genera~eselectricity, h solarpowered ~ ~ e c ~ o ~and y sthe er The e ~ e ~ ~ ~ o ~process, h e m ~which ~ a l operates like a giant rechargeable b a ~ has ~ the e ~ ~ potential to deliver commercial, operational and environmen~~ benefits for electricity suppliers worldwide. It stores electricity when demand and costs u e low and releases it when demand and prices are high, removing the need to call up more expensive pawer plants. The system, which can deliver power instantly, can therefo~eassist deman~ planning, improve the use of power station assets so that less capacity is n~eded,enhance operational. control and give customers greater security of delivery. It will also offer lower lifetime costs than convent~onalstorage. The single biggest i n v e s ~ e n ~ Regenesys is that: it will offer lower lifetime costs than either pump ry plants - energy storage that could curtail peak demands stored, power electronic developments offer 8 fast respo Work and electrical DC energy stored in batteries. These considerations underline the potential value of energy storage in curtailing daily peak periods and that it would most e~ectivelybe located near the source of load variations, the consumers in the distribut~onnetworks [183. Coupkd with advanced power electronics, storage systems can reduce h ~ o ~d iis ~co ~ i oand n s elimina~evoltage sags and surges. Most ~istribu~ed ene o n ~so ~ that, storage § y ~ ~ m can§ be made ~ u l t ~ - ~ n c t with ~ o nlittle a ~ or no ~ d d ~ t i cost, example, both uninterrutable power supply (UPS)and energy ageme me^^ applications can be served by the same equipment. In combination with renewable reso~ces,energy storage can increase the values of p~otovoltaic(PV) and w~nd-generated e ~ e c ~ ~ c ~ t y supply ~ o i n c i d ~with n t periods of peak consumer demand. Energy storage systems used to follow load, stabilise frequency and manage peak loads. ~egenesyshas a. number of distinct a d v ~ t a g e sover existing electricity storage technologies like hydro rind battery storage. It offers all the benefits of pumped-hydro, but can be located here on a power system thus avoiding environmental problems. Though similar to a battery storage plant, ~egenesysis much more flexible. Unlike a battery, the power o u ~ u ~ and storage capacity can be specified individually, Based on h e 1 cell technology, ~egene§yscan be built in modules to the required size ranging from 5 to 500 megawa~sof

.

Power System R ~ s ~ c ~ rand i n ~g~ r ~ g u l ~ t i o n

I4

capacity. It i s able to provide vital services to elec~icitygrids, ~ncIudin voltage control. Regenesys could meet peak demand and maximise inve allows better use to be made of the cleanest generating plant by reducing the need to operate less efficient peaking plant. It can also enhance the value of renewable generators such as wind and solar power.

1.6.3

Wind

Cunently some 50 countries have major wind power ins~allations~ Europe is presently the most important market but demand in Asia is growing strongly. Ease o f rapid installation (six to nine months) and a free local source of power make wind an attractive technology in developing countries. Over 1300 MW of wind-electric capacity has already been instal~edin Germany and more than 1000 M W is on-line in Denmark. The Danish goal is to provide 10 % of its elecwicity consumption through wind-electric energy by 2005 and more than 40 94 by 2030. At about 4 US cents per kW of installed power, electricity from Danish turbines now costs around the same as the average cost for electricity from coalfired power plant. However, there is no such thing as a single price for wind energy as the costs depend on both wind speed and the accessibility of sites. Wind-electric energy has the potential to supply 25 % of Europe’s electricity needs. Some countries could also export power to neighbour in^ countries.

Potential applications of PVs range €rom basic electrification for the 2 billion people of the world without electricity to the integration of PVs in building structures in deve~oped, urban areas. Customers need complete systems of PV modules, panels and arrays to provide electricity appropriate to their needs. Improved light-to-electricity conversion efficiency of individual cells is less important than reliable, integrated systems. The flexible thin-film amorphous silicon panel is at the forefront of PV technology. D i s ~ i b ~ t e d generat~onwith PVs has been tested to relieve substation o v ~ ~ h e and ~ ~ as ~ na gmeans to defer transmission or distribution system upgrades. Remote locations in developed countries are also prac~icalapplications for PVs. ~xamplesinclude water p u m ~ ~ n fence gq elect~~cation, and radio station power supply. PV is one of the most flexible technology s u ~ ~options ly available for electric power product~onbecause they can supply loads from several watts to megawatts.

More than 350 MW of electricity are generated by commercial solar-thermal power plants in the USA. To exploit s o i ~ - t h e ~ apower l hlly, broad~r coop~~ation g o v ~ ~ m e nelectric t, utilities and private industry i s ne~ded.The major investments ~ieeded to develop and market solar technology must be supported by stable ~ o n g - t re ~ e ~ l a ~ o ~ policies, which can only be provided by government. For example, in the UK recent

Energy G e n ~ r ~ t i under o n the New Environment

1

studies point to the need of tax equity to improve the economic ~ompetit~veness of solarthermal plants more than ~echnolog~cal ~reakthroughs.

World concern over carbon emissions, new domestic pollution regulations, ~mprov~ng small-scale technology, and the: prospect of open competition for energy markets are forces that converge to demand greater efficiency in energy generation - to lower h e l costs, iiicrease marketable products and reduce emissions. These forces argue strongly for a new paradigm o f dispersed, combined heat and power (CMP) plants that have double the efficiency and produce half the pollution. Although large units will continue to operate in the short term, most will eventually be replaced by new facilities and virtually all new growth will come in the form of small units. Readily available technologies now exist to combine the generation and supply of heat and power. By capturing unused heat energy, generators and consumers can, in effect, use the same fuel twice. Combining heat and power production reduces the net fuel demands for energy generation by supplying otherwise unused heat to residential, commercial and industrial consumers who have heating and air-conditioning needs. CHP technologies can be widely implemented. In almost every case, such teGhnologies will save enough money, now spent on fuel, to pay for their capital cost. By combining roduction and supply, 80 to 90 % of the useful energy in fuel can be put to beneficial use. When these plants extraGt steam from the turbines ar relatively low pressure to drive industrial processes or provide heat, they lose some electricity production, but capture all of the heat, eliminating the use of other fuel to make this heat. Total ef~ciencies can reach 90%, d e p e n ~ n gon how well the electric and thermal needs are matched or balanced. CHP takes energy from a central electric plant and distributes it to end users as steam, hot water and chilied water using piping networks. An increase in efficiency of 1% would result in a 2.5% reduction in CO, em~§sions,An UK study suggests that half of the CO, savings required up to 2010 can be met most casteffectively with CHP. CHP can reduce fuel use, cut emissions and save money. Policy makers should take a ~ ~ ~ asteps ~ i tov encourage e use of CHP. The technology is ~eadily available, has a net economic benefit and can cut fuel consumption and pollutant emissions in the e n e supply ~ ~ ~industry in half. There are many ways in which r e ~ l a c i nseparate ~ heat and power generatiQn with CHP systems can reduce emissions s i g n ~ ~ c ~ n tFor iy. example, producing 1 kWh of electricity, and a given amount of heat, from hard coal in a CHP system can reduce emissions by almost 30% compared with producin~both s ~ p a ~ a t efrom ~ y the sanie fuel. Using natural gas in the CHP system can reduce e~issions by almost two-thirds compared with generating the heat and power separately from coal. CHP meets energy needs and can save money for a wide range of energy c~stomersincl~dingpublic sector users - and also helps preserve the earth’s precious energy resources, reducing the impact on the environment of harmfbl pollutanls. The GHP shares of European power generation range from about 34% in the Netherlands to about 6% in Sweden, s ~ ~ g e s t i nscope g for large increases in some countries. Energy m ~ k e t deregulation could produce more favourable conditions for CHP, by increasing investmen< innovation and market entry, and decreasing the costs of backup power and natural gas.

structuring and ~

16

capital costs of these systems may deter

consumers’ inter

e

~

~

l

er such i n v e s ~ e n t under s

is fair. In some

renewable energy sources. In Italy, for example, new legislation requires that from 2001 all generators and ~ m p o ~ e of r s electricity will have to supply into the system a quota generated by renewable sources [X 91. The EU directive allows member states to n with public services where this is necessary in the general interest of the vided they comply with Community law. Examples could be an oblig~tionfor to p ~ c h a s ea certain percentage of electricity from r~newableenergy sources or an obIigation for distributors to supply all customers in their area at an equal

s been good value, and now it is even more so, with the UK g o v ~ ~ m e n tdecision ’s to exempt d-quality CHP from the Climate Chan starts in April 2001. This exemp will apply to electricity generated fro CNP and used on site or sold directly to other bus~e§§es.The govemm$nt belie~eswith f a fair and appropriate fiscal and r e ~ i a t o~ ~~ e w o r k ~ other measures such as negotiated ag~eementswith indu ewable genera~ionand efgcient CRP will be ~creased,This should deliver substantial increases in CBP capacity in the coming years. It should en govemment to a ~ ~ o u n cin e , the coming months, a new CBP target of around 10 of the draft Climate Change P r ~ ~ a ~that m would e resent more than ’s CHP capacity. Action by the UK government an essential to provide a market environmen~with incentives and penaiti that the new tec~ologiesbecome available at competit~vecost and in ample quantity. For dis~bL~ted generators, there have been concerns about treatment of ~ ~ s ~ i bgeneration ~ted o ~ i they do not by public electricity suppliers (PESs), especially distributed e n e ~ ~ t i that er the new a~angementsa dis~ibutedgenerat~rowned by a PES will be to formal arrangements with the distribution business in the same w ted generator. The same r e ~ u i r e m e ~tot p u b ~ i ~the h a ~ ~ ~ m minimise the risk of the ~S-ownedgenerator bein treated in a more favourable way than others. The p o ~ ~ t decisions ~ ~ a l set the economic framework in which n e ~ o r k sw311 d e t e ~ i success ~e or failure in meeting the target. Private deve~o~ers will install the CHP and the renewable energy plant if they see a return for their investment, If developmen~sare to happen, unpopular measures will be required, such as ssions, incentives for the deve~opmen~ of s u i i ~ n § ~~~ l a~t i o~nand s ~ the relaxing of restrictions imposed by ~ ~ ~ regulations. ~ ~ i Inn order g to meet the new o b ~ ~ ~ a tai ~uppIier on can either supply the requ~redamount of renewable e l e c ~ i cor ~ buy ~, upplier who fails to meet the obl~gationwill be required to make a government has recently announced the basis for its new renewable energy support mechanism. Suppliers will be able to meet their obl~~ation ~ ~ t h by er

~

chasing tradable green c e ~ i ~ c a t eA s .~ ~ e ~ i a ~ v e ~ y purchas~g~enewableenergy or by of their obligatio~~ able to buy out a ligaticsn and the associated inc e total cast of meet t ~ o to~ the g end ~ user. In addition, the provision of a le sources at p r e ~ ~ uprices m via the NFFO and also the D 's New and Renewable Energy ~ r o g r has ~ iresulted ~ There are a number of le~~slative and policy ~ e ~ ~ l o ~ mc e~net ns ~inl yhand h hat will impact on d i s ~ r i ~ u ~g e d~ ~ r a t i oand n influence its growth. The Utility Bill is aimed at p u ~ the ~ customer g first. The Bill will ~ntrodu~e i r n ~ ochanges ~ t to the I98 Act. These changes will include the in~oductionof new ~ a d i n g and seXling electricity, separation ofthe PES supply and dis~ibuti ion on s u ~ ~ l i etor smeet targets an renewable electricity. AI1 of these c e ~ m ~ l i ~ a t i ofor n s some if not all distributed generator^. In gove for ensunn that energy e f ~ c ~ e n c y d e p a r ~ e has ~ t a ' een ~inister'with responsibi~~~y targets are met. Tar~etshave been set in some ~ ~ ~ a for~ sourc~ng e n ~ene s renewable sources (such as wind) rather than conven~~onal genera~~on.

e~egulatio~ has led the e l e c ~ c i t yi ~ d u s to t ~focus a~entionon the costs of and provides incen~ives ors to reduce their costs and ~ ~ i r n itheir s e ri investing in smaller scale Capital costs, construction time, h e 1 costs, up r n ~ t e n a n c ecosts will d ~ c ~ s i oonn what p~antsare ~

osts ~ e on the ~ s pee c i ~~site ~ as~ well as the s ~ e c i ~ c a ~(size, i o n oper~tiona~ reliability, e n v i r ~ n m performance, e~~~ safety r e q u i r ~ m ~etc.). n ~ , Costs will be ant built an the gr~enbeltCO existing ~ i i f r ~ s ~can c ~ber used. e Plants close oses and avoid costs for CO rent sourc~sas each project is site s-fired plant can vary &om US$300 om ~ S $ 9 ~ ~for~ W e advanced c o a ~ - ~ r e

1

Power System Restructuring and ~ e r e ~ i a ~ o n

corresponding to the replacement of major plant components after 20-25 years, whereas coal-fired plants can reach up to 30-40 years of life. A l ~ o u g hgas-fired tecknology is cheaper in U S $ ~ W eterns there are other factors that should be taken into account. Natural gas is not available in every country and prices are not always competitive. Moreover the i n f ~ ~ ~ c ttou produce re and than the equivalent costs for coal. As discussed more capital ~ntens~ve upstream capital costs are considered in the competitiveness of gas coalfired plants then the capital expenditure associated with both ~echnologiescould be the ~ can outweigh the difference same. The high costs of the pipeline network to t r a n ~ p ogas in capital costs for plant construction. If in place, the electricity generator can benefit from d build cheaper gas-fired plant. However, as d e ~ a n d cture will be needed. It is estimated that to c in Europe (1.7% annual growth from t 999 to i n f r ~ s of~ US$l00-200 c ~ ~ ~ billion will be required [23]. Such inv unde~akeiionly in the fr~meworkof long-term contracts and it i s unc rofitable in competitive electricity and gas markets.

1.9.2

Technology Advances - Clean Coal ~ ~ c ~ n o ~ o ~ i ~ s

Clean coal technolog~esis a tern used for ~ e c ~ o l o g ithat e s achieve a higher effici~ncyand ns for converting thermal energy to electricity than conventio~alpul on (PCC) with subcritical steam and without emissions control. The also u s ~ dto include e ~ s s i o ncontrol systems such as 0, control equipment. Clean coal technol~giesare the way forward for coal as they can ensure compliance with the ~ig~tening env~onmentalstanda . There has been considerable effort to develop these t e c ~ o l o g ~at e scompe~itivecost eserves of coal are large and w i d e l ~~~stributed likely to continu~to be widely used, so more efficient and cleaner coal technologies (CCTs) are an i i n p o ~ a noption ~ in a future energy strategy. CCTs will enable the use of coal. with higher ene~gyefficiency and minimum e n v i ~ o n ~ e ni ~~paa~c t s . types of coal technologies applicable to large-scale power ~eneration: PCC t e c ~ o l o g ~ ewith s emiss control e q ~ i p ~ e(n ~ ng fluidised bed combustion C); pressurised fluidi and ~ntegratedgasi~cationcombined cycle (IGCC) t e c ~ o l o g ~ e s . status of these technologies today and a compa~son technologie~with gas-fired power generation in various c ing and construction of plants using these technologies worldwide (241. CCTs can also be used to repower existing coal-fired power stations a~proachingthe end of their lifetime, instead of buil~ingnew plant, and therefore r e d ~ c eoverall costs. ~ e ~ o f i pollution ~ ~ n g control equip men^ is also important as future and exis~jngcoal-fired plant may need to meet increasingly stringent environmentai standards.

nergy is one of the most critical r e s o ~ c e sfor that energy c o ~ s ~ p ~will i o nat least do facto~of up to five in the next 100 years. At present 1

-

19

Energy Generation under the New Environment

energy poses threats to the climate, with potentially severe enviroi~men~al consequences~ given the levels o f consumption likely in future, it will be an immense chal the global demand for energy without unsus~inablelong-term damage to the environment. This situation has attracted the attention of political leaders across the world, and at the Kyoto meeting of the parties to the UN Framework Convention on Climate Change in D e c e ~ b e rI997 there was agreement to tackle one aspect the amount of greenhous~gases emitted to the atmosphere. The levels of atmospheric CO,, for example, have increased from 285 ppm before the ~ndustrialRevolution to about 350 ppm now. Xt is now generally accepted that there is a strong case for acting to mitigate the threat of drastic clima~e change associated with the unrestrained continuation of this trend. The Kyoto meeting produced pledges by the industrialised nations to cut their GWG emissions, by 20 12, to an average of 5% below the 1990 levels. Deregulation could play a positive role by giving flexibility to different plants or even countries to trade emissions. In this way a generator could have a portfolio of plants including some using renewable energy and therefore meet overall environmen~al requirements. It could also help the development of less costly pollution coatrol technologies. In the single European electricity market, however, where electricity will be traded between member states, it is not yet clear where to allocate emissions. It could be the country where electricity is produced or where it is actually used, This is particularly important in the view of commitments to reduce GHG emissions. US e n v ~ r o n ~ e nregula~ions ta~ have caused a niajor shift in demand for lower sulphur coal supplies. Since the 1990 amendment to the Clean Air Act, there has been a noticeable shift in coal use by ~eneratingcompanies in the USA towards lower sulphur coal. ~ e r e g u I a ~increases i~n the o p p o ~ n i t i e sfor using CEiP, since the power ~ e n e r a ~ ecan d more easily be distributed and sold. GWP units can supply both electricity and heat at the same time, achieving high efficiencies and therefore reducing emissions to th compared with separate generation of electricity and heat. In all c o u i i ~ ~ s economical on industrial sites or community heating schemes where there is heat. In deregula~edmarkets industrial users can set up a small CHP plant on their sites to sup~lyheat and sell any surplus electricity to the local grid. Before deregulat~onthis practice was either not allowed, or at least not encouraged in many countries 1241. There are two ways to reduce GHG emissions. One way is to increase our r e l ~ ~ n cone nuclear power; the other is to develop a wide range of alternative methods of e x ~ a c t ~ n g energy from nature. The nuclear option is clean and feasible but it is hard to See opin~onwould switch from its present hostility to the acceptance of a massive pr of c o ~ s ~ c t of ~ onew n nuclear power stations. The role of nuclear power is ex nce decrease in Europe as the perception of its environmental and economic p e ~ f ~ ~ a has substan~iallychang~d,In the 1970s nuclear power was regarded as a source of cheap and em~ss~ons-free electricity. High costs invoived in decommissionin~nuclear reac~orsand the unresolved issue of nuclear waste have changed the image of nuclear plants. Italy has phased out nuclear generation since the early 1990s after the Chernobyl accident. ~ e ~ a n y decided in late 1998 to phase out nuclear power and is now d~seuss~ng possible ways for ~ nplans t to start phasing out nuclear power in implementation. The UK ~ o v e r n ~has It is clear that the construct~onof new nuclear plants in Europe will cause pubI~coppos~~ion and is unlikely to materialise, particularly in deregulated markets where such ~ n v e s ~ ~ n ~ s are not competitive, as they are too expensive. The contribution frorn nuclear power to the fuel mix is expected to decrease and will be replaced by other sources ~ne~uding coal.

-

Power System R e ~ ~ cand ~ ~i e rn~ ~ u l a t i o n _sl__

power, the power system must evolve to deliver 11 reinfor~ethe need 10 ensure diversity f b m ~ a market ~ ~ thed e n v ~ r o ~ e nimage t a ~ of fuels and t e decisions taken by developers and ~ o ~ ~ ~ ~foc i a n s ~ o m p e t i t ~ oinn retail will certainly create c o n s u ~ to ~ ~influence s ~ ~ v e l o p ~ e nAlthough ts. cost an ~~a~~f a c ~ ao~~f ~ c t i nc g~ ~ t o choice^ ~ e r as e ~ v ~ r o n ~ be ~ n f l u ~ c eby d the e n ~ i r o ~ considera~io~~ e n ~ suppliers are ~ ~ ~ n or~ have j n launched g environm e l e ~ ~ ifrom c i ~r ~ ~ e w a benergy le projects. An opinion poll in the 6% of c o ~ s ~ i ~we ro su prefer ~ ~ to buy e l e c ~ from c ~ ~renewable sources, but only 21% e p r e p a r ~to~pay more for it. In ~ a ~ ~ f o ran n ienergy a supply company has ~nergyscheme which gives eustomers the option to buy a part of or all om r e n ~ ~ a benergy le sources 1251. some c o u ~ ~there ~ e shas been opposition to the cons 1 r ~ a § o ~ §The . poor environmen~lim on. plants of an earlier g e ~ e r ~ ~ iThis erive from the pollu state-of-the-art pl act on new projects. where the r e s i d u ~are ~ re~sedin building ~ a t e r i a l scan ~ o ~ ~ n e ~ the ~romotionof CCTs and their excellent e n ~ ~ r o n m e nperfo ~al a role for coal plants in the re [24f.

f the ge~erating~ ~ n c t i ohave n chan d new capacity, and, if t h e i ~ ~ u d gise ~ ~ ~ t face an additio~al u n k ~ o ~that n is, . To ~ i n i ~their ~ s risks, e they ds and low unit capital CO e, it is expected that ne

ercut c ~ n ~§a ~ l ~u ~~ iicosts ol ~~and ~ have

grid is fin e l e c ~ ~~~ s~ ~l ~set ya of t eg e~n e ~ f i ~ ~ r s of c ~ s ~ o m[28-403. ~r§

Energy Generation under the New Environment

1

any studies indicate that distributed generation (DG) might play a s i g n ~ ~ c arole n t in the future power system structure. A study by the Electric Power Resea (EPRI), for example, indicates that by 2010,25 % of the new generation will be distributed [41]. Owing to variations in ~ o v e ~ m eregulat~ons, nt different de~nitionsfor DG are used in different countries. In England and Wales, the term ‘dis~ibutedgeneration’ is predominantly used for power units with less than 100 MW capacity. In Sweden, DG is oRen defined as generat~onup to 1504 kW.In Austra~~a DG is ofken defined as power generation with a capacity of less than 30 MW. In New Zealand, DG is often considered as generation up to 5 MW. There is no special definition of DG in the Californian and N o ~ e g i a nelectricity markets. A general ~ e ~ n i t i o forn DG could be an electric energy source c o n n e c directly ~~ to the distribution network or load centre. DG is decentralised and located closer eo the point of reater economic and env~ronmen~l sense. Several main reasons have combincd to make DG a technically, commercially, environmentally and, to an extent, politicalIy attractive proposition. Customers benefit from the success of DG because:

of The use of distributed energy will allow improvements in the dispatchab~li~ resources and improve the integrity of the ~nstnissionand dis~ribu~ion systems. Identii~cationand use of alternatives to power generation, transmission and systems controls will ~mproveload levelling, load manage men^ and overall power quality, The system will become more robust in its ability to tolerate natural disasters, suffer less damage and minimise the dependence upon the need for ~ ~ e d i ares~oration te of rhe grid system. Over~llsystem reliability will improve. To get a better unders~ndingof the possible fbture develop men^ of DG in a com~eti~ive market, some examples of typical DG applications are as follows: ~eiiewableenergy technologies, e.g. wind power or solar power. These projects receive certain subsidies, or customers might pay premium prices for renewable en Peak supply systems, based for example on emergency generators or on-site uch systems ~ ~ i c a lsell l y to the wer exchange for only a very short period per year to capture exlremely high peak pri CWP systems, e.g. district heating, whereby a high efficiency can be achieved and additiona~revenue from selling heat can be obtained. On-site generation based on microturbines or fuel cells. Electricity as well as hear are most likely to be used locally.

1.I U.1

Market Regulation

In competitive power markets, DG competes with cenlralised power generation. Hence, market regulatiosls should ensure that DG can act freely within power markets, similar to centralised generation. Tt is, however, often argued that most market ~ e g ~ l a t i ~used ns

Power System K e s ~ c ~ and ~ n~g ~ r ~ ~ l a

worldwide have been designed with large centxalised generation in mind and that, therefore, DG ofien faces significant barriers w~thinthe competi~~ve market.

1.10.2

The Power Pool

The power pool is used to create an efficient marketplace for trading electricity. The power is usually by a c e n ~ a ~ ~ s independent ed, or sation that defines the ards for ele rice bids and the eva~uationof thes s, as well as organising ns the bidding and eva~uationprocedure. The evaluation of power p o d r e ~ ~ a t i o regarding the t r e a ~ ~ofn DC t is a very complex issue. The main difference between various approaches for e l e c ~ c i t y~ ~ r k e ist sthat the trading o f electricity through a power pool (or power exchange) is optional in some a ~ , m a ~ d a t oin~ others, e.g. ~ n g ~ and a ~ d countries, e.g. in Nord Pool ( S c a ~ d ~ a v i and ~ a ~ ase well s as in the National Electricity Market in Austra~ia.In ~ a l i f o ~ i the a, ation in the pool market is optional, except for three large private utilities. They trade through the power exchange until the year 2002. The rea~onfor a regulator to set up a m a n ~ t o r ypool system instead of an op~ional market is usually to achieve a high market transparency~e.g. to prevent some large ~ ~ n e r ~ tfrom o r s gaining market power. In ge~eral,all market pa~icipantswill b e n e ~from t arent power market however, other options are also possible to prevent large rs g e ~ ~ market ng power, e.g. by splitting up the generators as was done in New The disadvan~ageof a mandatory pool approach is that all market p a ~ i c i p a n ~ have to join the pool. That leads to various fixed costs, e.g. members~ipfees, and or energy fees. Both fees are a way to recover the cost for the operation of the power pool. The me~bershipfee is usually a fixed annual fee and the energy fee is based on the energy a c ~ a l l ytraded via the power exchange. These costs may be a major b ~ndepe~dently owned generation companies that focus on DG to enter the electricity market. Therefore, exception§ to the mandatory rule were incl~dedin the re ~ n ~ l a nand d Austra~iafor small-scale generation. The exce~~ions depend on ~ n s ~ l l e d ver, there is no obvious reason for a capacity (30 to 50 M ~ of)the DG source. N source with a capacity of 25 MW to be treat fferently from one with a capac~tyo u ~ h e ~ o rtechnical e, limitations in a distribution n e ~ o r kmay aural iand an urban distribution network. Mence, regulations based on a certain installed capacity influence the way certain market pa~icipan~s to behave. The cost problem for p a ~ ~ c i p aint ~the g pool market, however emains, even if certain capacity limits are removed, This issue is of particular interest €or 6 concepts that aim at power generation, probably for only a few hours r year. To c a ~ ~the re ing e ~ ~ ~e o ma ~~p peaks, ~ c e these dis~ibuted~eneratorsmust p ~ ~ i c i p a t e change. Therefore, high annual fees can be seen as a major barrier for nerators to participate in a power market. As a solu~ion,the cost recover for o f the pool e x c h ~ g eshould mainly be based on energy fees, In additio~,it oned that within the national electricity market in Australia d i s ~ ~ u t e d to sell all generated power within the d ~ s ~ b u t i no ne ~ o r k[ i~cantlyreduces the market o ~ p o ~ n i t of ~ esmall-scale s gene~ation.With e treatment of the individual imbalance of each market ~ ~ i c i p isa ~ t ant for fluctuating power sources, such as wind or solar power. Such

Energy ~cnerationunder the New E n v i r o ~ e n t

t e c ~ o l o ~ i have e s the d ~ ~ a ~ that v the ~ ~power g e output during an upcornin urs, can only be pre~ictedwith some ~ c e ~ afor i ne ~ are three main ~ r o b l associated e~~ with the pool price:

ts effectively bypass the pool. a1 price is paid to all, it i s ~ a t h e m a ~ ~ c a l l ~ 3. Average pool prices bear no relation to any real price p hence of gene ratio^ has been falling steadily since 1990. until about 1994, s t e a d ~ and e ~ ~now seem set on an U

Figures 1.3 and 1.4 show the pool and an er at ion, but most renew cen~ra~ised low (33 kV or below) voltztge networks. So c l e c ~ i c i ~ wholesale prices, w h is wrong. The c h ~ a c t e ~ s to~ c s c o ~ p l eof~ bids i~ 001 capacity ~ a y ~ ~ These nts. lack of tr~sparencyin contracts for onsumer c o n ~ ~ e n c eAs . a result, n Contracts for

re I.

001

a

needed reforms

Energy Generation under the New Environment

5

e n f ta sv o ~ renewable and CWP enerators, are c o n c e ~ e dsince the a ~ ~ ~ e ~will and those gene~torswith ~nflexibl tors with flexible and predic~ble redictable o u ~ u will t face ill b e ~ from ~ ~thet oked at in a wider context. er d i s ~ b u t e dgenerators are likely to grow s coming years and the government has, therefore, paid cmfbl attention to t the economics of DC. It is i ~ p o ~ aton t new e l e c ~ market c ~ ~ that may adversely CHP, obtain access to the el e n s ~ r ethat ~ i s ~ b ~enerators, ~ ~ e d inch META, a ~ ~ g e i ~ e that n t s wil d ~ s ~ i b umarket ~ ~ o on ~ fair terms. As p to ana age their risks and achieve fair osals too, to deal with the needs of generators.

A n ~ i ~s l~a~~i care e sthose

nctions ~ e r f o to~ s~u pdp o the ~ basic services of c a p a c i ~energy ~ supply and power delivery. The costs for ancillary servic s i ~ n i ~ c a nfor t ; e x ~ p ~ine the , USA the total costs for ancillary services are about

and mark^^ ~ a ~ i c i p a nthat ~ s we able , the ancillary services are split up i

erat~onof ~ l e c ~ i yc ~istrib~ited i~ ~eneratorswith~ndistri issues concerning real and reactive r qua~ity[ 1 ~ , 3 0 , ~ 4 ] .

ution networks operate on a radial or open-r designed broadly on principle that load reduces along the I of each distr~~utor. d i s t ~ ~g~~ t~ ee r~a t effectively reverse th point on a distributor or interconnected network and this could affect c o n v e n ~ ~ Q ~ ~ 1 automatic voltage control schemes which cater only for conveneio the design of protect~verelaying systems i s much more complic going both ways. buted generators, such as the majority of wind gene~atorsa d sma1I”s~a~e ased on induction ~ a c h i n e which s have no stead~-sta~e reactiv generation c a p a ~ i ~There i ~ . is a need to import react~vepower to provide Geld exci ators, partic~lar~y s ~ ~ c h r o n ogenera~ors, us can lead to localised increases ich can potentially exceed the sho~-timeratings and ~ a ~ i ~n tg i n o~f s

~ o t e pn r ~o ~ l~arise e ~ ~with systems us create h a ~ o ~disto ic s y s t e ~ ss ~ b ~ etoc trap

inversion (e.g. PVs

erspective, the effect of DG is that networks will be~omemore active in le in behavio~r.From a g e n e ~ ~ o r ’perspective^ s althoug~it may to overall capacity, the c h a r ~ c ~ e ~ s of t i cge~erators s and tible, and network c o n s ~ a ~ could n ~ s result in ~enera~ors

I n c r ~ a s ~use d of CMP and co-generatiQn will result in lower usage of the ~ e ~ in? o r ~ terms of energy ~ a n s ~ o ~ ~ aand, t i o ntherefore, po~entiallylower lev~lsof income. The e to i n c r ~ s rather e ~hanr e ~ u c~ ae p ~e~ l olling a more compl~xand increasingly ~ ~ e r a t Qsuitably rs located may also offer benefits to a d ~ s ~ i b u t oby, r for ex o f f s the ~ need ~ ~for~ re~nforcementor provis~onof other s e ~ i c e such s as voltage ayments to generators will be s u ~ s t i ~ t i nfor g other ex~enditur

1.10. uppose t

is a need to replace a circuit breaker as the fault level. It is i ~ p o ~ a to n ts~ress

Energy ~ e n ~ ~ tunder i o nthe New E n v i r o ~ ~ n t

7

installed because of all ge~erators.The contr~butionof each generator can be re~dily onventional short-circuit analysis tools. These con~butionsto the s h o ~ ate the cost of replacing the circuit b such as there would not be a need to ear, cannot be credibly used to re entry to recover all system reinforcement cost. In this case, the distri rep~acesthe circuit breaker, and in the following price review peri system charges accordingly to all generators with respect to their con~ibutionin order to recover the system investment. The der~vationof charges for assets that provide the connection of a discre~eplant to the system should be differentiated from those for the use of the system. In the former case the asset i s provided for a sole user and could have been financed directly, and even owned, by that user. In this instance charges should be based on the histo~ccost of the asset and a fair return on the cost of the capital provided by the d~stributioncompany. In the latter case the assets are used by a number of system users, past, present and fbture, and charges should be on the basis of a tariff differentiated by voltage. The d ~ ~ arises c as u ~ to how reinforce~entcosts of the infrastructure of the system should be treated when a new user joins. There is also a d i ~ ~ with u 1the~ costs of s ~ r ~ d iend~ a s ~ c t u assets re when an existing user departs.

In regions where renewable energy resources are abundant but usually situated in remote locations, connection to the central power grid is expensive and in many cases ~ ~ ~to ~ c provide. Small-scale, autonomous generation schemes, on the other hand, are both economical and practicable. They utilise the energy resources available and supply the consumers in the local regions. The system cost can be reduced by using c a g e - ~ e sdf, excited i n d ~ c t i oge~ierators ~ (SEIGs) [47-521 since these machines are cheap and r ~ a d ~ l y available. ~utonomouspower systems often employ single-phase g ~ n e ~ t i oand n dis~bution schemes for reasons of low cost, ease of maintenance and simplicity in protect~on[53]. When a three-phase SEIG is used to supply single-phase loads, however, the stzator c are s e r i o ~ s ~unba~anc y causing degrada~ion in generator perfo o v ~ r c u ~ e novervoltage t, efficiency and machine vibration. These xtent by the use of the Steinmetz c o ~ n ~ can be alleviated to a c the excitation c a p a c i ~ c eand load are connected across different phases. For isolated operation, however, perfect phase balance cannot be achieved when the load is purely resistive. The objective of this case study is to introduce a modified ~ t e i n m ce o~ ~ e c ~ i othat n a s ewhich supp~iessingleenab~esperfect phase b a ~ a ~to c ebe achieved in a ~ e e ~ ~ hSElG phase loads. A general performance analysis is presented and experimental results are given to validate the princip~es.

~

~ ~ r ~ ~ ~ i ~t and ~

~

~

n ~

~~

~~

~~

l ~

Figure 1.5 shows the mdified Steinmetz connection (MSC) for a ~ ~ l ~ - c o n nSEIG, ~ted which supp~~es a s ~ n g ~ e - ~load. ~ a s eIt is assumed that the rotor is driven in such a d~~ection that it ~ ~ v ~the ~ stator s e swinding in the sequence A-B-C, i.e. in h e same direction as the positive-sequence rotating field. Hence, if A-phase is taken as the reference phase, B-phase is regarded as the lagging phase. The main excitation capaci~nceG2and the auxiliary load t e dB-phase (the lagging phase), while the ~ u x i ~ i a ~ r e s ~ ~ ~ a nRL2 c e are c ~ ~ ~ across excitation Gapacitance 6, and the main load resistance R,, are connected across A-phase (the reference phase). Compared with the original Steinmetz connection [54],it is no that the auxiliary load r ~ s ~ s RL2 ~ cand e a u x ~ excitation l ~ ~ capacit~ceC, have introdu~ed.These circuit elements provide additional current components that result in the flow of bdmced line currents into the SEIG. In a practical aut~nomouspower system, the auxili load resistmce RL2 cm be local loads such as lighting, storage heating or battery charging Alternatively, it could be a portion of the remote loads. For the purpose of analysis, all the circuit parameters in Figure 1.5 have been referred to the base (rated) frequency hose by introducing the per-unit frequency a and the per-unit speed b [55]. Thus, each voltage shown in Figure 1.5 has to be m~ltipl~ed by a in order to give the actual value and the per-unit slip is equal to (U b)/a. Besides, the motor convention has been adopted for the direction of phase and line currents. The ~hase-balancingcapabiIity of the MSC for a three-phase S E E may be studied by re€erence to a ~ o l t a g e / c ~phasor ~ n t diagram. It is assumed that the values of C,and C, are su~cientlylarge so that the SEIG has built up its voltage and is supplying the loads. Figure S.6 shows the phaasor diagram for the SEIG under balanced conditions. Because the is delta ~onnected,the line currents I,, I, and I3 lag the c o ~ ~ o phase n d ~ ~ voltages V,, V, and V, by (ld, f d 6 ) rad, where lli, is the positive-sequ~nceimpedance angle of the S E E . The line current 1, is contributed by the current Ia through C, and the current lR2 through RL2. ~ ~ e a n w h ithe ~ e line , current I, is contributed by -Icl(where Ictis the current through 6 , )and -IR,(where IR,is the current through RLl). It can be shown that the angle y between 1, and I, is equal to (4 2 d 3 ) rad, while the angle Sbetween -IR,and I , is (5x16 bP) rad. The phasor diagram in Figure 1.6 can be drawn only when la leads 12,which implies that perfect balance can be achieved for values of #p ~xceeding2x13 rad.

-

~

-

Energy Generation under the New E ~ v i ~ o ~ e ~ t

_/_3__

a

~~g~~~1.5 Modified Steinmetz connection for three-phase SEIG

From the current phasor triangles in Figure 1.6, the following relationships can be deduced:

For a given total output power, (1.1) to (1.4) can be used to determine the values of the load and phase converter elements required for perfect phase balance, provi and a of the SEIG are known. Equation (1.2) shows that B, vanishes when 53, = 5n/6 rad, which innplies that the auxiliary capacitance C, can be dispensed with. When #, exceeds 5a/6 xad, B, becoines negative, i ~ p l y i n gthat perfect balance can be achieved with an auxiliary induc~ance. In practice, however, the full-load power factor angle of an SEIG ranges from 2 d 3 rad to 4n/5 rad, and hence it is very likely that an inductive element need be used.

B Phasor diagram of SEIG with MSC under balanced conditions

A general analysis of the SEIG with MSC can be carried out using the method of be ~ o n s t a ~ t s ~ ~ e ~Gompone i c a ~s. All the equivalent circuit parmeters are as nce air gap except the magnetisi reactance, which is a fbnction of the posit voltage. With reference to Fig. 1.5, the following 'inspection equations' E561 may be w~~en:

where,

1 = G1+ J y ,=Z1

and

~ q ~ a t (1.6 ~ oj nimplies that z ~ r o - s e q u ~ voltages ce and Gu~entsare absent in the SE1 solving (1.5) to (1.8) in terms of the delta system of synmetrical ~ o r n p [57], o ~ the ~ ~ ~osi~ive-se~uence volta~eV, and nega~ive-sequ~nce voltage V, c m be d e t ~ ~ i n e a :

1

Energy ~ e n e r ~ tunder i o ~ the New Enviro~ent

Y,+-Y2

A

v,=&v.

(1.11)

Y2 -t- Y p + Y,,

v,=J?v.

(1.12) Y2 +

Yp+ Yn

where Ypand Ynare the positive-sequence and negative-sequence admittances of the SEIG. The input i ~ p c d ~ Zc, e of the SEIG when viewed across stator terminals 1 and 3 (Figure 1.S) is given by

Yz + Y p + Y n 3 Y pY , -iY p Y2 + Yri y2

(1.13)

Appiying ~ ~ c ~voltage o f law ~ tos loop 1345 in Figure 1.5,

For successful voltage build-up, I, f 0; hence L

+

z,, =0

(1.15)

Equation (1.15) can be solved for the excitation fkequency a and m a ~ e t i s i n greactance X,. d X, have been d e t e ~ i n e dthe , positive-sequence air gap voltage is found from tisation curve. The generator performance can then be comput~dusing (1.5) to (1.12).

The input impedance Z,n as given by (1.13) involves the generator admittances 5 and Y,

whose real and i m a g i ~ parts a ~ are high-order polynomials of a X,. As a result ofthe algebraic manipulations involved, both R,, and &,! in (1.13) are extremely complicated functio~sof the above ~ W Ovariables. Serious difficulties will be encount~redwhen solving (1.15) using conventional techniques such as the Newton-Raphson method [47] owing to the lengthy mathematical derivations required. To overcome these d ~ ~ c u la ~function ~ ~ s , minimisation t e c ~ i ~ is u eemployed in this case study for solving (1.15). This is based on c ebe ~ o ~ the o b s e ~ ~ t that, ~ o n€or given values of a and X,, the input i m p ~ ~Z,~ can readily. The following scalar impedance function is first defmed:

x,>=

z(a,

(1.16)

~

Power System Restructuring and Deregulation

32

and X , are respectively the equivalent series resistance and reactance of&. olution of (1.15) is next formulated as the following opti~isation

For given values of load resistances, excitation capxitancm and speed, determine the values of a and X, such that thefunction Z(a, XJ is minimum. It is obvious that Z(a, ) has a minimum of zero and the corresponding values of a and X, also satisfj (1.15). Any o p t ~ ~ ~ s a talgorithm ion that does not require the evaluation of ~ ~ c t i do ~nv a t i v e s may be used for the above problem. In this study, the pattern search method 1583 is used for ~ n c t i o n~ ~ i m i s a t i o The n . method employs two search strategies, namely exploratory rn moves, in order to a r b e at the optimum point. A ~ n c ~ i evaluation on is required each time an expioratory move or pattern move is to be made. For normal opera~ionof an SEE, a is slightly less than the per-unit speed b whilst X, is less than the u n s a ~ a t e dmagnetising reactance Xmu. A c ~ o r d i ~ ~bl yand , X,,,could in general be chosen as initial estimates for a and X, for starting the search procedure. In practice, it was found that a smaller initial value for the variable a (say 0.97b) would give more rapid converge~ce. To simplify the calculations and for easy comparison, all the machine p~ameter$are expressed in per-unit values using the rated phase voltage, rated phase current and rated power per phase o f the induction machine as bases. TabIe I. 1 shows typical computed results for the $xperi~en~al machine. The hnction minima obtained imply that very accurate so~utionsare possible. Over a wide range of load, the number of hnction evaluations Nrequired to reach a solution varies from 350 to 450. ~ a b 8.1 ~ eComputed results for SE16 with MSC

RL,

xtn

U

(P.U.1

1000 10 5 2 1

0.5

N

(P.U.)

0.977 193 0.975 109 0.973059 0.9672 18 0.958454 0.944063

-z,,,

1.2021 1.2205 1.2404 1.3084 1.4576 1.9230

z(Rm @.U-)

412 402 345 377 401 449

9.94e-6 7.73e-6 2.09e-6 3.5Oe-6 4.48e-7 1.88e-6

b = 1.0;ail= 0.97b: X,,, C, = 47 PF;Cl

= 2.48 p . ~ . 146 PF;R u = 2.3 P.U.

To illustrate the phase-bal~cingcapability of the MSC, ex~erimentswere c 2.2 kW, d e ~ ~ - c o n n ~ c induction ted machine whose equivalent circuit data i s given in the Appendix. The speed of the S E E was m a i n ~ i ~ eatdrated value (b = 1.0>and the values of RLi9C,, RI.2 and C, were carefully adjusted until perfect phase balance was obtained. ical results are given in Table 1.2, The good ~greementbetween ~ o ~ p ~ and ted

neration under the New Environment

33

results confirms the principle o f phase b a l ~ c i n gfor a three-phase ~ondi~ions for perfect phase balance in three-phase S E E with MSC

v

Zph

YP

@P

@.U,)

@.U,)

@.u.)

0.918

0.967

1.053

(deg) 130.8

0.835

1.037

1.214

134.7

0.805

0.954

1.186

135.5

0.796

0.789

0.992

133.8

RL, @.u.> 0.59 (0.56) 0.51 (0.49) 0.52

c, ($1

RL2

c 2

@.U*)

@.U*) 146

50

2.73

(0.50)

(49) 49 (46) 44 (4%)

(2.82) 1.78 (1.87) I .64 (1.83)

(146) I68 (167) 161 (160)

0.62 (0.61)

41 (39)

2.26 (2.

136

Normal: experimental values; bracketed: computed values

~iguresI .7- I ,9 show the steady-s~atep e r f o ~ a n c eof the SEIG with elements fixed at the following values: C, = 47 pF, C2 = 146 pF and RL2 seen that the SEIG i s balanced at a load current (experimental value) of 1 co~espondsto a phase voltage o f 0.86 p.u. and a phase cunent of 0.92 electrical power output is 1.63 p.u. (1940 W, or 88% of rated power), of which 80% is delivered to the main load RL, while 20% i s consumed by the auxiliary load RL2. Under the above conditions, the p e r ~ o ~ a n cofe the SEIG is the same as if it were excited with balanced capacitances and supplying a balanced load. For loads close to the balanced opera tin^ point, an experimen~alefficiency of 80% can be obtained. Very good c ~ ~ e l a t i o n between computed and experimen~~ results is obtained; hence the validity o f the a1 component analysis and solution technique is verified. 1.7 and 1.8 show that, when the values of the phase converter elements are ents and voltages in A-phase and B-phase may exceed the rated values when the load is reduced, particularly when the SEIG has been balanced at heavy loads. m e t h o ~to alleviate this unde§irable effect i s to balance the SEIG at part load (say 80% o f full-load current). The ~ e r f o ~ a n of c ethe SEIG will then be § a t ~ s f a c t oetw ~ we^ this load and full load. Another method is to balance the SEIG again at smaller loads, which involves ~ ~ l t i ~ vphase a l uc ~o n~v ~ ~elements er controlled by a s ~ p s w ~ i e~ h i n gs ~ a t e ~ .

Power System Restructuring and ~ e r e ~ ~ a ~ i

4

volt

4.1

igure 3.7 Phase voltages of three-phase SEIG with MSC

Phase ciurents ofthree-phase SEIC with MSC. P,: output power to main load RLI:P,: output power to auxiliary load R,, 2 1.6

1 0.6

re 1.9 Output power and efficiency of three-phase SEIG with MSC

Energy ~ e n ~ ~under a ~ the o nNew ~ n v i ~ o ~ e n t

1.11.6

5

~ i m ~ i P~a~e-balancing i~~d ~ ~ ~ ~ r n e

In circumstances where it is not pract~cableto provide auxiliary loads, or when a u x i ~ i a ~ loads need not be supplied, the simplified Steinmetz connection (SSC) shown in ~ ~ g be employed. In this case, all the electrical power output o f the $E to the sing~e-phaseload &,. The phasor diagam for the MSC (Figure 1.6) and ondiiig equations (1.1)-(1.4)may be used to identify the conditions for perfect phase balance for the SSC. Since the auxiliary load resistance R,, is absent, in (I .3) is forced to a s s u ~ ea zero value. Accordingly the posit~ve-seque angle bpo f the S E E must be equal to 2n/3 rad for (1.3) to be satisficd. F and (1.4), the values of the load conductance and phase-conve~ers in a b a ~ ~ c oepde r a ~ i ~ofnthe SEIG are: G,= 3 YJ2, B , = 43 YJ2 and

T ~ j ~ p iSteinmetz i ~ e ~connection for three-phase SEIG

The a u x i ~ i ae~x c i t a ~ ~capacitance o~ C, is thus one-half of the main exci~a~ion y s e l ~ c t i ~proper g values of C, and C,, perfect phase balance can be of stator current. with the SGG is simiIar to that for the SE1 ow equal to (0 +&). Figures 1.11-1.13 s e x p ~ r i ~ e n pt ae~~ o ~ a n of c e the SEIG with the SSC at rated s excitation capacitances fixed at 110 pF and 55 pF re at a load c ~ e n( te x p e r i ~ e nvalue) ~l of 1.13 P.u., whic voltage of 0.985 p.u. and a phase current of 0.77 p.u. Under this p.u. (1320 W) is delivered to the load and the efficiency o f the SEIG is 79.6%. Again very good agree~entb e ~ w e ~the n c o ~ p u t e dand experimental results is observed.

1. 1”

0.

El. 0.4

I

-/

4 Phase vohges of t ~ ~ ~ p hSEIC a s ewith S

hase currents of ~ h ~ ~ - p hSEIG a s e with S 1

1.

0

0

Fi

3~

upowe~

~

t

7

Energy ~ e n e r a under ~ ~ ~ the n New ~ ~ v i r o ~ e n t

1 ~ ~ a c h has ~ nthe e follow~ngpa~icula~s: .4 A, four-~ole,50

~

three-phase, d e l t a - ~ o n ~ e ~ t e

e ~ i a c ~p ia ~r ai ~ ~(in ~ eper-unit ~ s values) are:

_I

-

-

-

0.0844 0.112 0.0~2 1 0.098 1 0.1 22 0.013

(1.17)

de A l m e r ~(PSA) ~ in Spain, xi@ and success. The key c its sea§ona~cycles. This results in s i the collector f X d and the plant a

set of system p~ametersoptimised for a prescribed range of operations have proved to cently, fuzzy logic control (FLC) schemes, which enco like approach of processing and handling of information, have been of n o n - ~ i n pe ~l a ~ with ~ s pKomising results. However, early studies show that in such FLC schemes the optimis~t~on of the ‘if-then’ rule base is often a c ~ b e r s o and ~ e la~orious rocess ~ ~ v o l v i‘trial n ~ and error’. Genetic algorithms based on the i i a t u r ~law ~ of volutio~lend themse~vesas an ideal op~i~isation tool to be used in c o n j ~ n c ~with i o ~ FLC syst~ms.This s ~ is one ~ of y the first of its kind to show the d e v e l o ~ ~ eonft a scheme aimed at o p t i ~ ~ s i the n g response time of a solar power plant to input power and t e m ~ e r a demand. ~r~

The solar power plant under investigation, Plataforma Solar do Almeria (PSA), in Almeria, Spain

Figure 1.15 shows the block diagram o a d i s ~ ~ ~collector ~ t e d field called the solar collectors a r r ~ gd in 20 rows foming 10 p long. The oil is pump d through the receiver tub the receiver tube walls. The storage tank is filled with oil in the far end. The oil is heated and then in~roducedinto the storage tank to be used for e l e c ~ energ c~ plant [64]. The system feeding the heat exchanger of a desa~~nation three-way valve located in the field outlet that all its outlet t e ~ p e r a ~ Kiseadequate for entering into the m in a dis~ibutedcollector field is sired level in spite of disturba level, cloud ~ ~ v e ~ emirror n t , reflectivi~or inlet oil te

Energy ~ ~ n ~ ~under a t the i oNew ~ ~nviron~~nt I

I

1

4

Steam Generato

Steam turbine ///*

ACUREX Collectors

I--

I I I

Distributed Collector Field

!

(20mws, 1 D loops)

j

Pump

Storage System

A

-

1

Power Conversion

1

System

-

Cooling

tower

Block diagram of the solar power plant

1.12.3

~ o n t r~~t lr ~ c t uorfethe Plant

ed to the plant with shows the overall control bloc diagram of the FLC ted that the o ~ t ~ e ~ the proposed GA optimisa~ionscheme. For this solar plant, it t e ~ ~ e r aof~ the r e field depends on other variables such as solar r a ~ a t i o nI and the inlet ~ has an influence but ch tempera~reto the field 7;.,. Mirror r e f l e c t i v ~also that it may be co~s~dered constant. Hence dynamically the out1 e ~ ~ ~ e sass ea dn o n - l i ~ ~e ~n c ~ if oofn oil flow U,solar radiation al (of the ch The linearised model is based on p a ~ ~derivatives ATo with respect. to changes Au, dl and AT,,):

(1.18)

The ~ a ~ de~vatives i a ~ can be co~isideredas transfer fitnctions re~a~ing the var~a~ion in outlet t e m ~ e r a ~AT, ~ e to variations in oil flaw Au, solar r a d i a ~ i o ~ ATin, res~ec~vely. The mathemat~~a~ model which accoLin~s B ~ ~ f l u ies ncomplex. ~ ~ ~ To approximate these effects, in series with the FLC s shown, has been develo

Uf

=

0.78691 - 0.485(~- 15 1.5) - 80.7 U-Trl

where !,is

the oil flow

U i s the t e ~ p e r a ~set r epoint,

(1.19)

radiation

Tin

1.16 Control structure of plant

‘ i f - ~ ~ e mles n ’ in

where aj, pi

xi ,6 i ,E, E

ase ofthe FLC as

[Q,I],

~0000 I } , Rule 1 is not s i g n i ~ ~ a nwher~as t; as . It is found that using a higher number of r o v e ~ e n ~ins ~ e r f o ~ a but n ~ ei ~ c r ~ a s ~ i ~ n ~ ~ The ~ a entire ~ t i ~y h. r o m o s o ~ Xeis o f the f o ~ a ~

e there~orehas a total of 245 bits of i n f o ~ ~ a t i o n .

icai ~ ~ o r n ~ ~

Energy ~ ~ n e r a t under ~ o n the New Environment

a

~01010101000Q0101 1101Q1Q~0101010~01 1110101010~01Q111~0~01012 1111110110000 11100011100101111Q10101010l000000000000101~11101100111110 11 1Z 10001 1 111 I 1000110000~00011111Q0 I1 ~ 0 0 0 1 0 0101 1 ~ 110001I1 0001I 100 1000111101110111111001010~

The chromosome

Rule base consisting of 49 rules

7 ~ ~ r o ~linked o ~to rule ~ ~base e s

re~ro~uction, crossover 1.18. Firstly, the GA r nto an ini~~alisatio~

Power System Restructuring and ~ e r ~ g u ~ ~ ~

2

Step 1:

Initiaiising pool by randomly. (size=30)

geiierating chromosomes

Step 2:

run = run i1 ;/*mn = 0 initially */ Ben = gen + 1; /*gen = 0 initially" / /*Reproduction*/ Calculating fitness of individual chromosome fram initialisation or mutation pool. Copying high fitness chromosomes to reproduction pool. (Roulette Wheel method) /*Crossover*/ Selecting chromosomes randomly from mating pool and crossover. /*Mutation */ Copying new chromosome into mutation pool, mutation probability is 0.001.

Step 4:

Step 6:

If gen = no-gen Goto step 3 If run = no-run Goto step 2 End

~ s e u d o ~ c €or o ~ the e GA

~ x p e r ~ ~results e n ~ on l the simulator of the plant have been taken to verify the proposed GA-FLC s c ~ e ~Ine .Figure 1.19, the effect of GA o p ~ ~ ~ i sofa ~thei orule ~ base on the e r f o ~ a n c of e the plant is illustrated. The upper graph shows the eiTor versus the number of generations. The error, an index of the fitness of the c h r o ~ ~ ~ ois~seen o ~ to e , decrease as . The middle graph show bottom one shows the corre emcnt of the dynamic response o scheme on a day wh tional PI control s c h e ~ e .It is the plant's robustness when external . Since there is only on u n ~ ~ u ~d~fferen~ ly in any time interval, we concurr~n~ly in real-time. The validity of the comparison ~ e ~ e the e nPI and GA-FLC ~ 0 ~ s ~c h ~e ~0e slies 1 in the fact that the simulator is a proven ~ o of the ~ plant e [63] ~ and ~s one can ~ ~ pthe~ solar r eradiation in a particu~arperiod and use it as one of the i n p ~ to the simulator. The simulator's output is then compared under different control schemes. The current ~nvestigationis based on this principle.

Energy ~ e n e ~ a t i ounder n the New E n v i r o ~ c n t

35000 30000 25000 j

-

20000 -

isaoo 10000

5000

300 -

0

300 1 25Q

1

00 150

I----

100

~ ~ f c con t sthe ~

e

et point

~of thef planto by the ~ GA ~optimisa~~on ~ ~

arison of GA-FLC with PI scheme un er extreme external dynamic c

~

a

~

neration under the New ~ n v i r o n m ~ n t _ I _ _ _ p

such as l a r ~ e ~ s expen ca~~ e ~ ~ s t cannot o ~ be~ sec ~ s

Energy Generat~onunder the New E n v i r o ~ e n t

4

_ l _ l *

shnell and S.S. Qren, ‘ idder cost revelation in electric power auctions’, J o ~ r n uof~ ory Economics, Vol.6, 1994, pp.5-26. o and R. Wilson, ‘Priority service: Pricing, investment, a d market organization’, Anzerieun Economic Review, Vo1.77, 1987, pp.899-916. ‘Priority pricing of ~ n t e ~ p t electric i b ~ ~ service with early [S] T. Straws, and S.S. notifica~ion’,Energy ,V01.14, 1993, pp.175-196. [9] A. Midtun and S. Thornas, ‘Theoretical ambiguity and the weight o f historical heritage: a c o ~ n ~ a r a ~study i v e of the ritish and ~ o ~ c g i electricity an l i h ~ r a l i ~ a t i EmerB o ~ ~ , Policy, V01.26, 1998, pp.179-197. aas, N.Auer, C. Huber, and M. Tranger, ‘Limits for competition in restructured electricity ~ ~ k e- tthe s European pe~ceptive’,19th annuui North Amer~cunCon~erence,United States Association for Energy Economics and International Association for Energy Economics, 1998, pp.103-112. [I I] IEA, World energy outlook: I998 edition, Paris, France, QECDIIEA, 475pp, 1998. [I21 C, Singh and N. Gubbala, ‘Reliability evaluation of interconnected power systems including jointly owned ~enerators’,IEEE Transactions on Power Systems, Vo1.9, No.), 199 412. an, ‘Evaluation of the reliability and production cost of i~i~eKconnc~~ed ems with jointly owned units’, IEE Proceedings, Vol. 134, No.6, 1997, pp.377-382. nt contracts: The case of coal’, ~ ~ ~ r qfLaw n u l and Jaskow, ‘Price a ~ j ~ s ~ine long-term ~conomics, Vo1.3 1, 1998, pp.47-84. [ 1151 IEA, rroje~tedcosts o ~ g e ~ e r u t ielectricity, i~g Update, Paris, OECD/ IEA, 243pp, 1998. Vol.6, 1998, pp.25-29. [ 161 J. Lane ‘Sweeping thc board’, Power EngineeringInt$~u~ional~ on Embedded Generation, IEE, 28 February 2000. utanto, ‘Battery storage plant within large load c e n ~ ~ s~9 , E E ~ terns, Vol.?, No.2, May 1992, pp.762-767. [ 191 Gmzettu Uficiule delta Republica Italians, Decrelo legislatho 16.3.1999 n. 79, Attuazione comuni per il mercato in afico e Zecca dello Stato [20] EC, ‘Guide to the electricity directive’, available from: http:J/www,~uropa.eu.in~en/ c o ~ ~ d g ~ ~ / e l e c . m e i n oBrussels, r . h ~ , Belgium, European Com~ission,D ~ r e c t o r a ~ ~ ~ ~ e n e r II ( ~ n ~ r g yIOpp, ) , March 1999. E211 P. Baruya and D. Goidsacks, ‘European coal issues - European tibcralisation af coal’, ~ o ~ l d Coal, Vol. 7 (10); 1998, pp.29-34. road ben^, ~ G o ~ p e t i ~ i v eof n ecoal s ~ - the evolution of price’, CS/05, London, U Goal Research, 20pp, [23] UNEGE, ‘Security supply in a changing European natura1 gas ~ ~ P . 3 / ~ ~ . 4 / 1 9 9Geneva, 6 / 6 , Switzerland, United Nations Economic C o ~ m ~ s s i o n for Europe, Committee an Sustainable Energy, 17pp, June 1999. E241 Couch G., ‘OECD coal-fired power generation - trends in the 199Os’, IE esearch, 83pp, April 1997. [ E ] Modern POWCK Systems, ‘World digest: Green power launched,’ Modern Power Sjwterns, VoI.19, ~ e 1999. ~ ~ a ~ lobal Private Power, ‘Own coal?’, GIobuE Private Climate, report, The ociety and The Royal ~ c a d e of ~y

hugar, 'The value of grid-sup T r ~ n s ~ coiz~ Energy i o ~ ~ o n ~ e r s iVol. o ~ ,10, 1995, eloping CHP in the public sector and beyond' up Ltd., April 2000, pp.2-22. Ilan, Pcter Gaossley, Daniel icks, Fuel Cell Systems ~ x ~ ~ Q i n hluture far ~ i s t ~ ~g ~ u nt eer a~~ ~ oEniq ,e c ~ ~ i ~ ~ ~

~ a n a g ~ ~ ~ n t IiwM.w.nemmco.com.a~aulne

ket

9

Energy Generation under the New E n v ~ r o ~ e n ~

. Wa~son,J. Arrillaga and T. Densem, ‘Controllable d.c. power supply from wind driven se~f-excitedinduction m~chines*, 1EE Procee~ings,Vo1.126, W0.12, 1979, p p . 1 ~ ~ ~ - $ 2 4 ~ . [SO] T.F. Chan and L.L. Lai, ‘Phase b a l ~ c i t ~forg n se~f”exc~~ed ~ n d u c g~ e~n oe ~~~~t o~r r’ ~o ~ e e ~ i n g s of the In~erna~ional Conference on Power Utility De lation, ~estructuringand Power ~ e c ~ n o l o g i2000 e s ( ~ ~ T 2 0 0 0City ) , ~niversity,London, TEEE [SI] T.F. Chm and L.L. Lai, ‘S~eady~state analysis o f a thr~e-phase co~iection’,IEEE Power Engineering Review, V01.20, NO.10, [SZ] T.F. Chan and L.L. Lai, ‘A novel s i n ~ ~ e - ~ she~l € s e- r e ~ ~ lself-excited a~~d i using a ~hree-~hase machine’, IEEE ~ r a n ~ ~ u c ~on i QEnergy ns ~ o n v e r s ~ oVol. n , 16, in ~ e v e ~ o ~c ~o n~ g~ ~ eZEE s’, ropriate technology - rural electri~ca~ion evzew, Vo1.35, No.7, August 1989, pp.25 1-254. E541 T.F. Chan, ‘ P e r f o ~ ~ a n cAnalysis e o f a three-phase induction generator se~f-e~cited with a ~~E~ Power Engineer~n~ Society 1998 ~ ~ n~ t e e e ~ t ~i a~p ~~ ~r single capacitan~e~, 0 ~ 8 ~ ~ C - 0 - 1 0~~ e~ 9b ~r 71-5, u, ~1998, Tampa, Florida, U.S.A. nt ~ ~ c h i n eLondon: s, Pitman (EL S), 5th Ed,, 1983, p ~ . 3 ~ ~ ~ ~ 3 method o f analysis o f 3 ~ ~ h a ~nduc~ion se ~ o with~ ~ Proceedin~s,VoI.~OOA,PI. E, 1953, o f a 3-phase induction motor connected to a single~ r o c e e ~ 1959, ~ n ~Vol.l06A, s~ 1959, p ~ . 1 8 3 ~ 1 9 ~ ” 1W e i ~ I ~ n t rao ~d u~~ ~to i ~~n ~ ~ i m i ~~ uh t ei oo ~ , L.L. Lai, T.E. Tong, ‘ A opt~~isation o f rule base in a fuzz plant’, ~roceedings ox the I n ~ ~ r n a ~ i o n~a l ~ n ~ ~ on r ~ Pno w c ee~ Utilip ~ ~ ~ e~ e~ s ~~ a~ and ~~ Power ~ ~~ roTechnologies ~ ~ n ,g 2000, City University, London, IRE]E, p d 2000, p p . 2 2 ~ - 2 2 ~

A. ~

~‘Experience ~ of ~using thea neural i

-

and genetic algorithm €or fault secdon estimation’, P E PrQceeding~ ~ e n ~ ~ f f ~ ~ r a ~ ~and~ ~ ~ ~ . $~ ~s rvol. ~ ~o145, ~~ No.5, u ~ ~~ eo ~ ,~1998, e ~p ~~ e . s ~ ~ ~ - ~ 2 ~

Miss Vee Shan Uuen Prof. Mwok Lun Lo i ~S The University of ~ ~ a t ~ c ~ ~ d e The U n i v e ~ §of ~ l a s g o w~ ~ c o ~ l ~ ~ ~ a s ~ ~ocwo ,t ~ a ~ d

The success tisation o f the airline, teleco ~ e r e g u l a ~ ~ ~tructming of the electricity ers in ~ r i v a ~ ~its~ vertically ng i n ~ ~ g r ~el~ e d ~o~lowed in 1990 and 1996 ~es~ectively. The d Norway has encouraged other countries worldwide at have been ergoin in^ energy ~ e r e ~ i l a ~~ i c~ n~ ~ d e ain, Taiwan and ~ a l a y s i a . s used to refer to what one wouM regard as ‘ d e r e ~ u l aof~ ~~~ ~ l ~~ utiliti c ~ c le the two words are d ~ f f e r e~iteral~y, ~~ Ironically, neither is th ~ e ~ t 2.5, i o none ~ of titi ion or o ono^^^^. It i s ther~fo sive exercise of rn

~ e ~ ~ ~afaElectric t ~ oUtiliti~s n

1.

v an

structuring and ~ e r e ~ l ~ t i o n

the details of der

latiara of Electric ~ $ ~ l i t i e ~

2.4. I

exercise market power and control the price of e l e c t r ~ ~ i ~ . itions where the providers of a service can c ose that would be established by a coinp~ti~ive mark actual prices and the prices that would arise from the assumption that the generators are priceta en a major impediment to price reduction in the Engl Pool. Efforts are being made to eliminate market ~ a ~ ~ c u lhere a r , will be a d r a ~ a ~reform ic of the energy market the year 2000.

Access to the ~ansmissionsystem is one of the main issues in energy open acces and sound r~gulationsare required to facilitate ~ransmiss~on ng, TOA refers to the regulating construct such as the rights, obli dures, economic cond~~ions9 etc., enablin~two or more parties to use a transniission etwork. With equal rights of access to the transmission network, it has become ~ ~ a s i b l ~ or loads to arrange transactions with each other and hence c o ~ p e ~ i t i o n is am on^ the key elemen~sfor facilitating competi~ionin the ~ ~ e r g y on will look at the details of the two issues.

2.5.1

it^^^ in the ~~~e~~

~ o m p e ~ is~ the ~ ~main o n goal of energy p r ~ ~ a ~ ~ s a Ideally, t i o n . from view, perfect ~ o ~ p e tis~the ~ ~iiiost o i desirable ~ market structure. cmre ~hara~terised most notably by a situation in whic are p r i ~ e ~ a ~and e r sthere is freedom of entry into and exit from ih ng to these three criteria: inde~endence,product s ~ i ~ s t ~ t u ~ a b i l i wever, in any real markets, it is rare that all of the Considering also the te ints caused by the intrinsic properties can be d ~ d u ~ ethat d etition does not exist in the e n e r ~ ~ by its social welfare. Social we~farei d the benefit of the energy to socie for it. ~ a x i m ~social m welfare i s ~equentlyoperates at a s ~ i b o p t ~level. ~al been introduced in most deregulated m their own s u ~ p ~Retail ~ ~ r ~. o m ~ e ~ i t i o n c u s ~ o m ~ rare s abie to select their ated by the issue of direct access tec~nQlogy.In some countries9solid regul costly for res~dent~a~ customers t

e issue of e n e r g ~subsid~es the depos~t~on of s~randedcosts have also c o m ~ ~ ~ c efforts a ~ e d on energy ~ ~ v a ~ i sne aform ~ ~of~energy ~ subsidies refers to those given to generators to purchase highly priced coal in order to sustain the Iocal coal industry. Generators receive s ~ ~ n i ~ c a nfewer ~ i y subsidies after

e r ~ ~ l a ~OS i oElectric n Uti~i~~e~

that time were su

e less WQ~hy and i ~ v e s t o could ~ s end up b e ~ ba n~ n involves the d e t e ~ ~ ~ a t of i othe n degree of recove e ~ a l i f o ~ Pool, ~ a n the 8 in the e ~ e c ~ ibill. ci~

a price and the ~ a x i IIU~ u ~ tted offers are r ~ e lowest d

, each seller i ~ ~ ~ i asub lly are w ~ ~ ltoi ~make i ~ avail

om a seIler 10 a buy ed ~ a r k ~the ~ sQ,

~ n s u f ~ c ~because e ~ t ne

r ion ~acilities~ henever each whee

~ ~ r e ~ l a of t i Electric o ~ i Utilities

o f the line is now res~aine by its t h e ~ alimit l (Fi per€ect~yinelastic ( e ~ a s t i c ii~s zero) mand which is no generator 61 is met when 6 1 is forced Buy some of its el ~heoret~cai~y raise its price as hi as ~ossib~e. ~enerator6 2 is said to have a c ~ ~ i r e d ~ i i I ~ ~m~ t w e d~ o~w e r~. ~

Transfer L

i = 1~

~ 0 / ~line ~

~

~

olution

~ l ~ ~ s ~ r aoft im o n~ kpower ~ t caused by congestion

~ i g ~2.3e ~ l l u s ~ a t e s iflerent ~ongestion pricing ~ ~ t h o d o l o c l a $ s ~ ~ ~ a t iPriva~~sation on. a s is ~ ~ n e (e.g. n t diffe coordination betwe Europe and different states in e$sential to alleviate congest co~gestiofl~ ~ c i n ~ . calculated as dual ~ u l t i p ~ ~from e r s optimal po ons. These we based ~ ~ c iIn~places g . where nodal pricing is adopte~,differences in nodal prices c m ~ e s uin ~t arket ~ a ~ ~ c i ~can a n hedge t§ cong congestion contracts which are also and give their holders ~ a ~ s m i s s ~system, on

[9, Z 51.

Q e r ~ g ~ l ~oft Electric i o ~ Utilities

a h a ~ f ~ h o u basis. r ~ y Many customers will pay for electric power based on this price^ e~ther irectly ~l~rough their distribu~~on utility or t ~ o ~ ag private h power supp~y e Pool price. The IS0 can also operate markers for a n c ~ l t a ~ wer, spinning/non-spinning reserve and losses. The roles of the

2.6. I

~

~ and i

~

~

~

g

n Section 2.5.2 ~ifferentkinds of auction ~ e c h a n i ~ r were n s discussed. T uty to set the e l e c ~ c i ~price. y Pricing is done essentially in eith ex ante or ex post. An ex ante market is one in which the price o f the CO is set prior to its del~verywhile an ex posl marke~is one in which the n the e d time o f delivery. In an electricity market, a c o m I ~ i o d iis~d e ~ e ~ ~ at is like a b~~ateral con~rac~ market in which ~aders/pa~ic~pants agree on the a ~ ~ o ~ ~ electricity to be delivered at a certain time in the future at a certain price Nord Pool combines ex ante and ex post pricing. In its spot market, syst prices are set up the day prior to delivery. Any differcnce in the forecast wi delivery results in a discrepancy with the pre-set price and the spot price. This is c o ~ p e n s a ~ cby d the presence of the ex post mechaiiism. In the ord Pool, there is a buy~ackmarket to make up for this difference. ~imilarly,genera~orbids are also s u ~ i ~ i ~ e d on a o ~ ~ ~ ~ a y basis - a lin~the ~ ~ England d and Wales Pool, and participants are paid at the end of each day for their transactions plus CoInpensation. The England and ~ a ~ Pool e s is the~eforealso an ex ante market with an ex post mechan~sm.Ex post markets also exist and examples are the New Zealand and Australia markets. In the New Zealand e l e c ~ r ~ ~ i ~ mar~et,ge~eratorsand loads are aflowed to change their bids until 2 hours prior to d e l i v e ~ and the market is cleared re larly d u r i ~ gthe bidding process. Ex post prices are c a ~ c ~ ~ a t e using arke et-c~ea~ng software with the latest offershids and the actual r n e ~ demand ~ r ~ ~ of ex a m and t?x post p r ~ c i nfor ~ together with losses. Figure 2.5 illustrates the ~n~eraction e i e c ~ r i cmarkets. i~

~ ~ ~ all particip ts Rave to bear a certain degr As in any other c o m r n Q market, rket. The $ys~emo~erator o has a share o f the risk rep an^^ of the forecast wi chal demand. The degree IS0 depends on the pricing ~ ~ c h a n i sof m the market. For ex sate s of market p a ~ ~ c ~ ~ina ~ n ti s~ a tceo ~n ~~alc ~depe s stem price in the future. In a ~ ~ m ab~ k elike ~~ ~ d ior to delivery, and any real-time power imbalan xhibiEs the ex post m ~ c ~ a n i ~ i ~ is therefore much more susc y financial comrni e a serious issue w h ~ nc~ngestionis comm

Power System ~

~

s

~ and c ~ e~ r en~ ~ gl a t i o ~

-

',

LOSS

'..

\

Compensation

actual demand

I

-

1

~_____-

final clearing price I e.g. Nord Pool, E & W Pool

1

~ ~ r e ~of Electric l ~ ~ Ui ~ iol i t~i ~ s

3

tier prov~ders~ play an nine years after the e s t a b ~ ~ s ~of~ Be n ~ c o n s ~ e r rather s than be c B ~ t u 11c o ~ s u ~ eshould rs have access

is a type of b i l ~ t e r ~ l

I

eregulation of Electric Utilities

5

to meet forecast demand. It is a ma

By 10.00 am. every

=3

+ 0 . 0 2 and ~ ~ ~

.7(a) The network; (b) unconstrained dispatch; (c) constrained d i s p a t ~(Source: [151)

h is shown in Figure 2.7b Wh (~igure2.8a). A § § that ~ ~ e ex ante price the day tice during real-time opera

costs of u ~ ~ and constrained ~ ~ dispatch s ~ ~ e nare t ~listed in Table 2.1.

a

~

~

~

~

Deregulation of Electric UtiIities

67

(a) SMP; (b) G1 cost function and adjustment; (c) G2 cost function and adjustment (Source: 1151)

Me 2.1 Gmerator payments and demand charges (Source: [IS]) Demand Payments Demand Charges (PSP) (Eh)

L1

Total Charge (E/h)

1464

~eneratorPayments

G1

62

G e n ~ r a t i nCosts ~ (Light-shaded Areas) (Eh)

616

487

Generating Payments from Pool (Ou~ut*PFF)(E/h)

952

340

Adjustments (Dark-shaded Areas) (E/h)

25

147

Total Payment (Sum of Generating Payments and Adjustments) (f/h)

1464

308

L2 1156

has been a significant drop in electricity prices in ~nglandand Wales this price drop does not hlly emulate the cost reduction of g are not passed on to customers entirely but are partially r es in the form of higher profit. Also, there has not yet been a decrease of price in the retail market. A possible reason for the i n e f ~ c ~ wholesale market is that the three largest generators could game and m ~ ~ p u l a the te w h o l ~ s amarket. l~ The market lacks small IPPs which could potentially f ~ v o ~ u o~~ p e t i t i o n and reduce the market power of the large generators. In view of the existing problems of the pool, the director of the Office of Gas and ~ l e ~ ~Markets i c i (~ ~ f published ~ ~ mthe~NETA [$I, for England and Wales in 1898. The reforms should commence in 2001 and should lead to significant chang~sin the exis market. First of all, a d ~ ~ ~ ~ m iauction n a t owill ~ replace the uniform auction. Sec d e ~ ~ a n d - sbid i ~ edin^ is allowed so the market will transform into a bila~eralrn reform§ are designe~in such a way that pa~icipan~s can choose over a i f f e ~ ways e ~ ~ they ate in the market. In a different time frame before actual delivery ~ a ~ c i p a can n~s choose to trade in the following markets:

Forwards markets: these are optional and are operated by i n d e ~ ~ d e on tr ~ ~ i s a ~ ~ 5 ~ sign~ bilateral contracts that are up to sev~ralyears ahead as desired. P a ~ i ccan i ~ ~ S h o ~ - t bilateral e~ market: this is optional and open from 24 to 4 h ~ u r sahead of the period, All trades will be organised by a market operator (MO). s is also optional and it is open 4 hours ahead to he SO obtains full control of the system after the close market. 4t would engage in trades to ensure that generation and d ~ ~ a nared balan~cd, into a c c o ~and t resolving any constraints on the ~ansmiss~on network. al-time power ~ ~ charges~are imposed a on ~p a ~ c~ i whose p ~ ~~ c o ~ ~ ea c ~ e d ~ o u nist different from the actual metered amount and they could be based on the costs 0 to settle the imbalances. The r ~ f o feature ~ $ full dem bids and simple offers and bids, and they aim to higher flexibility over different ways of tradin , Neve~he~ess, many eptical about the proposed r e f ~ and ~ sbelieve this is not the solu~ion to get rid of market power and reduce prices [20,21], Moreover, there i s concern^ under the sed reforms, over the possibilities of exploitation of generators’ market power in the ing r n a r ~ ethrough ~ the incremen~and d ~ c r e ~ ebids. n t Inerement bids r e p r ~ s e the ~t p a ~ i c i p ~wish t s to be paid for an increase in output or are willin increase in demand. ecrement bids represent the prices they are willing to pay for a d e c ~ e in ~ output s ~ or wish to be paid for a decrease in demand.

d eration e o m ~ e n c in e ~Norway two years a&er the pass d from the former ‘ C optional pool was sufficient because o f the la~gen Norwegian power system ha ission n e ~ o r k which , w tatnett, which is also ket. The ~ o ~ e spot g market, i ~ the Nord Pool or Elspot, is icipants are free to trade in the bilateral contract power is in ~ a l ~for c eg e n e ~ ~ o rlarge s , custo~ers land and Wales Pool, the Nord 001 utilises ex ante pricing to set the or to delivery and compens s power i m b a l ~ c e susing ex post generator offers and ahead of actual delivery, the Nord Pool acc our of the following day. The system pric emand curve meets the ag price auction by paying all generators the last en bidding areas d ~ r i n gthis process, reas. In the s u ~ area, ~ ~thes area price is by an amount equal to the line capac y the right shifting o f its supply curve

Deregulation of Electric Utilities

9

price in the su lus area is set up in such a way that it should demand which has a quantity equal to the capacity of the c o n s ~ ~ ~ ~ e hand, in the deficit area, the area pkce is set encouraged to supply an ~dditionalamount equal to the capa p a ~ i c i ~ a incur n ~ s an ad itional cost and this charge is called the ' is the dif~erenceb e ~ e e nthe system price and the area price. (A iilus~atethis m e c h a n ~below) s~ on is broadcast to pool participants by 2.00 p.m. on th power imbal~ncesare compe~sa~ed in a separate rators can submit buyback bids after the d a ~ - ~ e amd is ~ ~ i s ~These ~ e dbids . reveal how much a generator is willing to pay to buy surplus power and how much a g e n e r ~ ~costs o r to produce the deficit a ~ o u n t . system operator selects the cheapest a v a i ~ a ~generators le to buy or sell in case Q and c ~ ~ g e s ~ ~i o~na g e m e nand t , all in-merit generators are paid the price set by the h i ~ ~ ecost s t block. S ~ ~ ~ e misedone n t ~ s ~ ~ ainl W l yOweeks.

~ ~ Q I t ~e ~ Q ~~the s ~, area c

situation. This is reflected in the area prices. Also, because of the physical flow of 10

Power System ~ e s ~ c and ~ i ~i ei r ~~ ~ ~ a t i o

7 Area 1

Sumlus Area

I

DCftClt

Area

140MW

(a1

16MW

2

LI

2

LI

1b0MW

40MW

@>

116.8MW

(a) Unconstrained Dispatch; (b) Constrained Dispatch (Source: [ 151) EMWh

EMWh

f,l/MWh

61, Ll: surplus area

G2, L2: deficit area

5.8

4.6

MW

80

(b>

(c)

.I0 (a) System Price; (b) Surplus Area Price; (c) Deficit Area Price (Source: 1151)

Various Prices and Settlement Capacity Fee in Surplus Area, C, Capacity Fce in Deficit Area, C,

= P, -Pi = 0.72 f M h

Settlement Price

P, = 6.52 E/IMwh

Charge Credited to Ll and G2 (Mc) ebited to C1 and L2 (Md)

= PLl*C,

Net Income o f Grid Company

=Md-M, = Capacity * (Ph - PI) = 337 E h

I

= PI, - P, = 2.65 f/&W?h

+ PG,*Cd = 73.32 51% = PGI*Cs+ PLZ*C,= 410.32 E h

The ~ o ~ e g i energy an markets have been a successful example of energy ~ e r e g u ~ a ~ i o n . ket power has not been an issue, ~ e v e ~ h ~ lthe e s management s of power im~a~ances arouse^ concerns since it costs the SO money to resolve bo~Ienecksin the regulating market. ~ o ~ n a it~has $ ~only y contributed to a sniall amount of S tmett operating ~ u d g e t SO f a [I cong~stion~ ~ a g e ~will e nbet costly when con~est~on becomes more serious. er, the selection of ~egulatingbids using merit order, which is easily c o m ~ ~ ~ ~ ~ by n s participants, ible does not necessarily result in the lowest cost to alleviate co~gestion.

Deregulation of Electric Utilities

2.8.3

71

Galijhiu

The ~ n e r g yPolicy Act PACT) of 1992 clarified the d e t e ~ n a t i o nof the USA for a com~etitiveenergy market. It is not m a n d a t o ~to implement a whole§ale c energy market in the nation. Individual states pursuit difl'erent policies an ending on their electricity prices. States with relatively high California, New York, Massachusetts, etc., arc more aggressive in implemcnting reforms. In 1998, California embarked on a four-year transitional period of deregula~ion. ~ t r a n d ecosts ~ have com~licatedderegulation in California. The state gov solved this problem by issuing bonds to inflicted companies to compensate for thei Customers' bills include a small amount of charge (e.g. 4 cents~Wh),the so-calk competi~ion~ a n s f e charge r (CTC), to account for stranded costs. During this transitional period, participation in the pool is optional, ap large private utilities, which have to trade through the PX until March 2002. One ~i§tinct d i ~ ~ r e n cb ee ~ e e nthe Californian Pool and the England and Wales Pool i s that in the former case market clearing and bids matching are under a separate entity, the PX, rather than embedded in the duties of the ISQ, as in the England and W C a ~ ~ ~ o r ntwo i a , types of bilateral contracts exist: Contract for Access Contracts. The fact that CFDs are tied to pool prices has 1 game the market using their market power, The idea of Direct Access Contrac~sis to c o ~ t e ~ athis c t problem: Direct Access Contracts are not bonded to the PX and pa~icipan~s only have to request their transactions through the ISd).

ge (CalPX) is responsible for holding auctions for the competitive forward markets (day-ahead and day of markets). Th,e day-ahead market is similar to its c o u n t c ~ ain~ Norway and England. Market pa~icipan~s provide hourly supply/demand bids to CalPX one day prior to physical delivery. MCP is actually the equivalent of system rice in Norway or system marginal price in England and Wales. ~ n ~ ~ pricing o r m is adopted and all pa~icipant§are paid or debited the market provides pa~icipantswith the chance to make up for system imbalances by ~ o l d ~ g auctions at various times during the delivery day. Zonal pr~cingis cm~loyedfor congestion ~ ~ a g e m e nMarkcl t. p ~ c ~ ~ acan n tsubmit s the so-called schedule a d j u s ~ e n ids (SABs) which are similar in nature to the r e ~ l a ~ ~ g bids in the Nord Pool, The S represents the desire of the ~ a ~ i c i ~to a nadjust t its price varies. When there i s congestion, the region is ~ ~ v i d einto d zones ates the zonal prices using SABs. The PX uses this i n f o ~ a t i o nto work out the final prices for participants so that upon settlement the PX remains revenue n e u ~ a l

P31. tion stage in California and it is premature to c o on the ~ various markets. However, there is concern over the operation of the spinning and non-spinning reserve markets. ~ e n e r a ~ ohave r s to reserve a c e ~ a i na m o ~ n tof their c a p ~ c in i ~order to bid in the reserve ~ a r k e t s ,They are not

72

Power System ~ e s t ~ c ~and i n~g e r e ~ ~ a ~ i

e n c o ~ a ~ to e ddo SO unless they can make more money in the reserve markets than in the ecause of that reason, generators submit very high bids to the r e s e ~ e markets, resulting in n o n ~ c o ~ p e ~ i treserve i v e prices. ~on~spinning reserve has a relatively higher price than spinning reserve because there are insufficient pa~cipantsin the noninning res~rve~ a r k e For ~ . maintenance of system security, the IS0 has certain amount o f both reserves. Since non~spinnin~ res spinnin~reserve, the consequence is a higher price for a 10 r e s ~ is ~ not e as ‘worthy’ to the system as spinning reserve is). These exemplify ~ ~ k e t ine~~ciencies caused by unapt market rules.

Scotland, u n ~ i ~ e Since the commencement of energy privatisation in 1989 in the ~ n ~ l a nand d Wales, has not acquired a competitive and e f ~ c ~ e n t sale m a r ~ ~Also, t. for various reasons, Scottish customers have benefited much less than their counte and Wales, despite the fact that the England and Wales PO mar~inalgeneration COS& in Scotland, even after into ac~ountthe smission losses, interconnector access charges, r transmission and distribution are regulated using the ‘price-cap’ c o n ~ o l which depends on the inflation rate and electricity prices are set based on the pool ~ r i c e in s ~nglandand ales with a d j u s ~ e n t made s after ~ i n into g account the ~iffe~ences of the markets. land is chara~te~sed by a surplus of generation capacity on c a p a c i ~almost two limes the total maximum demand [2 g e n e r a ~ types o ~ inch ing dual oil and gas, coal-fired, hydro, pumpe~-storageand nuclear. The two ~ e n e r a t i o c~o m p ~ i e s , Scottish Power and S c o ~ ~ s~h y ~ o - ~ l have e c ~ c ~ n t e r c o ~ e c ~grids e d and Scottish Hydro-Elec~iccan access the grid in Eng~andvia i~ cottish Power’s transmission system. Even ough these two d o ~ i n a ~privatised ~ e ~ e r a t i o~no ~ p a n iremain e s ~ vertically inte~rat after ~rivatisation,they are re~uiredto keep separate accounts for separate busine s, i.e. tr~smission, d i s ~ i ~ u t i oC n .o ~ p e t i t i obetween ~ the two companies is made possible t h r o ~ g‘$econd~ tier suppliers’ who are autho~isedto supply ~ ~ e c ~toi customers c i ~ ou~sidetheir supply areas. ent trading in Scotland. Firstly, the ~ W Q otential obstacles to e ondly, the market is loo small to be ertically integrated. eneration capacity indicates that there i s compe~i~ve, Moreover, the substantial surplus no need to build new generators in the coming hture. Finally c ~ o c to ~ ~ ~with e t e Scottish Power or Scottish ~ y d r o - ~ l e c etween the two countries. In view of the above, and ~lectricity~ a r k e ~will s , focus QXI reforms for the Scottish markets which will remove the obstacles and be consis~en~ with the NETA [25].

The voluntary wholesale electricity market in New Zealand c o ~ ~ e n c eind 1996, but before that there had already been limited competition in the supply sector. It is operate^

73

etplace Company L (M-COLtd) which has recently b s in the C a l i f o ~ Pool, i ~ market p ~ c i p a n t sin New outside the pool through bilateral contracts, provided that the system o p e r a ~ is r ~ f o ~ e d of the ~ansactions. In New ~ e a l a n d~enerationis d o m ~ ~ a t by e d hydro power, which is located in the Island. The load c o n ~ e n ~ a t on e s the North Island which is connected to the ~ o u t h~sland by an HVDC interconnec~or.Even though the three gove~ment"ownedgeneration ~ o ~ p a n i ~ominate es the wholesale market, the market remains s spa rent t ~ o u g hthe broa~castof predicted prices and load forecast. Effort was spent only on i n ~ o d ~ ~ ~ i n g co~petitionin the retail sector between distibutor§ and the state-owne~generation c o ~ p a n ybut , it was soon realised that retail compet~tionalone was not e n o u ~ hto re elecwicity prices and hence the wholesale market was developed subsequently. The New Zealand spot market: is an ex post market featuring nodal pricing. Nodal on the theory of spot pricing [26].Under nodal pricing, if the arke et is s h o ~ - t eprice ~ signals so generated should enhance the efficient . ~owever,there have been ongoing discu§sions on effe~tive opera~ionof the m als and the management o f the losses and ~ o n s ~ a i n ~ .Expost pricing in the physical spot market is acco using the latest supply and demand bids and the actual measured plus losses actual demand is vital and it is one of the main roles of onciliadon Agreement. Final prices are published a few actual dispatch.

and debate, The Council for the European ~ n i o neven Afler years o f negot~a~on adopted Directive 96/92/EC in ~ e c e m ~ e1996 r to liberalise the e l e c ~ i c ~ tin8 y According to the ~~rective, members of the EU are required to open their y the year 2006 at least one-third of the EU-wide energy market will h rent European countries can liberalise their markets at their own pace, as long nts set by the directive are met, Apart from i n ~ o d u competition ~~g in the wholesale and retail sectors, the directive also features U . ~ o u n ~ ate sthe forefront of liberalisation include Spain and the Netherlands the existing one in England and Wales, will be developed hourly supply and demand bids, while in the N e t h e r l ~ d sthe Elec m ~ d a t e sa complete l ~ ~ e ~ ~ l i s aoft ithe o n generation section by the year 2 0 ~ However, ~ . there are also coun~ies,like France, Italy and Belgium, which keep their l~be~alisation ess to the minimum level requ~redby the direc~ivebecause o f domes~icpoIi~ica~ reasons. e ~ a n opened y its market to all suppliers and end users. As it is n ely few natural resources, two-thirds of the energy con~umedis imported from other countries, Effort in deregulation is therefore focused on the ~8intenanceof security of supply. Under the Energy Law A~endmentnet owners are required to provide o en access to facilitate competition. However, only 8 few out of about 700 net users have so far published the charges for using their networks [29]. At present, and nmst net owners also operate the grid; t~ereforethe issue of se~arationof owne~sh~p

7

Power System R e s ~ c t u ~ xand ig ~ere~lation

opera~ionwould need to be looked into, Also, practically small custo~er$have not bcen able to change their suppliers easily under the current legislation. The ~ e ~ project a n group on the energy market is ~ r a ~ an poten~~al g project sketch and it is likely that concept for the pote~~tia~ energy ark et will be similar to the EX [30] (European Energy Exchange). It is envisaged &at the d ~ v ~ l o p ~ofe n t ill be done step by step. The first step will be the ~ e v ~ l o p ~of e natfutures market where bilateral contracts can be traded ahead of time. Then a spot market will be founded for physical and short-term power trading. efore reaching that step, Gemany has to work on the i n ~ a s ~ and ~ ~r er~ e~ a t i o nfor s fast and rel~ab~e w ~ e e l ~which g is essential for efficient ~ ~ i ofnthegspot market.

Energy Information Administration,~ t t p : / / ~ . e i a . d o e . g o v / e m e u l ~ e ~ e l e c ~ c i . BTM consult Aps, hnp://www.btm.dWArticle~~ed-globaf/Eed-glo~al.h~. Stefmo Zamagni, Microeconomic Theory: An Introduction, Basil Blackwell, Oxford, 1987, John Bernard, Robert Ethier, Timothy Mount, William Schulze, Ray D, Zimmerman, Beqiang Gan, Carlos Murillo-Sachez, Rober J. Thornas and Richard Scbuler, ‘Markets for electric power: Experimental results for alternative auction institutions’, availablc via h ~ t p : / / ~ ~ . p s e r c . w i s c . e d ~ i n d e x ~ u b l i c a t i Proceedings o n s . ~ ~ l , of the Hawaii ~n~ernation~l Conjerenceon Sysfern Sciences, January 1997. John Bernard, Timothy Mount, William Schulze, Ray D. Zimmennan, Robert J, Thoinas and chard Schuler, ’Alternative auction institutions for purchasing electric power’, available via bttp://www.pserc.wise.edulpsercbin/tcsl/, 1998. Frank A. Wolak, and R. H. Patrick, ‘The impact of market rules and market structure on the price determination process in the England and Wales electricity market’, selected paper presented at the POWER Conference, March 1997, University of California, Berkeley, Berkeley, California, February 1997. Tim Mount, ‘Market power and price volatility in restructured ~~~e~ for electricity’, available via hnp://~.pserc.wisc.edulindex.gublications.html, November 1998. NETA, New Electricity Trading Arrangements for England and Wales, are based on proposals published by OFFER, Office of Electricity Regulation, July 1998, available via ~~:/t~.ofgem.gov.~~ lielix F. Wu, ‘Coordinatedmultilateral trades for electric power networks’, 12th Power Systems Compu~a~ion Conference, Dresden, August 1996. K.L. Lo and Z.Q. MO,‘Methods for determining wheeling rates’, submitted to the special issue o f International Journal of @stem Science on the Beslnicttmring of the Electric Power Industry, 2000. lgnacio 3. Perez-Arriaga, Hugh Rudnick and Walter 0.StadIin, 'international power system transmission open access experience’, IEEE Transactions on Power Systems, Vol. 10, No.1, February 1995. Young-Moon Park, Jong-Bae Park, Jung-Uk Lim and Jong-Ryul Won, ‘An analytical a ~ ~ r o a c ~ for transmis5ion costs allocation in transmission system’, IEEE Transactions on Power Systems, Vol. 13, No.4, November 1998.

Deregulation of Electric ~ t i ~ i t i e s

75

[13] J.W. Marangon Lima, M.V.F. Pereira and J.L.R Pereira, ‘An integrated f r ~ e w o r kfor cost

[14]

[lS]

[16]

[17]

a~locationin a mu~~-owned transmission system’, IEEE Transuct~onson Power ~ s ~ e m s , V01.10, No.2, May 1995. J.W. ~ a r a n g o nLima and E.J. de Oliveira, ‘The long-term impact o f transmission pricing’, IEEE ~ ~ n s a c t j o nons Power Systems, Vol. 13, No.4, November 1998. K. Lo, Y.S. h e n and L.A. Snider, ‘Congestion management in d e r e ~ l a ~ eelectricity d Conference on Power Utility ~ ~ r e ~ ~ a markets’, Proceedings of the I~~erna~ionul ~ e s t ~ c ~ u rand ~ nPower g T e c ~ n o l o ~ 2000, i e ~ City Universiw, London, IEEE, April 2000, pp.47-52. Michael D, Cadwalader, Scott M. Rarvey, William W. Hogan and Susan L. Pope, ‘Coord~~ation congestion relief across multiple regions’, Harvard Energy Policy Papers, available via ~ . k s g . h a r v a r d , ~ d ~ p e o ~ ~ e / w h o g a i ~October, ~ d e x . h1999. t~, R.S. Fang and A, . David, ‘Qptimal dispatch under transmission ~ontrac~s’, IEEE T r a n s a ~ ~ ~ oon n sPO Systems, Vol.14, No.2, May 1999. sco Galiana, Lester Fink, Power Systems R e s t ~ ~ E~i ~~~ i~n e~e rnand i r~~ ~: r Academic Publishers, 1998. ie and Ivar Wan~ensteen,‘The energy market in Norway and Sweden:

Energy Institute, September 1999, available via http://www.ucci.berkeley.edu/ucei.

[22] The Nordic Power Exchange, ‘The spot market’, available via w~.nordpool.com. [23] For derails and examples refer to, ’Zonal clearing market prices: A tutorial*, available via h~:/lwww.calpx.comtnews/publ~cations/in. [24] $ c o ~ i has ~ d 10,000 MW generation capacity against maxim^ demand of around 5,750 data taken from ‘Review of Scottish trading arrangements: A c ~ n ~ u l ~ adtoi cou~~~e n ~The ’, Office of Gas Electricity Markets, October 1999, availabIe via http://w.ofgem.gov.uW. [ZS] Details of future proposals can be found in the latest documents publishe~by O f g e via ~ its web site: h ~ ~ .ofgem.gov : l t ~ .uM, t261 Fred C. Schweppe, Mchael C. Caraminis, Richard D. Tablors and Roger E. Bohn, Spot Pricing o ~ ~ l e cKluwer t ~ ~Academic i ~ ~ ~ b l i ~ h e r1988. s, associated with a discussion of the losses and c o n s ~ ~ isurplus’9 n~ Marketplace Company Limited, July 1999, available via

[2S] Greenpeac~,G e ~ ~ a n *y ,S ~ r o ~inaDeutschland: r ~ Vom Monopol zum ell', ~ o v e ~ b ~ r 1898, a v a ~ ~ ~via b I h~://www.greenpeace.de. e [29] ~ n ~ o ~ a tobtained i o n in the ‘Strombijrsen’section at: h ~ t ~ : / / ~ . s t r o m . d e .

Un~versityCollege Dublin Ireland

Prof. Chen~ChingLiu Universi~of ~ashingtQn Seattle, USA

~ l e markets c throughout ~ ~ ~the world ~ ~are undergoing major chan es 111. These changes are varied in their nature but h e uiiderly~gtrend is towards a more CO and this results in electricity being traded as a c o m ~ o d ~ ~ e markets to facilitate this trade. Political forces [a33 are driving these changes. A compe~~tive electricity market is one in wh (ge~~erators) are c o m ~ e ~ i ntog sell their e l e c t r ~ cto~a~ number (loads). Here we are concerned with c o ~ ~ e ~ i tin i oanwholesale electricity ~ a r ~where e t the c~istomersare lap consumers or a retailer who will resell the e l e c ~ i to c ~th~ co~s~~ers. A l t ~ o ~ electric gh energy can be stored in batteries it w tities and hence ~ l e c t ~ is c ia~ r ~ a l - t i ~corn e i~stan~ly. The electrici~demand d also has a significant random Id in an ~ ~ ~~ ~c ris~energy, ~ ec t iTh ~ active ~ o w e rand au~omaticgen~ratorcontrQ~ Er that the electricity system can need to be ~ r o v ~ d eand d an e$ec~icitym of t ~ e s eservices [ 6 ] . The g ~ e r a t o ~ ically and K ~ c ~ olaws f ~ s system. The consequence o ystem and altering the s u ~ (g~~ierator ~ l ~ ou~uts) iates this c ~ ~ ~ ~ e s[7]. tion y, a n ~ ~ lservices l a ~ and

Competetive Wholesale Electricity arke et§

77

with the real-time stochastic nature o f the electricity deman makes des~gningan arket a great challenge. s in a wholesale electrici~market will be connected to the highsystem as opposed to the Iow-voltage distribution system. This ~ansmissionsystem an sports the electricity. In some markets single entities generati~gunits, transmission systems and supply the customers directly. These are ~o~ as vertically integrated utilities (VIUs) and can be monopolies. Where opol~esexist or where a ~ o ~ i n market a n ~ position is held in one part of the ~ n d u s t r ~ , c ~ i ~ agenera~ion, rl~ au~oritiesare implementing new market s ~ c ~ r toe es n c o u r a ~ ~ corn~et~tion [2,3]. It is ~ i f o accepted ~ ~ y that the transrnissi~n sys n ~ o ~ o p oand l y in this new environmen~it should be regulated to ensure open market [9]. Here it is assumed that all other aspects of the w market are competitiv~,a it is recognised that many who tive. For example, in Norway redefined limit are compensate limit are not [no]. Co~sumerdemand is largely inelastic but demandc o ~ p e ~ ~ t~i av ~ e k ei st tse c ~ i c a l l yfeasible and is becoming more CO In a monopo~istic ~ameworka re lated VIU makes pl isions based on a least cost objective, subject to constraints ( ~ ~ ic ~l ~ ti e~~ r 1~ ~a ~ 1This 3 1 . p ~ a ~and i ~operationa~ g process f scheduling algorithms, each one s roblem over a distinct time frame. ~nvo~ves econom ch ~~gorithms which achieve a real-ti and demand in a least cost manner. More advanced economic e consider the optimal the optimal ~ o w e rflow c o n ~ ~ a i n~ncluaing ts transmissi e limits, voltage levels, e frames unit commitment (UC) ~~~~~

s which are limited by these type

replaced by ~robabilisticmodels [IS]. Ln this ns are made. This planning and o ~ ~ r a t i o n e time for delivery approaches, the sc~edulesand ~ ~ s ~ aare tch to current circu~stan~es.

7

Power System ~

e

~ and ~

c

~

these markets result in cost m~nimisationin the short tern but their CO aspect should in the long Tun serve to reduce these costs even further. In the c o m ~ e ~ i t ~ v e market situation ~here€orea set of markets need to be developed that mimic the VIU least cost objective, subject to opera~io~al and re~iabi~~ty constrain~s.In p ~ i c u l are being replaced by markets for energy, transmission and Just as with scheduliiig algorithms these markets have di The real-time or ba~ancingmarkets are run very frequ~ntlyto main~hn~alance~ e ~ e e n supply and demand and to ensure system security and are similar to economic d~spatchand OPF al~orithms.In many markets there may be a need to run day-ahead ~ a r k e t that s will be like the unit ~ o ~ i t m eprocess nt [22]. L ~ n g - c~a ep a~c i ~markets may also be a feature hn some systems where €or reliability reasons generators are compensa~edfor keeping available capacity 1231. ~ o m p e t i t ~ velectricity e inarket design is a highly complex exercise not only by economic and engineering considerations but also by histo social cons~aints.Many of the current designs have ~ecQgnisabIe flaws ibutcd to both technical and non-tec~icali ~ ~ u e n c e s . to be assessed with these factors in mind. Lessons can be generally every market has particular c i r c u m s ~ c e swh ity market designs in di~erentcircums~ncescan be e q ~ i a ~e ~f yf ~ ~ ~ i v ired result, an efficient and reliable electricity supply. Different rna ~ i r c ~ s ~may c e also s roduce the same desired results. There i s no i ~ ~ k de ~ t olution to the complex problem o f e l e c ~ c m ators will agree that competitive e ~ e c t r i cmarkets i ~ ~ will resat1 society there are some very s i ~ i ~dif€erenc~s c ~ t of op~nionon some issues, These differe~cesof opinion can be d o ~ ~ ain~nature i c and s to cloud the issues. Each regiodcountry should choose a design tha ition but suits their particufar social, e c o ~ o and ~ ~political c e re a broad o ~ e ~ i eofw wholesale elecbi on of the independent system operator in which describes wholesale e~ectricitymarket charact~r~st~cs follows in ~ e c ~ i o n c~arac~erist~cs incl~deauctions, b i d d ~ nprici~ig, ~, fo~ard ential markets, congestion man ary services, physical and ~ n a n c ~m al s are given to illustrate these cha ty markets Section 3.4 describes le e ~ e c ~ imarkets c i ~ are still an active area o f rese the challenge^ in the design and opera~ionof these A c ~ o ~ ~ ~ d are g eInmSection e ~ ~ 3.6 and a CO S e c ~ i 3.7. o~

~

C o ~ ~ ~ t e~~hi vo el e Electricity s ~ ~ Markets

79

As more and more regions/countries open up their electricity markets to competit~on,the ~uestionof how to des~gnthe market in the best interests of the consumers and s u p p ~ i is e~ of prime importance. Central to this are the energy, transmissio and ancillary services markets and how they are coordinated. The competitive market demand ~ n c ~ ~ effect~vely ons in many markets, e.g. stock market delivery of a product is required by a stock market whereas in an electricity market a p r o ~ u must c ~ e v e n ~ a ~be l y de~iveredins~ntly(i.e. no storage) and its ower system. The closer we get to physical delivery th the operational and reliability constraints. These basic principles are n many ~ ~ c ~ i o nmarkets i n g and it is universally accepted that e i~dependen~ system opcrator (ISO). Although an accepted pri market structures require a large role for the the IS0 is a hotly debated topic. S w h ~ others l~ require roach. This operator disc rim in at^^ to all e market hence the operator. In general responsible for tasks such g of load for all users and e n s u ~ n gcom standards. The IS0 will o ~ ~ r the a~e , open access to the transmission grid to all users gestion and constraints on a n e ~ b o ~ ~ system reliability the IS0 should also hancements, The I S 0 may also pro led basis and perform the s e # ~ e ~ e ~ t of aspects that need ~ ~ a g i n ranging g, from c o ~ e c ~ i o~no l i c ~ econ~est~on s, management and the a ~ i n i s t r a t i o nof ene n po~iciesare an ~ m p o ~ a naspect t of the ESO res~onsibili~~es. and charges that all participants must meet in order to connect to the grid and in the marke~.The trans~ssionsystem is made up of a e n e r a ~ r sand cons~mersare located, and these buses are lines. These lines transport the electrical energy around the hi t r a n s ~ ~ s ssystem i o ~ and have limited capacities, which for security [26], When a line is at its limit the system is congested er inj~ctionsat every node in the system. Relieving th generator^ andfor consumers to alter their quantities. ~ ~ e f o congestion r e puts a c o ~ s ~ r aon i ~the t e n e r markets ~ and in many instances may render them non-compe~itiv~ [27]. Losses on an e l e ~ ~ i grid c i ~are u n a v o i ~ b l eand can be substantial. A market nt’s p~ysicallocation on the transmission system, i.e. portan ant factor in wholesale electricity markets. Th ~ d ~ e n of~ establishing a ~ s the instantaneous Iocati of electricity. For e x a ~ p i e a, generator that is injecting p o w ~ r location at one i n s ~ nin t time can cause substa~tiaIlydifferent losses and c o n ~ e ~ tthan i o ~a similar i n j e c t ~ oat~another ~ location andlor time. The cost of these loss sm~ss~on s y ~ t ne ~e to~be~~ l ~ o c ain t esome ~ manner to the

Power System ~ e s ~ c and ~ ~ n g

electricity market and this is not a trivial task [29]. The revenues collecte~by the TSO from the ~enera~or§ and loads for these ~ansmiss~on s e ~ ~ c e(co~ection, § age, ay for the ~ a n s ~ i s s i o§ ny s t e ~in tbe short an

In the VHU environ~entthe least cost objective ~ p ~ c a rl e~~ye ~ ~e idi l yto the cost of a n 6 i l l a ~services such as r ~ s e and ~ e vol~dge~ o n t r owere ~ ~ e a t as e~ opt ~ ~ is atioprocess n and their cost may not have been e x p l i ~ i ~ ~ y illary services is costly, and the ~ u a ~ ~ ~of ~ ~ ~ reserv~are services that generat~ngunits provide they have significant costs associated wit11 them D23. ~ ~ ~ will~ a not provide these services unless they are ade~uatelycompensa~ed[33]. In s o ~ cases, e howev~r,g ~ n ~ r a ~ may o r § be obliged to provide these services in order to be all~wedto arket. Ancillary services can be self-~~ovided by the e ~ ~ r g y nsible for a ~ q u the ~ ~ n ce. ~ ~ h y s i c a ~selfly ~ r o ~ i s i af Q nthese s ient and ener~y~~k~~ these services from others. Therefore in c o ~ p e t i t whole ~v~

a state where load shed and reserve that must rocess, then the ex

~ncentiveto ~ a i n t a units i ~ [41]. A strong ar fines are not n ~ ~ e s sasa pure ~ , market forces in the competitive enviro the event of a shortfall in gen

In a ~ ~ o ~ eelectricity s a l ~ market ~ u l t i ~ l e being traded over d ~ $ ian~e li e~~ ~ r i ~ i ~ number of choices a t ebasic c ~ a r a c ~ e ~ s t i c s s ~ s model ~ e ~is used to i ~ ~ u s ~ the

ity markets are used here to illus refe~encesto existing wholesale ele eneral these exist in^ m characteri§ti~s but it should be no ecul~ar~t~es which do not allow them to be c lso be noted that even at the time of this writing rn re a v a ~ ~ a bthel ~ relevant web sites are giv the r sh

3.3.1

~ r nTest ~ ~~ yl s ~ e r n test system c o n s ~ s ~ ~ofnag supply si d a simple ~ ~ e e - bnetwork. us

~ i n i ~ uand mr n a x ~ gen ~u~

tors have quadratic production cost con§~aintsgiven by

are the power outputs in MW of generator #1 and

ic utility curves and

n i m ~ mand max~mum

Power System ~ e ~ t ~ a ~d ~ i n g

2

Line A

(3.5)

Line AC

(3.6)

The coef~cientsof P, and Pzare the line sen~~~ivities of the respectiv~lines to inj~ctionsat buses r~spectively[27].

us A

Line A us

s ~central s ~ auction istinct ~ a d ~ ~n g~ c h a n ithe iers and ~ustomersboth s u b ~ i t the market clears, i.e. d e t e ~ i n ~ § m [46]. In their simplest forms these centralised auctions to a §im~le merit order economic ~ s p ~ a~l gc oh~ i [12]. ~ h ~ The d auction for auction m e c ~ a n i s ~ ~ ~~~~~~

Competetive Wholesale Electricity Markets

3.3.3

3

~ddin

idding into a simple central auction i s similar to the process of each generator submi~i cost data and each load submi~ingutility ( ~ l l i ~ ~ e s s - t o - pdata a y )to the used by the VIU to dispatch the system. In an ideal world with a electricity market the bid data should be the same as the ~roductioncost (utility) data or o p p o ~ n i t ycost, wRicRever is eater. The o p p o ~ n cost i ~ is the r~venue p ~ i c i p would ~t expect to get by selling in a different market. This price assu~ptionin a competitiv~market is an optimal strategy for a market particip n the The p ~ c i nmechan~sm ~ i an important factor in this p r i c e - ~ i n ga s s ~ ~ p t i oand to the seminal paper by Vickrey [49]. The fixed costs are not d ~ u a n ~ i t yi.e., clearing the market. The incrementa1 costs (ut~~ities) are all that are needed to clear the market. Here it will be assumed no opportunity costs and that all market p ~ i c i p a n t sbid at ~ c r e m ~cost n~l case where bids vary from incremental cost (utility) is dealt with later in the section on ~ S e ~ t i o3.3.9). n The cost (uti~ity)curves and the increme~talcos small test system are given in ~ ~ ~3.2r and e 3.3 s respective~y. T (utility) curves result in linear increnienta~cost (utility) curves. 20

w

0

100

200

300

400 500 BOO Power (MW)

Cost ( ~ t i lcurves i ~ ~ for the small test system

700

5 n u

0

100

200

300

400

500

600

7

3 I ~ ~ r e ~ e icost i ~ a(utility) l curves for the small test system

C o ~ ~ e t e ~Wholesale ive Electricity Markets

Profits for the ~ e n ~ r aare ~ ~c ar ~s c u ~by a ~t ea ~~ i the n ~di~erencebetween the r e ~ e n ~ ~ e cost. The cost in t h ~ calcu~ations §~ is taken to be the i~nore§other fixed costs such as eapit ce ~ e ~ e the e nutility p r ~ c~i~~c h~ a n i sisi npay as you bid w h ~ pra~~ i c i ~ a pn ~ s is prQ~osedthat this type of d i s c r ~ ~ i n a ~ opricing ry wi

ts (3.11, (3.2), (3.3) and (3.4) and (be load balanc~c o ~ s ~ a ~

roce~§ (a ~ ~ a d r a ~p ir co g ~ a ~ i n g ils of solution) the no-load and fixed et in h i s m ~ n e with~ut r amb ~ ~ c r e a s ~[ndg~ c r e a s ~ ~ g ~ ,

Power System R ~ s ~ c and ~ ~n e rge ~ l a t i Q n

constraint (3.9) and the assumption of a lossless system, the pool (central auction) is revenue neutral, i.e. what is paid in by the loads is paid out to the ~enerators, le 3.1 Market clearing, transmission uncQns~ained G e n ~ ~ ~ Q r / Quantity ~ Q a ~ (MW) Generator #I 313.6 ~eneratQr #2 409. I Load #\ 522.7 Load #2 200.00

3.3.5

Price ( 18.3 18.3 18.3 18.3

$

~

~ Profit ) ( S u ~ l ~($A) s)

683.7 21 10.3 437 1.9 4345.5

~ a rTiming ~ ~ t

to the stochas~icnature of the demand [ and the need to s c h e d ~ e~ e ~ e r a t i o n resources in advance, electricity markets can characterised by timing. Forward s are run in advance of the delivery time. This enables suppliers to e~erationto meet the demand and for the IS0 to coo~dinate~ a n s r n i ~ s and i~n ervice needs. The forward markets also perform a very important financial ants by locking in prices an ractice [SZ]. In power I-time (spot) markets and is systems with large themal plants that are in unit constraints these 1. There may be a m u ~ t i ~ of d ef o ~ ~ ~a r kd e at ~s ahead, month ahead and day ahead. In C a l i f o ~ i athe power (PX) m s three different types of forward markets [53,54]. The day-ahead lishes prices and quantity of electricity for delivery d ~ each ~ how n of ~the . The day-of~our-aheadmarket o erates similar to the day ding closer to the delivery hour. pa~icipantscan buy and sell energy months in ad order of min~tesin ance of delivery are deeme a ~ p r o ~ ~ real-time hes markets are needed to ensure supply and adapt to unforesee~c~rcumstances.These real-time markets are in d Norway the respecti~elSOs operate market for real-time a d ~ u s ~ [55, e n ~101. The p r e a ~ -basis ~ i ~but~ this is set to change with the ~ t r o d u c ~ i oofn a binding day-a~ead~ a r ~ e t E567.

The core product being soid in electricity markets is energy. U l t i ~ a ~ e the l y coordi~&ti~n of units (sc~eduling)and of the ~ransmissionand a n c i 1 1 services ~ enables its seque~tia~ e l e c ~ c market i~ structure is one in which the energy t ~ a ~ ~ d ndently of the transmission and ancillary services. The provision of the e ~ e r g y~ a d in~ an ~ tlal trans~issionand ancillary services needs follows sequentia~~ ~ a n n eIn r . ~ a l ~ f oforward ~ i a energy ~ a r k e t are s con real-tinre energy, congestion management and ancillary services a l i f o ~ ~[58]. i a There i s a strong physicai coup1

7

Compete~iveWholesale Electricity

s e ~ i c e and s congest~onmanagement and this is reflec~ed some markets where cbre is one in which ded s ~ ~ u l t a n ~ o u sAl ys~muItaneo~s . electricity market a n ~ ~ u s lwith y the transmission and ~ c i l l a r ys ania, New Jersey, Maryland, USA) ~ n t e ~ c o ~ e [60] c~~on this simultaneous c ~ a r a c ~ e ~ § t i cThe . i ~ t ~ ~ r n r pool is to use a hybri se~uent~a~simul~aneous arket structure in Alberta may also be a h y b ~ $as the the energy market and the ancillary services markets as eously [56]. In the uncons ined m a r k e t - c ~ e ~ n g the ~ a n s ~ i s s i line o n power ws are given in Table 3.2. le 3.2 Power flow, market clearing, transmission ~ n c o n s ~ ~ n e d

Line AC (PAc) Line BC (PBc)

-100 200 400

337.5 385.2

The line c ~ n ~ e cbus t iA~to~ bus C is overloaded by 137.5 M with this ~ o n g e s ~ is o nto clear the ~nergyand transmission mar h the e x a ~ that ~ ~ will ~ e be given here will deal with ~imulta~eouslyit c to deal with energy smission constraints in the ed line. This can be a c ~ ~ e v e d social welfare (3.8) subject to unit const~aints(3.1), (3.2), (3.3) a ons strain^ (3.9) and the transmission ~ons~raint (3.5), (3.6) and hic that illus~atesthis arke et-clea~ng m e c h a n i s ~ ~ c o n s ~ a ~ nen€~rced ts Table 3.3 gives the ~uant~ties, prices and p the power flows. Market clearing with transmission constraints

Generator #2 h a d #I Load #2

467.0 378.0 200.00

20.0 22.9 22.9

2871.1 228~.0 3419.1.

Power flow, market clearing with ~ransmissio~ cons~aints Line

Line flows (MW)

Line limits &IW) -100

Line AC (PAJ

200.0

200 400

structuring and Deregulation

ce the ~ n ~ o d u ~ tof i othe n ~ o t i c that e in Table 3.3 the price at each bus is differe price at each bus is the term locationa~marginal pricing or nodal 1 cost of the next ~ ~ g a w aoft t po I is active then typically the price at each bus ipants at different buses receive (pay) a d~f~erent price and this is the ~ c r e ~ e n tcost a l i s different at d~fferentlocat~~ns, ~o~ational the appropriate price signals regard in^ their location. ~ e n e ~ t o r #l is poorly located in co~parisQnwith generator #2 as it is

~ e ~ thee ~oe buses. ~ ain as the revenue for the generators will be les

and New ~ e a l ~ ~

can r e a ~ a n gthe ~ result of the ener

~ o m ~ e t e t i Wholesale ve Electricity Markers

~ q ~ a n and t i ~price) ar is trading ap~roachhas the ut the cheapest generators. ot c ~ e n t l y~ e ~ i in~ e d must be traded through the cen is set to change in Engl may be net inject~onswhich may c n management process these If these transacti m have been changed as se bilateral trade transmission c o ~ $ ~ a i nfor t s the central auction: Line A

ine A Li

(3.12)

arket clearing with transmission constraints and bilateral ~ ~ ~ s a ~ ~ ~ o

~ e n e ~ a#It o ~ ~ e n ~ r a tfc2 or Load #I Load #2

92.6 472.2 364.8 200.0

13.9 20.2 23.3 23.3

-214.3 2945.3 2129.6

Power System Restructuring and

9

Table 3.6 Power flow, market clearing with transmission constraints and bilaterals Line Line AB (P&) Line AG (PAJ

Line Rows (MW) -97.4 200.0

Line limits (MW) -100 200

The bilateral trades have altered the central market result. In order for be allowed they need to pay for the tran§m~ss~on service. The ~ ~ s ~ i s b i l ~ t e ~W1 a l is the product of the q u ~(10t MW) ~ ~by the i n c r e m ~ ncost ~ l of ~rans~ission between bus A and bus C ((23.3-13.9) $ i.e. 94 $/h. The ~ ~ s m i s § i charge on for product of the quantity ntal cost of ~ansmi§sion d bus G ((2~.3-20.2)$ / ~ i.W ~ ~ ~ ilatera~trades re in a d~rect~on that relieved congestion the price diff~rential~ would be negativ ~ ~ ~ s m i s s charge i o n would be negativ~,i.e. the bilateral trade would be re estion the bilateral trades can be ~Qnductedindep stem becomes c ~ n g e s ~ ethen d there to pay the ~ a n s ~ i s s ~ charge o n these ilatera~~ a d e have s been eEe central auction. This concept is r e c o ~ n i s ~indNorway where zonal ement ~ u ~ o § bilateral es trades b e ~ e e nzones c e ~ ~auc~ion al [20J P ith locarional (nodal is mandatory partic

able to match the p a ~ [67]. e ~ It is ng schedul~from a central a~iction

Competetive Wholesale Els&icity Mmkets

1

scheme in the ~ a l i f o ~Pi a [54]. It is interesting to note that this iterative bid~ingscheme proposed for California proved impractical and has not been In the VIU envi~onmentgenerators were typically This UC a ~ g o ~uses th~ cost ts and accounts for the inter r a m ~ ~ nrates g [l5]. In so need to be ~ n t e ~ a l i s eind the bids of the p a ~ i c i p ~16 ts the prices in advance [69] and bid so that the pro~table. This self-schedu~~ng approach is in existence in the ~ a ~ i f o r n ~ ~ and Norway [70]. Bilateral trades are by their nature self-s security reasons, self-scheduling may be subject to approval by the IS0 "711. a central auction process can also involve a firm that owns mui~pleunits submitting portfolio bids. These bids represent an aggregate offer. Afier market clearing the firm can then decide how it will schedule its own units to supply the q u ~ t i t i e s ,The CalPX allows portfolio bids. An alternative to self-scheduljng is centralised scheduling where a UC-type algorit i this auction m e c h ~ i s mis very is used to clear the market [41]. ~ i d d i n gi n f o ~ a t i o iin ~ e t ~ ~ ~n ~~ e~ ~all u 9dcost ~ gdata and a p p r Q ~ ~ atechnical te cons~aints.In the clearing examples above the optimisation problem variables were which are of social con~inuo~s.In a centrally scheduled system the objective is th welfare, subject to ~onstraiiits9 but the variables are both continuous (quan~ities~ and discrete (turn a generator on or off) [72]. In PJM some units can choos scheduled while others with bilateral contracts can self-schedule. In the e n e r market ~ i s a centrally optimised UC process but this is set to change

In perfectly competitive electricity markets the most profitable swate p ~ i c i p a n tis to act as a price taker and bid at incremental cost [48]. a s s ~ p t i o nassumes an infinite number of competitors so the bidding beh player cannot affect the markets, i.e. influence prices. In the real world, however, there are only a finite number of market participants and each participant has to some degre to increase their own profits. There is a mu power, i.e. they can bid s~a~egically possibilities for this type of gaming behaviour in ellectricity markets [73,74]. De bids from ~ncremen~al cost can also occur because a p ~ i c i p a n wants t to en schedule [66] or it knows it has another opportunity in anorher market incremental opportunity cost. As an example consider generator #I in the constrained market above. i n c r e m e ~cost ~ ~~ e n e r a ~#oIr is making a loss (Table 3.3). In this: case ge alter its bidding strategy so as to avoid this loss. Table 3.7 gives the result for one strategy wbere gen~ratQr #I ~ncrease~ the linear part of its bid (3.1) from 12 ~ ~ to W 14 h

Power System R e s t ~ ~ and ~ n g

--

atkec clearing with ~ansmissionconstrain~§and ~ e n ~ s a tktoltbidding strategically, i.e. the linear coefficient is changed from 12 $ Mto 14 ~$MWh ~enerator/Load ~enesator#I Generator R Load #1 Goad #2

Quantity (MW) 106.3 48 1.3 387.6 200.0

Price ($/NIwh) 16.1 20.4

22.6 22.6

Profit (Surplus) ($/h) 25.2 3075.1 2403.2 3480.4

erator # I by a~tenngits bid away from ~ n c r e m e ncost ~ l has tu le 3.3) into a profit of 25.2 $/h (Table 3.7). ~enerator## 2 incre~ingits profits from 2871.1 $k(Table 3.3) to 3075.1 $k(Tab 3.7). Load# 1 and 2 have both also gained as their surpluses have increased. y b ~ d d ~above g its ntal cost generator # 1 has increased its price and reduced s quantiw and most are a balance betureen ~ c r e a s price ~ g and reduc q~antity, ~ i t inelastic h e is more scope for driving up prices w~thoutexcessi ads can also bid strategically. The price differenti bus B: has reduc~d(T 7) and hence the loser in this g revenue from the c ~ ~ g e s t i o n ment has reduced from 23 15 ties and these exist in here pa~icipan~s are in collusion 176,733. Cen the market power. ~ o w ~ v e tr r, ~ ~ ~ i s s i o n mall pockets with very few ~ ~ i c i In p this ~ ~ . s with little market power ~ l o b a ~cl y ransmission systems that are prone to tive electricity market difficult. S their nature [26] pose s i ~ i l adiffic r ed that must run for r e ~ ~ a b i lreas ~w R) generators are ~ o ~ p e n s a t outsi ed market results [8OJ. In the example above it should be noted that the strategy of g~nerator# I is not op~imal. i n ~ re~u~es ptirnal s ~ a ~ e g ican e s be found, however [S1,821. Successful ~ a ~behaviour ts to have good information about other p a ~ i c ~ pbi~ ~ and s ’ to consider the n a ~ r of e the p r o b ~ [83]. e~ onstra~~ts and g a ~ i n gbehaviour act to reduce social we~fare. Table 3.8 below the social welfare for some of the ex sion c o n s ~ a ~and t s all ~ a ~ c i p ~ t s social welf~reis a ~ a x i m u m .With the ~ a n s m ~ § lace of cheaper power and the social we1 are bi~dingaway from ~ n c r e ~ e n tcost a l ( ~ t i l i then ~ ) the social we~fareis her r~duced,

Coin~etetiv~ Wholesale Electricity Markets

Social welfare

3.3.10

Market

Social welfare ($h)

No t ~ ~ s ~ i s sconstra~nts iQn tab^^ 3.1) ~ r a n s ~ i s s ~constraints on (Table 3.3) Gaming and transmission constraints (Table 3.7)

11,511

10,715 10,711

A n c ~ i Services ~u~

s e ~ i c e sare required for the reliable operation of the power sys s ~ n ~dre d~ n i ~ of ~ othese n services is not globally accepted. AGC, resesve (s s t a ~ d b yload ~ , fo~~owing, v o ~ ~ control ge and b ~ a c ~ - sct aa~p a ~w~ oi u~ be ~~ ised services. The generat5rs ically provide these ~ n c ~ l ~ a can also provide some. New these s e ~ i c e are s not term contract. Some of ~ ~ k eInt ~. J AGC ~ is, se ~ the IS8 operates a they are L ~ w i orl ~ ~ ~then acquired by cons~aining e process of a c q u ~ this ~n~ energy and ~ansmi§s~on congestion manageme~tmarkets. Consider the simple test system. Assume that the IS spi~iingreserve is required for system r e l i a b ~ and ~i~ advance by acquiring it in a ~ o ~ amarket, r d e.g. day ahead. Spinning reserve is the a~~~~~ of an on-line erator tor (bad) to increase (decrease) its output (c period of time. The time per~odwill be d e t e ~ ~ by e dthe s y s t e ~ t for smaller s y s ~ e ~ s the time period is nemlly smaller in order to avoid large ~ e ~ u ~eviati5ns e ~ ~ c[38,86]. ~ Assume that gesler r # 1 and generator #- 2 can ramp up by 25% and 50% respectively of acihes in the s ~ ~ ~reserve i n gtime period. It is also ass loads are ~ncapableof p r o v i d ~ spinn~n ~ ~ reserve. Therefore the ~ e ~ ~ ~ e

1I- 0 . 5 ~ ~ 0-0P2)2 200 Table 3.

(3.13)

the r e s ~ of ~ ~clear s the market with the above constraint (3.13) m~ssionc o ~ s ~ a i n t s

Power System ~ e s ~ c ~and ~ ~i ne rg e ~ I a ~ ~

Market clearing, reserve constraint, transmission un~ons~ained _ _ . I I _ .

G e n e ~ a ~ o r ~ oQuantity a~ (MW) Generator #1 3 16.9 291.5 Generator #2 Load #I 408.4 Load #2

200.00

Price ($/MWh)

Profit (Surplus) ($h)

21.9

1842.4

21.9

2969.2 2669.3

21.9 21.9

3614.1

The first thing to notice about Table 3.9 is tkdt in comparison with Table 3.1 the quantities have altered substantially. In order to meet the reserve c o n s ~ a i n(3.13) ~ g e n ~ a t o#2 r has had its quantity reduced and loa is largely unchanged and load #2 is unchanged. Although generator reduc~ionin quantity it i s more profitable than the unc~nstrainedcas reason for this is that the price has increased. Although generator #2 cannot complain about its profits. The biggest gainer out o f this si whose profits have more than tripled. This high1 with ~ e c ~ i cparameters, al i.e. generator # 1 has It should be noted that if both generator #I and #2 had the ability to ramp up to m a x ~ ~ u m output within the s p ~ n i n greserve time period then the market would clear at the same price and q ~ ~ ~asi int Table y 3.1, i.e. the reserve constraint (3.13) will not be ~ ~ n d i n g . Here the binding reserve constraint has caused the social welfare to reduce to 11095 $/h from 11511 $/h in the ~ c ~ n s ~ i case n e (Table d 3.8). It shou~dalso be noted that in the event o f this reserve being used then generators # I and #2 woufd be paid the real-ti~e price for their energy. This scenario, where both g e n ~ r ~are t o bettcr ~ off because o f the ~ c ~services, ~ l i sanot~always the case and eref fore if a constraint causes a red~ctionin profits a p ~ i c i p a n tshould be compensated for its o p p o ~ n cost i ~ [60]. The ~ y b r ~ d approach in the New England Pool requires the ca~culationof this o p p o ~ cost n ~ for ~ An alte~ativeapproach for the provision of ancillary services is to set up m ~ ~ k e t In s. a competitive e n v i r o ~ e n the t bid curves for reserve and other ancillary services should reflect a pa~cipant’sexpected o p p o cost. ~ ~ Expected ~ o p ~ o ~ ~ ncost i t ywill require f o ~ e c ~ s tofthe i ~ g energy spot price [69]. In C a l i f o ~ i the a ancillary services markets fo in sequence after the energy and congestion management markets. In this way capaci progressive~ya s s i ~ e dto the various tasks 1551. In New Zealand the ~eservem ~ k e is t cleared simul~neouslywith the energy and transmission markets. With ~ a n s ~ i s s i oand n reserve constrai~tsthere may be a need to account for the interact~Qn between the two, i.e. in the event that reserve is needed it will require ~ a ~ s m ~ s s[87]. ion

3.3.11

ieal and Finan~ialMarkets

~ a r k e can ~ § be physical or financial. If the markeE is physical then the quantities are to be physic~llydelivered in contrast to a financial arke et where no p ~ ~ s i c da le ~ i v eis~ reqMired. In advance of physical d e l i v e ~the IS may well receive i n f o ~ a t i o nthat is ~ndicat~ve of the physical deliveries. However, at some point in time the I ~ n f ~ of~ the e dbin~ingphysical c o m m i ~ e n t sso it can c o o r ~ i ~ athe te

C o ~ ~ e t eWholesale t ~ ~ e Electricity Markets

security and reIiabi1~~.Deviations from these binding c o ~ i ~ are e n ~ with by buying or selli the differences at the real-time price. In C a l i f o ~ submit binding schedu and any imba~ancesare adjusted in the real-time m ~ k e that t is operated by the CAISO. ecause of price volatility many ~articipantsin a central auction rocess may wish to acquire financial contracts which hedge their position. In Alberta effectively hedged against the pool price. Alberta currently has only and this has been possible because of the large-scale hedging w p a ~ i c i ~fiom ~ t sprice volatility, This situation is set to change wi being introduced in the near future 1561. teral trading is one m e c h ~ i that s ~ can be used to hedge the vola~lityin a central a load that are participating in a central auction can ~iavea a price P,,. If the pool market has a uniform price of P, then and the load pays the same amount. his can be ach~eved at zero price and the load requesting MW and i n d ~ c ~ t ~ g he market price P, is higher than oad. If the market price P,, is lower than the P, - P,) to the generator. The net effect is W at a price P,,. The two that thc ~eneratorand load h of the uniform price. The s bil~teralc o ~ ~ aare c t perfectly hedge is known as a Contract for Difference (CFD). ~ o w e v e r ,this hedging mechanism is u n d e ~ i n e din a system that congested and has a ~ocationalcongestion ~ a n a g c m e nsystem. ~ If are at the same bus then the hedge is still perfect. If, however, the I different buses they will have to pay for transferring Q MW from the gen ~ be revenue depending on the price d~fferenti load bus. This p ~ y n i e ncould and the generator split this payment (revenue) between them is their own business. There involved: the IS0 which collects the charges for co~ges~ion ~ a ~ i o nprices a ~ can be very variable and ~ e n c ethe price of ly voIatile. A solution to this difficulty is the conc ~ r a ~ s ~ i s srights ion ~ ~ where T pa~icipants ~ ) can in advance ~ ~ r c h a fr se right to collect the ~ a ~ s ~ i s scharge i o n for Q MW between two buses [63]. load and generator are again hedged. If these transmission rights are compe~~tively traded then their price should reflect the expected price ~~fferential b ~ ~ e the e n load an generator buses. The IS0 must also ensure that these ~ a n s m ~ s s ~rights o n are feasible9i.e. its t r a n s ~ i s s i ocongestion ~ income in the physical market covers the p these rights. Trans~issionrights may also be subject to gaming behav~our ex~s~ence of ~ u l t i p l etrading o p p o ~ n i t i e (bilateral, s spot market, forwar to optimise their portfcdios the wholesale electricity market ~ a ~ i c i ~ win a n ~endeavour s [89,90].

From the el~ctr~city arke et characte~sticsde§c~~bed above it is eviden~that there is a esign choices for electrici lco r n o ~and ~ l the ~ i ~ a t e r a l sion and ~ ~ c i ls~e ~ai rc e~s

pool and bilateral aspects with

rates all ~ h y s markets ~ ~ a ~ forward

and cost ( u t ~ inform i ~ ~ ~

p a ~ c i p a n self-schedule. t~ This model oflers all the ~ e n ~of~ co~prehensive t § CO icipa~io~ in the cen c ~ ~ i coif sthis ~ type of ically not ~ i q u and e the is very sensitive to algorith~p ~ a ~ e t e which rs could lead consequence of the i~tegern opti~~§ation proce dificulty the prices are set by lex algorithm whic

oach should m i ~ i ~ i any se

~ l e ~markets ~ c are i ~highly complex systems that consist of a number of ~ t ~ e l a t e d m ~ k e t sfor different commodities (energy, ~ansmissionand a n c i ~ services) l~~ and different time frames (real-time, hour ahead and day ahead). There are still man lems in the design and operation of e l e c ~ ~ i tmar~ets. y when the pure economic theory is applied to a power syst e economists want the electricity markets to embrace the laws d with simply ideal examples they can show the benefits af such a . The real-time nature, physical constraints and reliability issue all act to make the development of an ideal market impossible. It should be noted that it is well accepted that all markets, even those for simple c o ~ o d i t i eare ~ , not ideal. T h e ~ e f o r ~ the goal should be to develop a market that is a bestfit to the ideal. Several wholesale electricity markets have been established around the world and most of these are in a con~nuousprocess of change. This evolut by the need to address some of the outstanding issues in the these markets. Here some of these challenges are outlined.

3.5.I

~ a rPower ~ ~Evulualion t and ~itig~lion

e t valuation of market models can have many differe~tv i e ~ o i n ~ sThe . ~ a r ~ must dmction in a reliable, efgcient and fair manner. The generators will want to maximise their profits t ~ o u g hthe markets. The consume^ will seek the best value for the service they receive which may conflict with the aims of the generators [SS]. This will ne~essitate analysin~the social benefit that the market offers and the prices that are charged. It will also be dent to ensure that market power and gaming do not exist and that m ~ k e t are s not overly volatile. on there are some a v a i l a b ~simuiat~on ~ and a ~ a l ~tools. ic simulation model that considers the market s ~ c t u r and e estimates and ~uantities.Kumar and SheblB [93] have dev~lopedan auctio~ s et al. @ I ] have developed a framework to in market simulator. supply Green when all p ~ c i p a n t are s maximising their own and ~ e w b [94] e ~investigated the UK market using the supply curve T h ~ is~ little e doubt that market power is bei exercised regularly in many electricity rices, which are well above co~petitive m ~ r k [95,96~. ~ ~ s This practice is characterised ~ y profitable for the ge~eratorand ult~matelycostly for levels, The result is ~ i c a l very s power can be exercised in many ways. aerators with global market power can manipulate the marginal (spot) price as in the gland and Wales p o ~ e pool r [96]. ~ransmissioncongest~~n can give p ~ i c i p a n t slocal market power and they can ~ a n i ~the ~ Iocational ~ ~ t e marginal prices 1971. Some possible solutions to this problem i n c ~ u dthe ~ following [76]:

Competelive Wholesale Electricity Markets

99

Better market design. Some markets have experienced difficulties, which could be resolved by better design [24]. The congestion management process in California has a gaming problem and the Federal Energy Regulatory Commission (FEW) appears to be encouraging the adoption of locational marginal pricing as a solution [3,64]. reaking up the large generating companies into smaller co~petitiveWI oliticd issue, which may not fully solve the problem. In the perceived that the two dominant generation companies exercised their market power to raise prices above competit~velevels [96]. ~ u ~ more ~ ~~ansmission n g so as to avoid creating o p ~ o ~ n i t i efor s local m a r ~ e t power. Over-building transmission may seem wasteful but with this ~ a n s r n ~ s ~ ~ o n capacity in place local market power can be removed and generators may act more e ~ ~ v e [8]. l y This additional ~ansmissionwill also increase the r e l i a b i ~ oi ~f the system. There is, however, significant environmental concerns related to b ~ ~ d i more ng transmiss~onlines. n here the load is Making the load more responsive to price. In the examples responsive (3.3); however, this masks the reality where in mo markets the load i s largely inelastic. Any generator hoping to find that a responsive load will reduce its quantity and reduce . For domestic customers this may be very difficult to i e customers may be capable of ~ n s t a ~ ~equipment ing that can respond to the m a r ~ e t e. In the long run new technologies may make distributed generation (e.g. fuel cells) more prevalent and this will reduce the need for further investment in transmission [%I. It will also combat market power, in particular, if this type of generation is owned by groups of consumers (i.e. if the market price is too high they will generate themselves). If this does happen then the electricity market will become part o f a larger energy market. In some markets if the price rises above certain levels the prices are capped; however, this distorts the price s a1 and may have long-term negative consequences~ Price capping has been used at one time or another in most wholesale electricity markets. For e x ~ ~ ~theI California e, ancillary services markets have had price caps i m p ~ ~ [79]. ed

3.5.2

~ y s t Capacity e~

The issue of p l a ~ i n gin generation and transmission must be a d ~ e s s e dwith a view to ma~ntenanceand enhancemen~sto meet increasing demand. On the generation side these functions are generally left to the market, the assumption being that energy prices will signal the best times to maintain units and when to build new plant. The energy price spikes in the Mid West (USA) in June 1998 highlight this issue. A market for generating capacity over a fonger time frame (more than one year) may provide the necessary market rs signals h ensure that the system will expand according to the needs of the c o ~ s u m ~1233. The concept of marginal cost pricing [99,100] for electricity is based on ~ndamental microeconomic principles [50]. In an ideal market bidding at ~ n c r e m e ncost ~ is an optima^ strategy [48]. However, the resulting schedule may be unprofi~ablebecause of costs such as no-load costs, startup costs and fixed costs (Table 3.3). In the VRJ environment with spot pricing Schweppe et al. (281 introduced the concept of revenue reconci~~a~ion where ~ a r g i n apricing ~ may not be sufficient to cover all costs and give a

Power System R e ~ ~ ~and ~ Dere r i n ~

1

reasonab~eprofit. ln competitive markets revenue reconciliation shou~dbe redun that m ~ ~cost~ pricing a l will in the long i-un resolve this issue. In the lon not suf~cientto cover business. However, this i in the ~ n g l a arid ~ d Wales power pool ere are c a p a ~ pi ~ a~ents p ~ i c i p a n t sreceive in addition to the market spot ~ a l i f o no ~ ~such a st in the energy market but there is ctures would appear to fail in in new trans~ission [8,102]. This may be a p r ~ a ~ r e tra~smissio~ i n v e s ~ e is n ~a long-term issue [30 d the markets have only been recently i n t r ~ ~ u c e dAlso . it could be argued that the tran sion system was over-bu~ltin the past and the excess capacity is only being utilised rec , In addition environmenta~c o n c e ~ s are also a factor in the lack of investment in the tran§~issionsys i n v e s ~ e n in t transmission does not keep pace with the increasing de there will be ~ o n g - t eeconomic ~ and reliability problems. The when transmiss~oncapacity is needed the market &I ction delays etc. this could lead to periods of u ~ e l ~ a b i ~and ~ty i~ef~c~encies.

.5.3

Reliability

W h ~ it~ is e desirabIe to encourage co~npetitionin the e l e c ~ i c arke i ~ et to reduce the costs e quality for consumers, also v ~ ~ ai~ml yp o ~ tot ~ a i n ~the in . In an operational envir ent, an important re~~ability ~ e a § u r eis em security refers to system’s a b i l i ~to w ~ ~ h l is~ et l y~ ~ A system is said to be in a secure state if it is able to meet the Load d c y ~ as a line or without viol at^^ the erating constraints in case of a like c o n ~ n ~ e n such with re spec^ to a set of next In other words, s e c u is ~ defin ~ g ~ ~ e r aot uo ~ g e11 ies that are likely to occur. Gatas hic failures of power c ~ c a d e devents that are co~binati n a ~ r acl a l a ~ ~ t i(e. es q ~ i ~ m e ~n at l ~ n c t ~ odesign n s , flaws andor h ~ a en~ o r s[ security assessment is to reduce the likelihood of catastrophic failures. uch effort in the past decades has been devoted to the develop~entof c for s y s t e ~§ e ~ ~ r i assessment. ty These tools include state estima~ion, select~on contingency evaluation, external network equivalents and 10 ustry evolves into a competitive environ~ent,system securi ~ n c t i o n .In this new env~ronmen~, the p~~~ responsible or a similar entity. Since the e n v i r o n ~ ~isnm~ ical challenges. For example, the level of unce as increased s i ~ i ~ c ~ tThis l y .is due to the fact that ~ e n e r a ~ patterns on and the market outcome may not be easily p r e ~ ~ c ~ ~Cbo lne~.e q u e ~ ~al y , s y s ~ eng~neer e~ at the IS0 who studies system security may find it d ~ ~ ~to upredict l t the eneration and load conditions for evaluation of system security. is defined fo city market ity Gouncil nes ATG as triC

Competetive Wholesale Electricity

1

‘the Total Transfer Capability (TTC), less the Transmission ~eliabil~ty Margin ( T ~ less ~ , the sum of e ~ i s ~~ansmission in~ c o ~ ~ t m e n(which ts includes retail customer s e ~ ~ can e) 11061. Note that ATC is d e ~ n e dfor a ~ c t ~ t ~ o u presents the amount of power that can be Van ng the ~ a n s ~ i s s i osystem n con§traints~such as line flow limits. ing condit~onof a power system; the syste de thermal, voltage and stability limits. the effect o f various unce~intiesin system condi~ionson ATC, smission ~ a n s f e rcapability reserved by load-serving entities to ensure ~ h e i ~ erations from interco~ectionsto meet the system r e l ~ a b irequirernents. l~~ e x a ~ p of ~ ethe d e t e ~ i n a ~ based ~ o n on power flow c ~ ~ ~ ~ l can ati~i~s in Bergen and Vittal [ To determine the ATC for a path from X to Y, one can ~nject an a ~ o ~ofn power t at node X and remove the same amount of power at Y and calculate the i n j e c t e ~ r ~ ~ o amount v e d i s increased to a level that c the power flows. h its capacity, the amount can no longer be increase ~ansin~ssion line en wer ~ a n s ~ise rthe TTG. hen a iven line cont~ngencyis ~ a ~ in flows of the post-continge~cy operating conditions also nee transmission line conskaints. ~onsequeiit~y~ the ATC may not be as high the ~oiitingencyconditio~i s not considered. The s ~ e a d y - s ~power te flow m e t h o ~can be exte~dedto include s y s t e ~d y n a ~ ~ cTime s ~ domain simulatio~scan various levels of power ~ a n s f eto r evaluate system stability including vo s y ~ c h r o n i so~f the ~eneratorrotors. When dynamic security is considered in ad~itionto the steady-state operating cons~aints~ the resu~t~ng ATC may further be the availability of ancillary services such EIS reactive power sources can

e aware of the ~im~~ation§ of the path-based ATC concept [ 1081. The existence of the multiple transactions is a reality in the market environment. When the ATC of a path from X to Y is being evaluated, one needs to consider other t r a n s a c ~ ~ that ns have to be ~ccommodated. For the power flow method, other transactio wer inject^^ into and removed from other areas of the system. These taken into ~ccountsimul~~eously when the ATC for path X to U i le kansact~on~ a ~will e lead ~ to s differen~values of illustrates the concept of multi-dimension ower transfers over tie lines 1 , 2 and 3 re projection of the thr~e~~imensional region on the P,,-P, plane resern describe the secure power transfer point inside the ~~ree-dimens~ona~ re ot violate the security conska~nts. The projec~ionon other similar manner. Now suppose the power 1 P,, is at the value of P,,,, the maxi eases to the value o f PTIZ, Waisfer level for P,, increases to Pn2. NQWit is not difEcult to see th p o ~ e~r a n s ~level e r P,, increases fro^ the zero level. To s u ~ a r i s ethe , parameters of the operating con the other ~EIn~~ctions. , represented by a tie line in F on the levds of other ented by other tie lines.

Power System R

T3

Illustration ofa power system security region

~

$

~ andc

~

~

n

~

Comp~te~ive Wholesale Electricity Markets

3.5.4

TechveicalIssues

~ e g ~ d l e of s s wholesale electricity markets power system p l a ~ i n gand opera~~on has many technical challenges. With the advent of wholesale electricity markets new and d~fferenttechnical challenges may arise which need to be addressed. The comp~~ationa~ aspects of the electricity markets are one obvious area of interest [l09]. There are also interesting technical challenges related to the management of a large number of transactions [I 101. The OPF algorithm which i s at the heart of the marginal cost pricing paradigm [ZS] and of power system security analysis will have to meet ever-~cr~asing challenges [ 1 1 17. In the m ~ ~ m a l iIS0 s t model with ~elf~schedu~ing the UC a l g o r i t ~is being implicitly solved in a d i s ~ b u t e dm a ~ e by r the market particip~ts11121, which may or may not produce results which are as good as conventional UC algorithms. In the interest of efficiency these decentralised UC approaches need to be analysed. In the r n ~ ~I S ~0 ~ ~ model a cen~alisedU ~ / ~ ~ F - t yalgorithm pe is required [1131, Although s e c ~ t y constrained UC afgori~msexist [ 16,171 a UC algorithm with a full QPF formulation for a practical-size power system is still a significant computational challenge. The UC algorithm itself is still a very active research area with many issues unresolved [114,14]. In particular, solutions are invariably suboptimal and not robust [92]. In the short-term, regulators, system operators and market ~ a ~ c i p a nwill t s have to face the challenges described above. However, any actions need to allow market forces to push the indusfxy towards possible long-term competitive solutions.

.6 The authors would like to thank ESB National Grid, UCD President Res~archA w ~ d and s Fu~brightfor their financial support. This work is partially supported by US ~ a t ~ o n a l Science Foundation through Grant ECS-9612636 with matching funds fiom Alstom ESCA Corp. The authors would also like to thank Prof. Richard Christie, Universi~of Washington, and Mr John Kennedy, ES National Grid, for their useful c~mmentsand insights.

.7 [I] [2] [3] 141 [5]

S. Hunt and 6. Shuttleworth, CompefitioPtand choice in Electricity, John Wiley & Sons, 1996. EU, European Union Council Directive 96/92/EC, 1996. FERC, Regional Transmission Organisations, Docket No. R ~ 9 9 - 2 ~ 0ava~~able ~, at ~ . ~ e ~ , f2000. ~ . u s ~ D.W. Bunn, ‘Forecasting loads and prices in competitive power markets’, IEEE Proceedings, V01.88,pp. 163-169. T. Gjengedat, J.Q. Gjerde and R. Flolo, ‘Ancillary services in deregulated power systems: What are they: Who need them and who pays’, Proceediags of the Power System Technology Conference, 1, 1998, pp.704-709.

[6] H. Singh and A. Papalexopoulos, ‘Competitive procurement of ancillary services by an ~ n d ~ p e n a esysteni n~ operator’, ZEEE Transactions on Power Systems, Paper No. PE-427PWRS-0- 06- 1998,2000. ~ i sPower Systems, [7] T.W. Gedra, ‘On tra~missioncongestion and pricing’, IEEE ~ r a n s a c ~ i oon Vol. 14, 1999, pp.241-248. 181 L. Hyman, ‘‘Transmission, congestion, pricing, and incentives’, ZEEE Power ~ n g i n e e ~ ~ n g Revinu, August 1999, pp.4-10. 191 R.D. Christie, B.F. Wollenberg and I. Wangensteen, ‘Transmission ma~agemen~ in the deregulated environment’, IEEE Proceedings, Vo1.88,2000, pp.170-195. I101 R.D. Christie and I. Wangensteen, ‘The energy market in Noway and we den: ~ n ~ o d u c ~ i o n ’ ~ IEEE Power Engineering Review, February 1998, pp.4445. [ 1 13 G.Strbac, E.D. Farmer and B.J. Cory, ‘Framework for the incorporation of demand-side in a s ~ o~n ~ s ~ r i b ~ ~ competitive electricity markt’, ZEE Proceedings Generation, ~ r a n s t n i ~ ~and VOI.143, 1996, pp.232-237. [12] A.J. Wood and B.F. Wollenberg, Power Generation, Operation and Control, Wiley Interscience, New York, 1996. [131 R. ~ i l ~ and ~ ~ R.N. o Allen, n Reliability Evalua~ionof Power Systems, Lo 1996. [ 141 K. Baldick, ‘The generalized unit commitment problem’, IEEE Transactions on Power Systems, Vol. 10, 1995, pp.465-475. [I51 M. Walsh and M.J. Q’Malley, ‘Augmented Hopfield network for constrained generator scheduling’, IEEE Tramactions on Power Systems, Vol.14, 1999, pp.765 - 771. [I61 J. Batut and A. Renaud, ‘Daily generation scheduling o p t ~ i ~ a t i owith n constraints: a new class of algoriths’, IEEE Tramactions on Power Systems, Vo1.7, 1992, pp.982-989. [I71 JJ, Shaw, ‘A direct method for security constrained unit c o ~ ~ e n tIEEE ’ , Trans~cfionon Power System, Vol.lO, 1995, pp.1329-1340. [ 181 L.F. Escudero, J. Salmeron, I. Paradinas and M.Sanchez, ‘SEGEM. A simulation approach for electric generation management’, IEEE Transactionson Power Systems, Vol. 13, 1998, pp.738747. [191 R.J. Green, ‘Competition in generation: the economic foundation’, IEEE Proceedings, Vo1.88, 2000, pp.128-139. [20] R.D.Christie and I, ~ a n g ~ n s ~ e‘The e n , energy market in Norway &nd Sweden: Congestion ~ n a g e m e n tIEEE ~ , Power Engineering Review, May 1998, ~ p . 6 1 - ~ 3 . [21] R.D. Christie and I, Wangensteen, ‘The energy market in Horny and Sweden: The spot and htures markets’, IEEE Power Engineering Review, March 1998, pp.55-54. 1225 J. ~ c o n a ‘A , bid solicitation and selection method for developing a eompetit~v~ spot priced electric market’, IEEE Transactionson Power Systems, Vol. 12, 1997, pp.743-748. 1231 N.S. Rau, ‘The need for capacity markets in the ~ e r e g u l aelectrical ~ ~ ~ industry - a review’, ~ r o c e @ d ~ofn the ~ s IEEE Power Engineering Society, ~ ~ ~~ te ee r~New ~ nYork, ~ , 1, 199% I

1241 J.M. Jacobs, ‘Artificial power markets and unintended consequerices’, I E . E fiansacfions POWWS y s t ~V01.12, ~, 1997, pp.968-972. [25] ET. Singh, S. Nao and A. Pa~alexopoulos~“Transmiss~o~ congestion managei~ent in competitive electricity markets’, IEEE Transaclions an Power System, Vol. 13, 1998, pp.672680,

Competetive Wholesale Electricity Markets

105

[26] C. Taylor, Power System Yoltage Stabifity, McGraw Will, 1994. 1271 K. Seeley, J. Lawanee, J, and C.-6. Liu, ‘Analysis of electricity market rules and their effect on strategic behavior in a non congested grid‘, IEEE Transactions on Power Systems, V01.15, 2000, pp.157-$62. [28] F.C. Schweppe, M.C. Caraminis, R.D. Tabors and R.E. Bahn, Spot Pricing of Electricity, Kluwer ~ c a d e m i Publishers, c 1998. G. Strbac, D. Kirschen and S. Ahmed, ‘Allocating transmission usage on the basis o f traceable [29] contributions of generators and loads to flows’, IEEE Transaclions on Power Systems, Vol. 13, 1998, pp.527-533. [30] J.W. Marangon Lima and E.J. de Oliveiw, ‘The long-term impact of transmission pricing’, LEEE Transactions on Power Systems, Vol. 13, 1998, pp. 1514-1520. [31] N.S. Rau, ‘Optimal system dispatch o f a system based on offers and bids - a mixed integer LP formulation’, IEEE Transaciions on Power Syslenns, Vol. 14, 1999, pp.274-279. [32] B.H. BaWtcn, Technical and Economic Aspects of Operation of Thermal and Hydro Power Systems, P h n thesis, Norwegian ~niversityof Science and T e c h n o ~ o199’7. ~, [33] M. Flynn, NI.. Walsh and M.J. O’Malley, ‘Efficient use of generator resources in emerging electricity markets’, IEEE Transactions on Power Systems, Vol. 15, 2000, pp.24 1-249. E341 J. Kumar and G.B. Shebl6, Tramework for energy brokerage system with reserve margin and transmission losses’, IEEE Transactions on Power Systems, Vol.11, 1996, pp. 1763-1769. [3S] N.S. Rau, ‘Assignment capability obligation to entities in competitive markets -the concept of reliability equity’, IEEE Transuctions on Power Systems, Vol. 14, 1999, pp.8~4-889. J. Qdom and G. [36] R. Billinton, L. Salvaderi, J.D. McCallcy, H. Chao, T. Seitz, R.N. Fallon, ‘Reliability issues in today’s power utility environment’, Systems, Vo1.12, 1997, pp.1706-1714. [37] C Concordia, L.H. Fink and G . Pouiilkas, ‘Load shedding on an isolated System’, IEEE T r a n s f f c ~ ~Power o n ~ Systems, Voi.10, 1995, pp.1467-1472. [38] J.W. O ’ ~ u 1 ~ ~ vand a n .J. alley^ ‘Economic dispatch of a small utility with a ~ ~ q u e n c y based reserve policy’, IEEE Transactions on Power Systems, VoI.11, 1996, pp.1648-1653. [39] G.B. ShebltS, ‘Priced based operation in an auction market structure’, IEEE ~ r a n s ~ c ~ ~ ooyy11 s Power Systems, Vol.11, 1996, pp.1770-1777. I401 H.A. Cooi, D.P. Mendes, K.R.W. Bell, and D.S. Kirschen, ‘Optimal scheduling of spinning reserve’, IEEE Transactions on Power Systems, Vo1.14, 1999, pp. 1485-1492. lynn, P. S h e r ~ ~ J.n ,Dillon and M.J. O’Malley, ‘Reliability and reserve in compe electricity market scheduling’, IEEE Trcansactionson Power Systems, Tdol. 16,2001, pp.78 [42] M.J. Sullivan, B.N. Suddeth, T. Vardeli and A. Vojdani, ‘Intemption costs, us tom er satisfaction and expectations for service reliability’, IEEE ~rafls~ctions on Power ~ y s ~ e m s , Vol.1 I, 1996, pp.989-995. 1431 J.W. O*Sullivan and M.J. O ’ ~ d i ~ e y‘A, new methodolagy for the provision of reserve in an isolated power system’, XEEE Transactions on Power Systems, Vol.14, 1999, p p , ~ 7 4 - ~ ~ 3 . [44] T.W. Gedra and P. Varaiya, ‘LiWarkets and pricing far interruptible electric power’, XEEE Transacti[)nson Power Systems, Vo1.8, 1993, pp.122-128. [45] N. Rudnick, K. Varela and W. Hogan, ‘Evaluation of alternatives for power sys~em coordination and pooling in a competitive environment’, IEEE Transactions on Power Sysfems, vO1.12, 1997, pp.605-613.

10

Power System R e s ~ c ~ r i and n g Dere R.P. McAfee and J, ~ c ~ i l l a ‘Auctions n, and bidding’, Journal of Economic Li~erature,June

1987, pp.699-738. R.D. Tabors, ‘Lessons form the UK and Norway’, IEEE Spectrum, August 1996, pp.45-49. G. Gross, D. Finlay 6. and Deltas, ‘Strategic bidding in electricity generation supply markets’, Proceedings ofthe lEEE Power Engineering Society, Winter Meeting, New York, 1, 1999, pp.309-3 15. W. Vickrey, ‘Counterspeculation, auctions, and competitive sealed ~e~ders’, J o of ~ Finance, March 1961, pp.8-37. H.R. Varian, ~ n t e ~ e d i aMicroeconornics: te A mo~ernapproach, 5th Edition, Norton, New York, 1999. OFFER (2000) Office of Electricity Regulation, United Kingdom. www.ofgas.g0V.uk. . Bjorgan, C.-C. Liu and J. Lawarrbe, ‘Financial risk managem~ntin a competitive electricity market’, IEEE Transactions on Power Systems, Vo1.14, 1999, pp.p.1285-1292. .J. Kaye, H.R. Outhred and C.B. Bannister, ‘Forward contracts for the operation of an ectricity industry under spot pricing’, IEEE Transactions Power System, Vo1.5, 1990, pp.4652. CalPX, California Power Exchange, www.calpx.com, 2000. CAISO, California Independent System Operator, m.caiso,com, 2000. Power Pool of Alberta, ~ . p o w e ~ o o i . a b . c2000. a, P.R. Gribik, G.A. Angelidis and R.R. Kovacs, ‘Trmsmission access and pricing with multiple separate energy forward markets’, IEEE Transuc~ionsPower systems, Vol.14, 1999, pp. 865876. J.J. Gonzalez and P. Basagotti, ‘Spanish power exchange market and information system design concepts, and operating experience’, ProceediBgs of the 21” Power dndustry Computer Applica~ionsConference, 1999, pp.245 -252. X. Ma, D. Sun and K. Cheung, ‘Energy and reserve dispatch in a multi-zone electricity market’, IEEE Transactions on Power Systems, VoLl4, 1999, pp.913-919. Pennsylvania, New-Jersey and Maryland Interconnection, ~ . p j m . c o m2000. , hey, D. Goodwin, X. Ma, D. Streiffert, and D. Sun,‘A security-constrained bid-clearing system for the New Zealand wholesale electricity market’, IEEE Transaciions 08 Power Systems, Vol.13, 1998, pp.340-346. K.W. Cheung, P. Shamsollahi, D. SUR,J. Milligan and M. Potishnretc, ‘Energy and ancillary service dispatch for the interim IS0 New England electricity market’, Proceedings of the 2Ist Power Industry Computer Applications Conference, 1999, pp.47-53. W.W. Hogan, ‘Contract networks for electric power transmission’, Journal of Regufoto9 mics, Vo1.4, 1992, pp.211-242. Order accepting for filing in part and rejecting in part proposed tariff amendment and on revaluation o f approach to addressing intrazonal congestion, California ISO, Docket No. ER00-555-00, available at www.ferc.fed.us, 2000. ME. Baran, V. Banunarayanan and K.E.A. Garren, ‘Transaction assessment method for allocation o f transmission services’, [E!?.. Tramactions on Power Systems, Vol. 14, 1999, pp.920-928. C. Li, A.J. Svoboda, X. Guan H. and Singh, ‘Revenue adequate bidding strategies in competitive electricity markets’, IEEE Transactions on Power Systems, Vol. 14, 1999, 497.

Competetive Wholesale Electricity Markets -

7

[67] TEEE Current Operating Problems subcommittee, ‘Operational aspects of generation cycling’, IEEE T r u y ~ a c t j Power o ~ Systems, V01.5, 1990, pp.1194-1203. E681 I. Q~~ro-Novas, C. M e s e ~ e r C. , Batlle and J.J. Alba, ‘A simulation model for a competitive generation market’, IEEE Transactions on Power Systems, Vol.15, 2000, p ~ . 2 ~ 0 - ~ 5 4 ~ ria and T.S. Dillon, ‘Electricity price s h o ~ - g forecastin e~ Szkuta, L A ,IEEE Trunsuctions on Power Systems, V01.14, 1999, pp.851 cial neural ne Fosso, A. Gjelsvik, A. Haugstad, R. MO and I. Wangenstcen, ‘Generation sched~lingin a d e r e ~ l ~ t esystem: d The N o ~ e g i a ncase’, IEEE Transactions on Power Systems, v01.14, 1999, pp.75-81. [71] J.W.M. Cheng, F. , Galiana and D.T. McGillis, ‘Studies of bilateral co~tractswith respect to steady state security in a deregulated e n v ~ r o ~ e n tIEEE ’ , Trunsuc~ionsPower Systems, Vol. 13, 1998, pp.1020-1025. [72] Z, Yu, ‘Mixed integer sociat welfare maximization (MI-SWM) and i~plicationsin opt~mal electricity pricing’, IEEE Power Engineering Review, July 1999, pp.53-54. [73] ET. Singh, IEEE tutorial on game theory upplications in electric power markets, IEEE Catdog NO. ~ 9 ~ F l 3 4 -1999. 0, [74] J.W. Lamont and S. Rajan, ‘Strategic bidding in an energy brokerage’, IEEE T r u n s ~ ~ ~ i on ons Power Systems, Vol.12, 1997, pp.1729-1733. [75] P. She~idan, M. Flynn and M.J. O’Malley, ‘Generator scheduiin~with demand bids’, Procee~~ngs of the IEEE Power Engineering Society, Summer ~ e e t i n gSeattle, , Paper No. 2 ~ O S M - 0 1 22000. ~ A. Lusan, Z.Yu and F.T. Sparrow, ‘Market gaming and market power gatio ion in regulated electricity markets’, Proceedings of the IEEE Power Engineering Society, Winter Meeting, New York, 2, 1999, pp.839-843. [77] Z. Younnes and M. Ilic, ‘Generation strategies for gaming transmission constraints: d e r e ~ ~ a t eelectric d power market be an oligopoly’, Proceedings of the 31’‘ ~ n n ~ tional Conference on System Sciences, 3, 1998, pp.112-121. and A. Papalexopoi~los,‘Reactive power pricing and management’, IEEE Tr~nsac~ions Systems, V01.12, 1997, pp.95-104. Nordhaus and C. Shapiro, Preliminary report on the operation of ancillary eh o f the California Independent System Operator, ~ a l i f o ~ ~ndependen~ ia System Qpe~ator.~ . e a i s o . c o m 1998. ) t paths, phase shifters, and efficient electricity trade’, IEEE 01.12, 1997, pp.749-755. . Pang, ‘Strategic gaming analysis for electric power Transactions on Power Systems, Vo1.15, 2000, pp.638-

~

I821 C.W. Richter and G.B. Sheblb, ‘Genetic algorithm cvolution of utility bidding s~ategiesfor the competitive marketplace’, IEEK Transactions on Power Systems, Vol. 13, 1998, pp.256-261. E831 M. Song, C.-6. Liu, I. Lawarrbe and R. Dahlgren, ’Optimal electricity supply b ~ ~ d iby ii~ M~rko*decision process’, IEEE Transactionson Power Systems, Vol. IS, 2000, pp.618-624. Gibescu and C.-C. Liu, ‘Optimization o f ancillary services for system security’, roceeditzgs Bulk Power System Dynamics and Control W - R e s ~ ~ ~ c t u r i~n gy ,m p o s ~1998, ~~s, pp.35 1-358. N.Allen and D. Ilk, ‘Reserve markets for power systems reliability’, IEEE ~ ~ u n s a c on ~~o~s ower Systems, Vo1.15,2000, pp.228-233. ~

x

Power System R e s t ~ and c Dere ~ ~

.R. Bobo, D.M. Mauzy and P.J. Trefny, ‘Economic generation dispatch with responsive

reserve constraints’, IEEE Transactions on Power Systems, Vo1.9, 1994, pp.555-559. agic, K.H.Abdul-Rahman and J.G. Waight, ‘Spot pricing of capacities for and transmission of reserve in an extended poolco model’, IEEE Transactions Systems, Vo1.13, 1998, pp.1128-1135. [88] J.B. Cardell, C.C. Hilt and W.W. Hogan, ‘Market power and strategic intcraction in electricity networks’, Resource and Energy Economics, Vol.19, 1997, ppiO9-134. [89] T.W. Gedra, ‘Optional forward contracts for electric power niafkets’, IfiEE Transactions on Power Systems, Vo1.9, 1994, pp.1766-1773. organ, R., H. Song, (2.4. Liu and R. Dah flexible electricity con~acts’, 1901 IEEE %amactiom on Power Systems, Vol. 15,20 f9I] R.J. Thornas and T.R. Schneider, ‘Underlying technical issues in electricity ~ e r e ~ ~ a ~ ~ o n ’ Proceedings of the 30th annual Hawaii ~ n t e r n a t i ~ nConference a~ on System Sciences, 5, 1997, pp.561-570. 1921 $3.Oren, A.J. Svoboda and R.B. Johnson, ‘Volatility of unit commitment in competitive electricity markets’, Proceedings of the 30th Annual Hawaii I n t ~ r n a ~ ~Conference o ~ a ~ on ~ ~ Sciences, s 5, ~ 1997, e pp.594~60~. ~ (931 J. K m a r and G.B. Shebli-5, ‘Auction market simulator for priced based operation’. 6EEE Transactions on Rower Systems, Vol. 13, 1998, pp.250-255. [94] K.J. Green and D.M. Newbery, ‘Competitionin the British electricity spot market’, Journal of Political Economy, 100, 1992, pp.929-953. [SS] S.Borenstein, J. Bushnell and F. Wolak, Diagnosing market power in California’s wholesale electricity ~ ~University k of~California ~ Energy , I n s t ~ ~ t1999, c, Wolfim, ‘Strategic bidding in a multiunit auction: AD empirical analysis o f bids to supply electricity in England and Wales’, RandJournaE of Ec’conollzics,29, 1998, pp.903-725. [97] K. Miyaknwa, Analysis o f strategic behaviors in an open access transrrrission grid. MSc Thesis, University of Washington, 2000. ration’, ~ ~ ~ c e e $ oif~the g s IEEE ~ ~ w e r 1, 1999, pp.568-572. ‘Advanced pricing in electrical systems. 1. 1.12, 1997, pp.48~-49~. .N. Siddiqi and J.W. Zarnikau, ‘Advanced pricing in e k t r i c d systems. 11. nscactions on Power Systems, Vol. 12, 199’7,pp.496-502. ughman, ‘The economic efficiency impacts o f alternatives for revenue [kola B.H. W m and nsactions on Power Systems, Vol.l2,1997, pp.1129-1135. [1025 1.1. Perez-Arriaga, F.J. Rubio, J.F. Yuerta, J. Arceluz, and J. Marin, ‘Marpi ~raiis~ission services: An analysis of cost recovery’, IEEE Tramactions on RQWeV Syslems, V01.10, 1995, pp.546-553. [I031 DOE, Maintaining reliability in a competitive U.S.eiectriciQ industry, Final report of the task force an clectric system reliability, United States Department of Energy, 1998. [I041 C.-C. Liu and R. C istie, Energy Management Systsni, in Encyclopaedia of Computer Science A. Kent, J.G. WiBiams, C.M. Hall and R. ~e~~~~26, ~ ~ ~ 11, p l I

report for the power system outage that occurred on the Mrestcm Interconnection on August loth, 1996 at 1548 PAST, Western Systm Coordinating Council, ~ ~ w s ~ c1996. . c o ~ ~

Competetive Wholesale Electricity Markets ~ Available ~ C Transfer , Capability Definitions and Determination, North American Ekctsk ~ e l ~ a b i lCouncil, ~ty 1996. [ 1071 A.R. ~ ~ r g and e n V. Vittai, Power System Analysis, 2nd Edition, senti ice Hall, 2 0 ~ pp.4360~ 440. . Ilic, T.Y. Yong and A. Zobian, ‘Available transfer capacity (ATC) and its value under open access’, IEEE Transactions on Power Systems, Vol. 12, 1997, pp.636-645. [ 1091 C.B. SheblC, ComputationalAuction Mechanismsfor Restructured Power Indusrry Operation, ~ l u w eAcademic s ~ b l ~ s h1999. e~~, [1101 G.W. ~ o s ~ n w and a l ~C.-C. Liu, ‘Consistency evaluation in an o p ~ r a ~ ~ o~nnav~~ r o ~ $ n t involving many transactions’, IEEE Transactions on Power Systems, v01.11, 1996, pp.17571762. [111] J.A. Momoh, R.Y. Kosselar, N.S. Bond, B. Stott, D. Sun, Papalexpoulos and P. ‘Challenges to optimal power flow’, IEEE Transactions on Power S’lstems, V01.12, 1997,

[1061 ~

. Angelidis, H.Singh, and A. Papalexopoulos, ‘Consu~$rp a y ~ e nm~i ~ m ~ ~ t i o n in power pool auctions’, IEEE Transactions on Power Systems, Vo1.13, 1998, pp.886-991. [ 1131 D. C~a~opadhyay, ‘Daily generation scheduling: Quest for new models’, JEEE T r u ~ s a c t i o ~ on Power Systems, Vol. 13, 1998, pp.624-629. Walsh aind M,J. O’Malley, ‘Augmented Hopfield network for unit commitment and economic ~ i s p aIEEE ~ ~ Transucfi~ns , on Power Systems, V01.12,~1997, pp. 1 7 6 5 - ~ 7 7 ~ .

Cliff Walton London Electricity Group UK

Robert Frief London Electricity Group UK

Dr Loi Lei Lai City University, London

UK

Transmission and d i s ~ i b ~ ~ are ~ i ostill n regarded as the natural monopoly elements in the r e s t ~ c ~ r UK e d energy market. Since privatisation in 1990 there have been a number of changes in the structure o f the industry which have impacted on the dis~ibutionbusinesses and the i ~ ~ t r o d u c of t ~ othe ~ Utilities Bill heralds a further change in the relationship with g o v ~ ~ ~ ethen regulatory t, body and consumers. One of the main objectives of privatisatio~was to promote compe~ition. This has focused on the supply (Le. the retailing) of electricity and gas and has encompassed the associated aspects of metering. A c o ~ p e ~ t i framework ve was developed for new connections to distri~utionnetworks defining certain elements as contestable work, but to date this area has not seen the widescale competitive activity expected and the regulator has indicated his in~~ntion to review competition in the gas and electricity connection markets by March 2001. The biggest effects on the distribution businesses have resulted from the price control mechanism. Distribution businesses in the UK are price regulated, a part of which is to allow a return on the assets purchased at vesting and the i~vestmen~s made in the subsequent years. The latest price review, which came into effect in April 2000, saw the regulator propose reductions in distribution business income similar to those following the last review, None of the UK ~ i s ~ ~ b u tcom~anies ion have c ~ d i ~ e the ~ ~o~ue~dc ~ of r nthe ~ latest review, which implies that the companies believe that they can achieve these savings. The only st~cturalchange at the time of writing has been the announce~entby London

Distribution in a Deregulated Market

Electricity and Eastern Electricity of a joint venture to operate their n e ~ o r k s with , the asset ownership remaining with the parent companies, Since the latest price control the UK r e ~ ~ a thas o r i ~ ~ l e m e n t ae dpr the ~ ~ f o ~ aprav~ded t i a ~by the regulated di§~ibutioncompanies and incentives to introduce an element of competition. The details of any such scheme are still to be decided, but it sends a clear signal that after 10 years there remains much scope for a ~ oand~the electricity distribution industry. menls in ~ e ~ lpractice

4.2. I

~ o ~ ~ e t iin~ Supply ion

The development of competition in the energy supply market in the develop~nentof two distinct activities in the UK public electricity suppl t y gas and distribut~on. The supply businesses are responsible for the sale of e ~ e c ~ c iand whilst the d~stribMtionbusinesses manage the cables, lines, ~ a n s f o ~ eand r s switchgear which form the power supply networks between the EHV grid system and the end users. The development o f competition in supply as part of the process of d subject in its own right. Certain aspects of the process to deliver a CO market have had a significant impact on the PESs’ distribution businesses. The most visible aspect of this has been the moves both physically and ~ ~ ~ c itoa l ~ y separate the retail and ~ i s ~ i b u t i obusiness, n and in some cases the sale of retail businesses to third parties. This process has involved the rebranding of the separate bus~esses. It was intended that the PES distribution businesses be rebranded and this will case, although those generators who have acquired retail interests (National and PowerGen) have essentially rebranded their retail arms (Mid~andsand respectively).

4.2.2

The ~ ~ s p 5 n s i b i l ioft ~~~es~ aand i l ~ist~ibution

retail business bulk hases both electricity and gas and supplies them to their custo~ersover the electricity and gas networks. A great deal of work has been required to the necessary systems to effect a competitive market. of this ~ a r k e ~ l a the c e retail businesses have taken responsib~li~ for meter g providers r e a d ~ gwith , the intent that this service be procured from ~ ~ t e r - r e a d i nservice on a compe~itivebasis. The development of competition in the mete~ngsector will even~allysee the pr~visionof meters to suppliers by meter asset ana age^ and opera to^. The distrib~tioncompanies manage and maintain the electricity d~stributio~ network. This involves both the technical asset manage~entand planning services and the b ~ ~ofl use ~ g of system charges from suppliers and the management and maintenance of the existing meter assets.

12

Power System Restructuring and

eration services organ~sationin London has been estabiishe from the n ~ ~ o asset r k manager, with an a ed scope and level o f s ~ ~b e~ ~ ce een

ive for separation o f businesses is a re¶ui~emento f establishing a ~ o r n p ~ ~ i ~ ~ v in the supply of energy [I]. In essence the incumben~d i s ~ i b u ~may or the position of the domina~thost supplier to the d e ~ r n e of ~ t custom~r§ Five potential means of achieving this have been i d e n t i ~ ~ ~ : comp~~itors. a combined dis~bution could disadvantage comp charges to support the suppliers’ retail tariffs. s pote~tiallyhas access to i n f o ~ a t i o nthat other ~upp~iers will not le, the names o f customers supp~~ed by a second-tier supplier. intentions o f a dis~ibutionbusiness. For e x ~ p l e ad , ze and nature of changes to use of system 4. Cross-subsidisa~ionby a d i s ~ r o p o ~ o n aallocation te of costs overheads to the d i ~ ~ i b u ~business. ion A small reallocat~onwill have i ~ p a in c ~a retail i n d u s ~that has very small ~ a r g i n s . The regul this issue in the 1999 price control review by realloc tion to supply before assessing the relative ef~ciencies businesses. ution business will in some way ~ o ~ g r the ~ service d e to a custo s ~ ~ p ~For ~ example, er. the response to power o u ~ ~ ~ ~ .

ion businesses ~ e s u l t i nfrom ~ c ~ m ~ e ~ i t in i o ns s e p ~ a t i of ~ nthe two business areas are:

on businesses of the i ~ ~ l e m e n ~oft the io~ e§tablished at the time e in resolving system-b constraining s u p ~ ~ ~ esales r s ’ for e ~ b e d d ee~n e r a t ~energy ~ ~ sales.

istribution in a Deregulated Market

.

It was decided that t ~ e was r ~

required s e ~ ~ e ~ a t i o n

e a p e r ~ Qs~~a~~v ito c ei ~ ~ i ~ i cd ~u sat ~o ~ e ~ s . torner^ er^ still receive the level of service the

i s conkoiled via

on ~alf-hou~ly c Q ~ s u ~ ~ tdata) i o n for 1 e s t i ~ a t ~for s ) smaller (mainly quarterly st o f a fixed portion or stan

Power System ~

11

4.2.7

e

s

~andc Der~gu~atiQn ~ ~ n ~

C ~ t oService ~ e ~

The ma~iagementof customer relations is another area where competition in supply cre~tes a number o f options. In New Zealand, the initial approach routed a11 customer contact throu businesses. This simplifies the contact issues for the customer, manage~entof the interface between suppliers and dis~butionb the correct i n f o ~ a t i o nis available to inform the customer. the UK, the distribution businesses have kept an interface with c u s t o ~ in e ~relation ly outages. Hence customers have two points of contact. ~ l t ~ o this u gs ~i m ~ ~ i ~ e s the m ~ a g e m e nof~ information flows on outages between istribution businesses and suppliers (there is none) the management of the routing of calls to the wro care. The future solution to some of these issues i s already apparent in call systems. These are already being installed to provide i n f o ~ a ~ i oonn outages and are of particular use in the extreme circumstances o f wide-scale power outages when call centres become overwhelnied. Internet technology will soon provide accurate supply of i n f o ~ a t i o non outages and torat~ontimes to both customers and suppliers - the inte~ationof fault reporting lephone network. It is possibie to generate specific y~ice-ac~~vated messages g to postcodes or dialling code i n f ~ ~ a t i o nIt. may even go as f8r as pro ~ f o ~ i a t i o(i.e. n ring the customer). London ~ l e c ~ hc i ~ d e ~ i c~~ ~e gnpower t outages on its internal web site for so make this facility available via the Internet once suitable security safeguards h proven. 4.2.8

~ a ~ ~ ~ int Metering i t i o ~

and 100 kW markets since I991 and 1994 Competit~on has existed in the 1 to own and operate meter assets in both these respectively. Third parties have bee areas since 1995. In conjunction with the hi1 c o ~ ~ e t i t i oinn energy physi~aland financial separation of the dis~ibutionand supply b~sinesses~ competitive metering are being extended to the remainin market sectors. This has created two distinct business streams, meter rea asset m ~ a g e m e n (otherwise t known as meter o provide data retrieval and data processing @R s will provide meter asset provision and m ~ a g e m e n(t k n o as ~ dis~ibutionbusinesses no longer have any part in the reading of meters, which is by the energy retail business via a contractual ~ a n g e m e nwith t a meter reader. ssion in that meter readers could provide services for r e a ~ ~ nany g This is a n a ~ apro l utility service meter. o n g - t e ~developments are likely to result in meters that are read remotely. The existing meter assets are presently owned by the distribut energy supplier, via a meter asset man~gement~ompany,can provide new cu$tomers DT as r lacements for existing meters. In the me will develop in the p~ov~sion of meters and the ~anagemento f these assets,

Distribution in a ~ ~ r ~ ~ u Market lat~d

1

To ensure the availability of these services, the PES energy retail businesses will ovide a meter-~eadingservice of last resort and the erat~onsservice of last resort.

d-side ~ ~ n a g e m e n ~ s to reduce the peak demand either of more efficient usage or by mov elements of the 10 the system load factor. real alternatives to rcernent. In general terms these the ~ansmissionand generation level where the cumulative effec ema and-side ma~agementhas been encouraged by the use of tariffs, e. peak tariffs for storage heating. In this i n s ~ c ethe move away . heat~ngto off-peak storage w impact on both generation and &arm wever, at a distribution ievel the wi e installed c a p a c i ~require^. mand occurring at nig ating has resulted in pe ated ~ a r ~ the e t promot~on of energy effi ess clear. The generation capacity availabl~b cts exist between energy §uppliers or retailers or respons~blefor bala~cingthe system’s m a ~ ~ system ~ ~ voltage, n i ~~ e~q u e n and c ~ security. Large c ing themselves available for disconnection as istrative complexities of participating in such this to a few very large bu§ine$ses. recourse to modifying use of system tariffs to promote alternati ,the energy § u ~ ~ l imay e r not be obliged to any re§ulting ~ n c e n t i ~Addit~onall~ e~. in the stomer contact and knowledge of their partic erator to m ~ a g the e syste bedded generation to offset the need for r e i n ~ o r c e m ehas ~~ become a major subject of debate. The principal difficulty is that p~omo~ing and ennsurhg the ~n§ta~lation of suitable generation in advance of the reinforcement ~e~uirement is far from a simple task. Not only this but the ~ p p r o p r i acommercial ~ a ~ a n g e must ~ ~ be ~ in ~ s place for the risk to be m i n i ~ ~ ~ss~e fd~ c i e nfor ~ l ythe generation to repre§~n~ an ~ q u ~ ~ a b l e ~ ~ ~ e ~ to a tt~aditiona~ ive ~~~nforceme~t. Therefore it can be seen that in the marketplace active d e ~ ~ a n d - s ~ ~ e re~uirea time of to be sent to c u s t o m e ~via tariff a ~enerat~on, ~an§mi lion levels designed to reflect the local cmt this would be in reducing the need ive if system security is allow for the effects of would require careful risk assessment. Variances between actual would need to be reflected in v a r ~ a ~ use l e of system charges in

1

d. It is not the intent to discuss here whether variable use of system a: on to ~ u s t o thro ~ ~ r ~ ticable as these woul

t in asset replace~enr,s

~ rein for^ § ~

~

~

~ i ~ ~ b u tini oa ~n ~ r ~ g u l aMarket ted -

117

Voltage tolerances, faul r e ~ a i nas they were before the priv~tisationof th r changes to h a ~ o n i s ewithin the E ~ o p e a nUnion). reflect EM1 standards and This has not been due to the ~ e s ~ c of~the ~ indu n g e unwiIlingness of What has chang~dth~oughr~gulat~on i s quality o f supply a number and d ~ r a t i oof~i n ~ e ~ ~ t iand o n customer s service. section. ~~ s u ~ ~drivers l y is discus in the f o ~ l o w on covers the drivers affecting planning ncepts of p l a ~ i n gasset r e p l a ~ e ~ e n t

4.3.3

~~~~~~~~

y ~ i s ~ i b u tsystem i o ~ design muse meet the following r e ~ ~ ~ r e m e n ~ ~ :

e able to supply the system demand whilst meeting the

s~stemsthe n e ~ o r must ~ s also:

ng costs of the network.

mic and risk assessment. Whilst much h

ndard was written in the 1970s, the st ai~houghit only addresses the scaIe and duration of a loss of supply and not the ncy of which such incidents can be e~pectedto occur. rmance, which are beginning to drive network the s ~ a n dpa~icu~arly ~~, in ~espectof the fre e d i f ~ c u here l ~ has been to find a common ance levels that from the different e n. The most recen er for change has come from to make the best use of embedded generation connected to the d i s ~ i b u t i onetwor~s ~ at 132 elow. This generation has not been considered in system security to date. This is due to:

1

s

Of tion of most gener~tion. 1 PPr commercial framework. The ~ i ~ ~ c u Ioft ~~e ss u ~ the n gpresence of the necessary generation before systems need ~ e i ~ o r c i n g .

be solved at the time of w ~ t i n g 1 be a ~ ~ e n d with e d a means of asse

has been measure A ~ a i l a b i l i ~S :y s t e ~average i n ~ e ~ p t i oduration n

r some time by the f o l l ~ w i ~ ) or customer minu~es

stem average i n t ~ ~ p t ~ eoqnu e ~ cindex y ~S~~~~ or its e ~ u ~ v ainl e ~ ~ r 100 connected custo~ers. These measures represent the average performance of the system and so do not a ~ c u r ~ ~ ~ l ind~~idual c u s t o ~ e may r e x ~ e ~ e n [2,3]. ce In the Of than I minu~eis cou~tedtowards these statistic 3 rm to allow for the benefits of system automation. dica~esthat customers prefer not to be ~ n t e ~ p tbut ~ din, the event of an d of restorat~onand accurate i n f o ~ a t i o n kely outage times ~ e c o m e The provision of such i ~ ~ o is ~po a ~ both ~ ~through n call centre

Ristribution in a Deregulated Market

11

re 4.1 demonstrates the relationship between incidents, ~ t e ~ p t i o and n s customer minutes lost. This cl shows that whilst by far the most inciden~soccw at low vol~age, ) 11 kV and 6.6 kV systems cause by far the most c u s t o ~ e r the medium-vol~ge c eere are siest area in which to improve network e r f o ~ a ~as dis~ption.It is also cos~~effec~ive solutio~savailable. It comes as no surprise eref fore that i n v e s ~ e n thas centred on reducing the impact of MV system incidents and disruption. The most ~ o u b ~ e s impact o ~ e on customers is that r~sultingfrom ~ e ~ u eor n t~ u ~ t i p l e The p ~ m measures a ~ discussed above have there for^ recently been tl measure of multiple interruptions, which will determine the ~ e r c c ~ ~ a ~ rs who experience more than a given number of i n t e ~ p t ~ oper n s amum, the second post-pr~vatisationprice control the d i s ~ b u ~ companies ~on s eved by the end of the second review period. During the third review place in 199842, the e l e c ~ cindustry i~ regulator decided to set targets for the e l ~ c t ~ c i t y c o m p ~ i erather s than to allow the co~paniesto set their own. In addition to this a syst of incentives based around p e r f o ~ a n c eagainst some of these measures is to i ~ p ~ e ~ e in n ~ e d [4]. Much discussion remained at the time of writing as to the e ~ ~ h a stoi sbe to which measures and therefore it is dif~cultto resolve haw these measures will further aEect the development o f distribution p l a n n ~ g .The de networks to meet quality of supply requirements targets the reduction of the i network failures at two main areas: The p r e v ~ t of ~ o~ ~n t e ~ ~ t ~ o n ~ The restoration of supplies has increased the focus an imp Since deregulation, the U rent approaches have bee aspect of c~stomerservi different companies driven by their particular regional and network pro s i ~ i ~ c aareas n t of ~ v e s ~ e have n t been in insulated or semi4 auto reclosers on overhead lines and network remote control and § Insulated overhead conductors have been used to reduc ~ n t ~ ~ p t i due o n sto trees tou~hinglines, which can lead to mor ~o~e-mounted auto reclosers have been in~oduced in conj intemptions a u ~ a ~ a t i c and a ~ ~are y therefore sometimes considered as a form of n e ~ o r k au~omation. ~ e ~ o secondary r k system (i.e. V systems~remote c o n ~ osys~ems ~ have been varying degrees by distributors in the UK. The most s ~ ~ n i ~ cinves ant ~ o n ~ oElec~ic~ty, n E a s t e ~Elec~icityand S o ~ ~ ~~ e c ~~ c i ~ . much faster res~ora~ion o f supplies following faults on cable network^ or on ely distant fiom o ~ ~ r a t i o ncentres. a~ They also reduce the n cing the amount of time an engineer needs to spend swi reducin~the risks associated with this activity. The case study at the end of this section considers the scheme impIemented by London Electricity, the b e ~ i e ~realised ts and the long-term potent~aiforeseen for the system. These projects have as part of their implementation s i g n i ~ c a ~ t cl yo n ~ b u ~ etod the ~ e p l a c e ~ or ~ n ut p ~ a d i ~ofg p r o ~ ~ e m a tnetwork ~c apparatus, be it overhead line or swi~hgear.

re, they can o p t i ~ i ~their e e l e c ~ c a~~licatiQn i~ sched~lingto cost savin~s~n~~~the , tend to maintai In a d ~ r e ~ ~c a t ~ dve ~ power m ~ k e t utilities

Power System ~

12

e

s

~ andc ~ ~ ~ n~ l~ a t i o

roposed method for a VSTLF has been s u c c e s s ~ l ~~mp~emented y in a ~ o w e r utility in the USA, and is used by dispatchers for on-line load forecasting. The developed f o r e c ~ t i ~system g predicts eight values of load for the time leads from 20 to 90 minutes in 10 ~~u~ increments. To provide dispatchers with the information about e~pectedforecast e ~ o r s ,mean absolute percentage errors ( ~ ~ E are s )calculated base forecasts for which load ~ n f o ~ a ~has i o nbecome available. For the 20, 30, .,, 40 minute [9]. Load data is fore~asts,the mean absolute percentage error lies in a range o f 0.4~~.~% from the automati~generation control ( A ~ C )system every g data is converted into lminute integrated loads which are consider~das COUS) loads. These loads are used as input i n f o ~ a ~ ~ for o nc o m ~ u ~load ~g pre~ictionsand they are also stored for training. e forecaster~sneural nehvo automatically retrained once a day. It is not the intent to discuss in detail load forecast~ngand its d i s a g ~ ~ ~ a tto i oanlevel where the installed capacity in the distribution n e ~ o r k can s be es~ab~ished. The advent of the increased demand for telecoms data and Internet services has also to add s i ~ i ~ c aloads n t to the d i s ~ ~ b u t ~systems on pa~~cularly f x the associated ing centres. The m a g n i ~ d eof such loads (10-4.0 ~ W and ) time scales (12-18 months) for such developments are such that the d~stribut~on com~anyhas to be in a position to ~espondcreatively and ~exiblyif it is to avoid l o s ~ nthe ~ either to another company in a di~erentlocation where supply can be a r to a competitor who is prepared to establish a separate dis The large urban centres such as London are not see in^ the m a ~ i demand ~ u ~ due to more efficient loads as these are being offset by these ~T-relatedincreases. ~ng businesses have adop a five-year p ~ a n ~ horizon, sed d~s~ibutio1~ culties of predic~ing~ e m ~ d l a r ~ ~ due l y to the ~ve-yearlyreview period and the shuction times related to major s y s t e ~changes, changes in bstation projects make the ~ ~ e - y e~a re ~ ~o ~d~ a ~short vely e tec~ologiesand s ~ c ~ or e networks must be tail in the ~o~lowing ieve this end are discussed in In

asset management and planning is to integra~eas replace men^ of poorly performi or high~riskasset a ~ d e ~~ a r ~~ ewill ~gt need s ay an i n ~ ~ ~ a s i n g ~ ~ asset r ~ ~ l a c e p~r ~o ~n rta m ~ with e s n e ~ o r kreinforcement and major new c o ~ e c t ~ o n s works. The p l a ~ i n gof asset rep~acementi s discussed later.

4.3.4

Long-tt?m int of l o n ~ - t ep~anning ~ is to d e t e r ~ ~ nhow e extema~~n~uences, of new business and changes in the regulatory env f the network and the levels of i n v e s ~ e nthat t will

-24

Power § y s t e ~R e s ~ ~ and ~ ~n ~gr e g ~ l a ~ i o

In time adopting such a me~hodologyshould Lead to an ~mprovedm a t c ~be ents and the d e ~ a n they ~ s must uch t e c ~ ~ q u hav es d in a number of projects worldw d are ~ i c a l l ystandard ~ r ~ c t ~ c e ians for ~ n t e ~ a t i o financing na~ agencies. on ~ ~ s c u s s ethe d use of a set of 10 A similar ~ e c ~using i q three ~ ~altem ness ~ l a ~ n i n gThree . views of all ent and investment drivers are normally eve loped being:

she goes’ view: a sable environme~tbased on exi ions encompassing a reasonable view of the effects of known de

mic and

s p e ~ ~ c l view: e s ~ a positive view of the dev~lopmen~ of the economy how this will impact on the demands on the business. oomy’ view: a more n e ~ a ~ vview e conside~n the impact of a ~ o n t r a c economic ~~g e n v ~ r o n m ~and n ~how is w o ~ l dimpact on the business. e would look at a range of business factors, d priate s ~ a t orc strategies. ~ A ~ ~ ~ l a tofi othe ns ss fxtors allows the p l a ~ e to r i d e n ~key i ~ s~~tegie an one scenario, as i ~ ~ u § ~ aint eTab1 d

winess Factor 1

S c ~ n a ~1 o §~ra~egy 1A ~ c e n ~2i o Strategy I Strategy 16 Scenario 3

Business Factor 2 S ~ a 2A ~ e ~ Strategy 2A Strategy 2 6

Business Factor 3

Strategy 3.4 Strategy 3 §trategy 3

Business Factor 4 S ~ r a t e 4A ~y § ~ r a 4B ~ e ~ Strategy 4C

are robust to more than one scena~o,

h t e ~ ~ i q u are e s widely used n i d e n t i key ~ ~long-term ~ le

d i f f ~ ~ c software nt tools exist to aid the esign of power s y s t e ~ ~ . s studies and a few a ~ a ~load ~ Row i c and ~ ~fault level use a fa^^^ rate c a ~ ~ ~ a t hols. ion which does not, in

a fault rate a ~ p r o a is c ~that it on ofthe under~yingcausatioii

The asset ~ ~ a g e m ed~scip~ine nt and network p l ~ n i n g ~most s s i ~ i ~ ci~iter~ace a ~ t i s in the planning of asset r e p l a c e ~ ~ n tThe . developrnent of asset ~ ~ n a g e m e n ~ s is to convert these policies &to covered in detail eIsewhere, The p l ~ e r ’ role programmes. In doing this the lamer must consider the asset mana objectives and the condition and carried out.

4.3.7

Risk Assessment

Risk assessment methodolo es are useful in any business and distribution b u s ~ e s ~ are es sk assessment is applied at two levels, the business level and for g ~ n d i ~ i d uasset a ~ asse~smentsas part of the asset replacement p l ~ i n process. ~usinessrisk analysis considers all areas, including network perfo~ance,finance, commercial (e.g. use of system income), contractual and regulation. Potential risks in each area are identified and probabilities and consequences determined. Fin measures and appropriate actions to control the risks identified are establish d distribution businesses, particularly where there is n supply, the largest risks are often associated with the income streams owing to the complexities o f the data acquisition and ag and the number of different parties involved. However, network risks must not be ignored. Historic control measures exist through planning and construction standards such as the UK’s Engineering Recom~~ndation PU5. In planning individual i n ~ e s ~ e n trisk s , assessments are normally con likely ~a~~~~~mode f ~ i etc. ~ Major ~ ~ nee ~~ ~ failures rk such as that Auckland, New Zealand, and the recent weather-related i n c ~ d ~innCanada ~ and France have prompted further debate on the appropriateness of existing design standards and the cost and be~efitof c ~ ~ these, ~ n g

4.~.8

ills and

ince the privatisation 0%:the d ~ ~ ~ binu the t o ~ there ~ has been angain ~educecosts. This has ~ e v i t a b ~resulte y in a very signi~cantr e d ~ c t ~ oinn up asset m a n a g e ~ ~ on rt ~ ~ n ~ s a t ithese o r i ~have a ng of the n ~ t w o r ~ It s . is e ~ i ~ e n t incipal activiti~s,new c o n n e c ~ and ~~~s en separated, the practical i m of ~ E the staf~ngof these org es a differ~n~ set of coi~petenciesthan

it levels of various c ~ m ~ a n ipl~ s ’ orn the d o ~ s i z i n gunder~ken ibt one can no Ion

~

Power System K

12

~

s and Deregulation ~ ~ c

h i g ~ yquali~edbut able technicians has been es~ablished.The p~anningskills of the more expe~encedstaff are gradually being transferred to the less experienced team members and the lost competencies being replaced. Whilst it can be argued that these skills have been r e ~ i n e dby some of those utiiities rt to best practice, the skills gap is being gradual~yredressed. With ing of competition in connections design and provision, the area of will be to ~ ~ ~ ~ nand t adevelop in the i n f r a s ~ c ~planning re skills necessary to review the overall network successfully and d e t e ~ i n ewhere action will be required to m a i n ~ j nand ~ ~ p r existing o v ~ levels of service. This is probably one of the areas most under pressure at the present time, especialiy with the increasing c o m p ~ e x of i ~ systems such as remote control and automat~o~ being introduced and the increasing asset-manag~nient-derived workload. The use of expert systems to capture experience and make it widely available has not yet been widely adopted but is an obvious opportunity to suppleme~tprocess charts in more complex and/or less routine operations.

4.3.9

~

e

~Design ~ r

k

There are three elements to network planning that need to be conside~ed~ these being the connection of new load, the reinforcement of the system and improvements to meet quality of supply targets, f new connections is driven by the regulato~requirement to offer the lowest cost connection and the need to meet larger customer’s needs. The former of these has given rise to a conflict with some aspects of network desi to meet ~ u a l of i ~supply targets. For example, the simplest design to connect a voltage load of less than 1 NW to the system is to create a new substation and connect it to the existing network via a tee off xisting circuit. If a large number of customers are supplied from this single source, such as a large housing development, then ideally the substation should be connected so as to be ~ o o into ~ ethe~existing circuit, as shown in Figure 4.3, in order to lessen the ri repair time outage.

.3 Tee vs. loop connections

~

D ~ s ~ b u t ~inoanDeregulated Market

127

ecision as to whether to fund this out of quality of supply monies will depend on the number of customers, the distance from the main circuit and the additiona~e n e r ~ losses incurred. The advent of competition in connection services would further compl~cate this issue. The network manager will have either to pay the contractor to hstall the additiona~cable at the same time or retrofit the additional cable at a later date. The customer will not be expected to pay for the additional costs related to the quality of s ~ p p l y as part of his connec~oncharge. This c o n ~ will ~ ~ also m apply to the installation of spurs to feed a number of customers at low voltage (LV) where no alternative back-feed arrangements from o&er n ~ ~ oare r ~ available s or where the installation of remote terminal units for SCA remote control may be desirab~e.Evid~ntlythis becomes easier to manage as the size of the load increases and the number of connection requests decreases. e ~ Genernl load growth and the connection of new load drive the need for n reinforcement. Typically the impact affects the thermal ratings of the network appara~s, security of supply or the voltage p e r f o ~ a n c eof the networks, but recently greater is having to be given to managing power quality issues, particularly harmonics. In most estab~ishednetworks the general growth of load is relatively low. In c e ~ a i n areas, p a ~ i c u l ~ lhighly y urbanised areas, redevelopment has seen prospective loads increase owing to new office developments and the associated IT-related loads. At the time of writing this would seem to be a developing trend, the forecasting of which represents a significan~challenge. The management of reinforcement with the connection of new load has become the most s i ~ i f i ~ achallenge. nt The management of the new connections process is b~coming progressively more detached and this is likely to lead to an increased need for manager to monitor connections activity and identify reinforcement requirements and implement them in an appropriate time scale. The failure of this process will ultimately impact upon a distributor’s ability to meet a customer’s connection requir$ments within its schedule. In large urban areas this may have not simply a financial impact on the distributor but also an economic one, The present regulatory process in the UK which involves five-yearly reviews to fix income for the following five-year period increases the risk of increased r e i n f o r c ~ ~ e n t exp$ndi~reaffecting other capital programmes,

As discussed in the previous section distribution automation in its simplest form has begun to be used to ~ m ~ r o vquality e of supply. At this level the automation inst~lIedc~n§istsof auto reclosers and auto change-over devices. The ins~allation of remote terminal units provides the basis for a dist~buted c o I ~ ~ u n i c a ~ i system o n s that could be used to implement some degree of automa~io~. For the convenienc~of the readers, an appendix is included below to detail ~ i s ~ b u t ~ o n automation and comm~~nication systems under a competitive env~onment.

Power System ~ e s t ~ and ~ ~~e r~e ~n l ag ~ i o n

Two levels of automa~onexist:

Via cen~ralc o n ~ Qsystems 1 Via e ~ b e d d e dsystems. bedded automation systems suc as auto reclosers ge-over sy§tems are probably the easiest dis~ibution autom~t ement. No ~ i g ~ - s p ce Q e ~r n m ~ i ~ ~ tsystems i Q n are required as the rn of c o ~ n ~ u n i c a tto i oinfarm ~ the central control room or system e but not essential LO realis bene~ts. systems have grea~er ilities if h~~h-§peedlocal C Q m ~ u ~ ~ c asystems t i o i ~ can be ~ m p l e m ~ ~ ~~oss~bilities ed. could include on of f a ~ ~ c~i re~di t all~wing s ~ e s ~ ne ed ~ o oper~tiarn r ~ w ~ t the ~ o ~ ~ ctional protection s c h e ~ e sor unit ~ r o t e c ~ o n he . hi ~ornmu~~~cations requir~dcould be achieved by a ~ ~ m boef means r in~~~~ing:

s , it WO Id produce many of the Same b e ~ e ~ tbut ickly. The c o ~ ~ i c a t i o path n § and c e n t r ~ ~ i n f o ~ a t i o ntralTc from the entire n rocess several scenarios simul~neously . However, such s or changes in n e ~ o ~c k e can be ~ a i n t ~ i n e d system. r the following benefits:

e ~ u ~ cO ~r ~~I on o acontrol ~ engineer§,

data and the presentation ofuseful information to the control engineer as to the actions that have been taken by the system. Improvements in the speed of restorat~onor securing of fault may in fume allow increase asset utiiisat~onby supplies f o ~ ~ o wa ~netw g permitting higher short-time loading levels as the duration automatically. 'This will of G Q U ~ Sdepend ~ upon the network configuration, but it increases the potenti~~ bene~tso f r n o v ~ ntowards ~ ~ e s h e dnetwork operation in the m e d i u ~to on^ term. A reduction in the need to carry out manual switching has a significant safety benefit, itchgear, as an operative does not need to be in particularly with ageing oil~ ~ ~vicinity. e d Thisi is ~ ~ ed ~by the possibilities for local cond~tionm~nitor and rcmote indication of alarms 1e:elatingto possible hazardous conditions.

4.3.16

Autor~ationCase Study - Remote Control in London Electricity

se study considers the planning issues concerned with the 10 term d ~ v e ~ o p mof e~t London ~ ~ e c t r i cnetworks, i ~ ' ~ p a ~ ~ c u ~the ar~ changes y m d e to i ve qua~ityof § u ~ p l ~ . The ~ ~ ~ o d uofc at remote ~ 5 ~ control and data acqui§ition system has been c ~ to ~ lectricity's development lans €or its secondary (MV and LV) networks over the

This case s ~ d yconsid~rsthe p~ilo$ophybehind the ~ r Q g the~ ~ e ~ deployed, the ~ d v a n t a gLondon ~~ believes these offer over alternative system performance d e ~ ~ vtoe date. r~~ review of secondary syst m m e was brought to th control review in 1 9 9 4 / ~when the re lator's focus on quality 06 supply ~ ~ p r ~ v increased. The remote o ~ e r a t i oof~network ~ w ~ t ~allows h e s s w i ~ c ~ i ntogo an engineer can reach the a ~ ~ e c t area, e d often in excess of an hour i reducing the intemplion time seen by a large number of aEected cust ~ ~ o g rwas a ~therefore ~ e tar eted to ~ e d u c~ues ~ o ~~eirn u t e lost s from am to a targct level o f 40 in 2000. all asset ~ a n a ~ e r n e plan n t required that the remote to deliver data ~cquis~tion and ~ ~ ~ e ~ l i oni g e itn or t in^ the basic control necessities. This approach has proved well foun as will. be discussed later. et performance improve men^ a ~ ~ r e e - $ approach ~ge was s across the 6.6 kV and I I kV ~ e ~ o r k ~ . approach a i ~ e to d d e l i v ~as~much benefit as possible in the initial phase. The first stage was aime targeting one in every four ring mail1 units in ~ e r f o ~~ e~~ go rTh~ s . networks in London E ~ ~ c ~ i c i tsyste y's groups of about four circ ese groups are run radially with a nu between them, e f f e c ~ ~ ~creating e ~ y an open four-feeder ring. However, d i v ~ d ~into d two er ~ a t ~ g o ~~i e se ~ on ~w~ether n ~ the ~ns s o~c i a t e ~

13

Power System Restructuring and ~ ~ r e ~ l a t i o n

operated radially or interconnected, i.e. operated as a mesh. The meshed LV systems are typical of the centre of London and assist in coping with the hi ~ e § t ~ i n s tand e r the City of London. The greatest initial benefit in quality of supply performance was to be gained fiom ~ r g e ~ i nthe g areas where the LV networks are operated radially (the radial areas) no ~ u p p in o ~the event of an MV fault, which is a characteristi~of the in~erconnectedLV system. The feeder groups supplying thc radial areas were ranked in order of their e over the pervious years, bearing in mind any major asset repla~ment to correct high fault rates. U installation programme was then targeted in these networks at open points urth ring main unit which offered suitable switching point. In order to achieve the switching ~ n c t i o n in~ the i ~ existin ring main units a p r o g r ~ m eof retrofit actuator solutions was developed to mitigate the ed to replace switchgear. Initially this e ~ units was t a r g e ~ eat~modern SF, ring main units and some of the more m o ~ oi~-fil~ed which were deemed suitable. This resulted in remote switching being ava~lab~e at an open point and at the approximate mid point of each circuit. A fault passage indicator with provision for remote indication was fitted with each ins~al~ation. This would allow 50% of each feeder to be restored by remote switching, approx~matel~ The second stage of the programme extended the provision of remote control facilities to ~pproximatelyone in two ring main units, again with the initial concen on the w o r s t - p e ~ o ~ i nfeeder g groups, In turn this would allow up to 75% of be restored by remote switching. The third stage extended these facilities to those ne interconn~ctedLV systems. This is a more complex task as ea equipped with an LV air circuit breaker (ACB). This is installed to prevent network collapse in the event of an MV feeder fault, due to either a fault infeed or resulting network overloading (it being preferable for the ACB to op network fuses which then have to be identified). It is operationally desirable for th to be con~olledto reduce the number o f site visits by e ers in the event of a while a ~ e m p t i nto~ secure supply. It is also necessary to know the status of the s u ~ ~following ~ i ~ sa fault, so remote indication had to be prov~ded. A one-in-two strategy was adopted as this was felt to be the m i n i ~ un ~e ~ s s to a~ e the degree of control required to secure supplies remotely without the need for an eng~neerto be present in the field. f 1999, London Electricity had equipped 3000 MVLV substations with emote control facilities as part of stages I and 2 of the ~ r o ~ a described m ~ e 3 was ~nitiatedin late 1999 and would begin to take effect in the least we interco~nectednetworks during 2000. The majority of these, 2000 of them, 1999. The performance of the programme has been excellent with customer minutes lost visibly reducing with the numbers of units in commission. Supplies are now restored to all customers within 1 hour for over 50% of all M Y network faults. Most pleasing of all has been the n ~ b e ofr routine switching operations that were soon being carried oat using the system. Figure 4.4 shows the theoretical performance estimates made when the project was co~ceived.The t ~ curve p shows the predicted p e ~ o ~ a against n ~ e the b o t t o ~curve,

~utionin B

0

I

2000

4000

BOO0 8000 RTU P O P U L A ~ I Q ~

._.-_ _ . (HV) trend.

10000

'I2000

. " - . - ITrend. (Overall)

24000

estorafion performance

The original vision for the development of the initial remote control system was to create an n e ~ o ~r ~~a g e m esystem. nt The advance , e x ~ a n d a b and l ~ to a l~rnited l e n ~of ent RTUs were chosen ta facilitate this deve ment. The ~ ~ i i te~l ea m the ans &realready being i m p l e ~ ~ n t e dAn . auto change-over ~ ~ e c h a has ~ ~ been sm i m p l e ~ e n t ein~the e ~1 s logic to allow supplies at open poi to be r e s t o ~ in 1 ~ ~ ufo~~owing t e ~ ~ ~ fault. o r This k de~iversan 1 as the customer minutes lost. ated res~orationo f expanded to deliver There are still many to be o v e r c o ~ einc location logic and c o ~ ~ ~ ~ i c a t i The o n sneed . t unicate to a ~ ~ m b than ~ n c in e less than a minu~ewill pose a s i g ~ i ~ c achallen nt There are, howe~er,other aspects of network ana age R W was specified to cope d ~ o n i t o ~ nand the LV s y s t e ~via ~ d ~ ~ t i o n g con to inte~acewith the e developed to include discharge levels in cables and v ~ r ~ ~ i n ~ ~ 6 a ~of o rthe s c o n ~ i ~ i oofn switchge~and r n o ~ i t o ~ nofg the LV load on a s~ngle-phase s even the location of faults.

e of the main rea§on§ for uti~~ties to advance a u t o ~ a t i ~ofn ~

~ s ~ mice. ~ e r e ~ u l a t i oisn evolving to establish some o penalise utilities in case o f ~ o w - ~ u§ae ~ iic~eh . cy will grow, and high peak prices fo cast load control will help manage the risk associ

ion auto~ation~ ~ o v i the ~ eability s to c o ~ u ~ i c avital t e in gives the ability to monitor and control that information from a central location. In sho the s y § t e ~could tell whose e l e c ~ r i cis~out ~ before anyone calls to ~ o ~ ~ ~And a i 1n . could save money by re§tori~gpower sooner than ever before. i§tribution a u ~ o ~ ~ t i o allows ~ o ~ t ~ ~ o n ~ ~ ~ to ~ t provide ~o s~ ~the greal-time ~ ~ w ~ needed e d ~to eo p ~ ~ i s € can ~ a x i m i s ecustomer satjs~ac~ion with ~

ns with remote e ~ech~ology require

i

structuring and ~ ~ ~ ~ ~ l a t i

134

improving revenues and reducing costs. ~istributionautomation i s a complex subject g the following major coi~ponents: remote terminal units (KTUs); A master station (open systems design, full ~ p h i c s ) ; ~nctionality(feeder sectionalisa~ion,cold load p i c ~ p topology , processor, voltage/var control, graphic feeder tracing, switching order preparation, special reports, automatic meter reading, etc.); operations and maintenance procedures (safety, tagging, pe its, c~earances,work orders, preventative, routine and restorative practices, spare parts, service ~greemeiits~~ s y s t e ~integration, design and management; and c o ~ ~ ~ c asystem t i o(e.g. ~ cellular phone, radio, power line carrier, ~ ~ ~ p ~ e ) .

4.52

emote mina^ Units

C o n s t ~ t ~involved y with improvements in RTCT technology are things such as the develop men^ of ladd~r/sequencelogic/PID algorithm ca~abili~ies, multiple serial interfaces to a c c o ~ o d a t esmart meters and relays, peer-to-peer rotocols arid direct ~A~ For d i s ~ b ~ i t i oautomation n purposes, small, low-power, w of, c o ~ p aRTUs ~t are available. These come in a variety of enclosure packages, fically cons~ained points counts, direct CTNT inputs, AC analyser modules to give a variety of calcu~ated i n f o ~ a ~ i o and n , more. The units can p e ~ data o ~logging to m ~ i m ~ the s e need for constant polling via the communications system, In some applications, peer-to-peer com~unicationshave been utilised to facilitate independen~islands of automation ( volta~e/varcontrol) that do not rely on the master stat~onfor decisio~~ a ~ i and n g control actions. initiated repo~-by-exceptjonprotocols are being utilised by sorne vendors to keep power draw (from constant co~m~nications with the master stat~on)to a ~olar-poweredunits are in common usage. Compact, ~ow-mainte er i s also available. With the advent of two-way commun~cat~on its, a great future lies ahead for dis~butionautomation and manage~entapp~icationsat the customer level for functions such as remote selective control of customer loads, surveillance of customer installations, choice of electricity rates.

The d e of modem ~ ~ ‘open ~ systems’ SCADA c o n ~ ~ a t ~ makes o n s use of consid~~able communic~~ons t e c ~ o l o g yto spread the risk that a single f a i ~ will ~ e wipe out or ~ s a b ~ technology permits hi a mission-c~tical system. odern ~~~A~ processing and achievem of graceful degradation upon failure hics d~pictionsof system assets, works~ationsand personal computers give users ful very ex^^^^ win~owinto the often in proper ~ ~ o g r a porient~ti~n, ~c and prov systems they are controlling. The dissem~nationof computing elements and the ~ e x i b i l ~of t y full graphics interfaces have in~reasedthe b ~ r d e nupon the system

Risti~bu~io~i in a ~ e r e ~ ~ a Market ted

135

these numerous features. The control system ~ i e ~ a r c hfor y the whole o f must be flexible, with its topology adaptable to meet the c h ~ ~ i n g d a ~ b a s e sneed to be kept synchronised so that all users view must be c o n s ~ c t e din a manner that makes operator na~igation nition simple under the most stressful situations. These items, if can lead to poor operator acceptance of the new tool. system architec~recan now be dynamic, allowing change and enha~cement over time as both user needs and technology change. Relational database ~ a n a g ~ ~ e n t systems have facilitated easier, more functional interchange of data with other c o ~ o r a t e s y s ~ e ~(accQ~~ting, s custonier records, maintenance m a n a g e ~ ~ nwork t, dispatc~,etc.). Compact, high-de~sitystorage media have simplified the tasks of lar b a c ~ pkeeping , historical data and managing archives. Disk shadowing pro hot standby data req~irementsfor key operational areas. Distribution ~ u t o m a ~ functionality io~ in the ~ C A must ~ A work with the actions reactions of the distribution system protectiQn equipment. Actions taken by the logic or s protection and SCADA) must not compete with c o n ~ oa ~~g o r i t h ~ofs both $ y s t e ~ (e.g. r to cause additional roblems. Those applying distribution a~tomat~on to wer systems must ensu that all protection schemes and systems are thorough^^ and catered for. One needs to have consid~rableexpe~encewith protect~on schemes involving:

smart relays which f e a ~ r ~e u ~ t i pse~ings Ie and c o ~ u n i c a t i o n sinte~acecapabilit~es; relay, recIoser and tap/transfo~erfuse coordination; and lel opera~~ons between buses or substations. ~~C

ddit

standard power system SCA

feeder load shedding - c o o r d ~ a ~ manual, ed rotational and under frequency schemes; processor - provides up to the moment topology and energisation status; cedure generation and management^ dis~butionpower flow; -load p i c ~ estimation; p former load ~ a n a g e ~ e n t ; supply interruption reporting and outage management; fault ~ o ~ a ~ i a n ; t~ansfero~timisat~on (load transfer and recon~gurationc a p a b i ~ i ~ ) ; automated feeder r e c o n ~ ~ r a t i and o n service restoration; voltage/var control - t r a n s f o ~ e tap r and sw~tched-capacitorm ~ a g e ~ ~ n t ; dis~ibut~on short-circuit analysi~; d e ~ ~ d - s i manage~~ent de - load management and time of use strategies; aphics capability to ide schematic display, switch posi~ionm a r ~ i n lemetere~),feeder c tivity status, and energisation status info in real-time an the operators’ VDUs; md training $imulato~s.

4.5.4

Softwui-e F ~ n ~ ~ i o n a l i ~

1st the ~ a ~ d of w a~ Se C ~ system ~ A is of gr primarily in system and applications software. The lity, s u p p o ~ b i i and i ~ mainta odem systems buil~to inte~ationaiiyaccepted Windows e n v ~ r o n ~ e n t ~ r ease of pro~rammingand p ~ ~ a b i l i ~ ; t any language to e-critical and comput~~gompete for co~puting ase co~nectionto allow easy passing of real-ti~eand ~ i s t o ~ ~ a l database and system t o p o ~ -oassists ~ w i ~ hease of s es ~ o p o l o ~p yr o b l e ~ much s easier; and that are based on tried and proven 8 n have significant co~sequen~es and gorous design ~ h e cv e~ r i~~ ~ a ~tproced~es, i o, ~

4.5.5 ~ u n i c a t i osystem ~ for ~ ~ s ~ b auto~ation u t ~ o ~is s ~ i ~ e rrece~vers s, and data links. The s y s t e ~s h o u be ~ ~desi ~ersonnelwill have to b nt~nancewill be as easy as possib~~. involved and new tools will need to be purchased (the o f a pote~tial system). ent will s ~ ~ i ~ c a ni ~ t lpyr o v e use of s~andardisedcomponen ~ o ~ not $ donly allow better compati~i~ity with existing communicatio that the s y s t e ~will remain also i ~ c ~ the u ~likelihood e and a ~ ~ o m ae~~i uo i~p mdeveloped ~~t in the future. This ~ ~ i n ~ e ~costs a n ctoethe u t i i i ~ . deve~opapprop~ate0 k m a n a g e m ~analysi ~~ roached by each u ~ i i ~ .

~b

s

~

~

~ i ~ ~ r i b u tin i oanDeregulated Market

There are several aspects of O& include:

for distribution automation to be considere bution automation installed (who does what, why,

ion automat~on~ u i ~ m e n t . ation equipment, software, d a ~ b a s e an ( ~ o n ~ g u r a tmanag~ment~ io~ spare parts holding^ service contracts~~ Training to suppor~the accepted p~losophieso f operations and ma~ntenance both cl~ssroomstyle and using simulator scenarios.

-

int-to-poin~’w ~ ~ between ng the and impossible to solution i s needed addi~ion,an inexp

into fbture techo

emote monito~ngand control. nications ports with n-line con~guration. g ~ d l i an large ~ number of points while ~ a i n t a i n i ~real ion of SCADA s o ~ a r efield , equipment, system integr ion, commun~ca~ons au~omationhas to be on of an integrat~dsystem with both high-voltage sub smissio~networks feeders. This would help to optimise operations requirin rew working at d~fferentvoltage levels. of power n ~ ~ o diagrams, r k plant data and teleme ntrol (remote or manu ltiple ~ a t a b a s ~and s

01s for p~anningand o~timi§at~on studies. faces to fault maiiagem~ntand custom~ri n f o ~ a t i Q n and map h an age men^ systenis. to the success o f any ~ e ~ e c o m m ~ ~ i c a tand i o nns ~ ~ ~ o r k ~ n Systems ~ n ~ g r a t is i ocritical ~ i n i t ~ a t ~ Since ~ e . most sy~temsare not developed in a vacuum, ~ntegrationo f ex is^^^ or essed. This ~ntegrationmust be r ~ ~ u n ~ c a t i level, o n s which ensures that ex of working in the new system to e n s ~ integr e

applica~~ons level, which ensures that ~ n f o ~ a t ~genera~ed on an a p p l ~ c a can ~ o ~be accessed by another application. 0th levels are critic~lto the success of the system and the organisation’s ope~ations.Access and a v a ~ ~ a b iolfi fflformat~on ~ in a timely~accurate and user-fiiendly manner are necessary for the system to be a success. The develop men^ and i m ~ l e m e n ~ ~ i oofn any t e l e c o ~ u n i c a t i 5 ~system will affect o ~ ~ ~ ~ soperations. a t ~ o The ~ ~success l of any project i s a direct result of the a~entionto detail given to system specification, design and i ~ s ~ l l a t i oThis ~ i . ~nc~udes ~erificat~on that what was specified and procured has been delivered, testing of system co~ponentsand ap~~~ca~ and ~ oensuring n s ~ that the system satisfies ~ ~ n p l e ~ ~ tr ~a ~t ui ~o r~e n i eand ~its r e ~ l a t guidelines. o~ ana age men^ i n f o ~ a t i o nsystems (MIS) are becoming an ~ n c r e ~ i n g li y~ p ~ ~tool a n t in the daily operations of electric ut~iities.The i n f o ~ a t ~ osystem n is more a col~ec~or,r e p o s i ~ o and ~ transpo~ ~ e c for~ i n f~o ~ ~ ai ~ i o~An . w~ d i n f o ~ a ~ i system on is a combination of hardware, software and c o ~ n ~ ~ c a t i~o nas ~ a b j rnis the foundation of efficient opera~ionsand dec~sionm ~ ~ i ~ ~ . ode^ ~ntegrate~ network management s~stemis used to control remotely and to supervise manual operations on MV distribution equipment. The system au~omaticaliyprocesses topology and highlights d~-ener~ised feeders when devices change state after telemetry input or manual dressing. System Alterations and s w i t c h i ~ sc~edules ~ a~ly are prepared in advance and operations can be a ~ t ~ n i a t i ~ ch d ~ ~ safety n e ales. ~ Power analysis functions can ~ a l y s ethe n e ~ o or r ~individual distribution feeders. One of the major ben world-map schematic diagrams, plant parameters and network CO one co~sis~ent system. Data is held at a variety of levels of de analys~sand detailed device operation. twork operation functions are ‘those functions which enable control an d i s ~ ~ u t i onetwork n facilities’ and inc~udecontrol, mon~toring,fault erating statistics. ~perationalplanning functions are ‘facilities to de optimise the sequence of operations required for carrying out maintenance work on the system’ and include network s ~ u l a tand i ~switch ~ action ~ c ~ e d u[IO, l ~ 1g13. The primary purpose of a network management system for network operations is to patch o f field crew (people) to ~ a i n and ~ ri ~ the network, safely and whose primary ~ u r p o s ~ differs from an energy management s y s t e ~( is the dispatch of power (MW and MVAr}. The modem ~ o m p e t ~ t ~market v e emphasises that utilities need to monitor and improve levels of customer satisfac~ionas well as o p t ~ ~ s i network ng ~ p ~ r a t i o and n s controlling operationa~costs. ion creates a new wave of electronic brokering as electricity is bought and odities market, Tracking of these ~ a n s a ~ t i o w n s~ t h ai given ~ utility should be m a ~ a ~ e a b lheo; w ~ v emost ~ , of these ~ ~ s a c t i o will n s span mul~plec o ~ p ~ i eIns order . to achieve interoperability, implement of a common information el has a data structure that is c , The common information S p r o ~ r ~ esystems. ta~ Most infornaation networks wil! be connected to the EMS to provide accurate real data to stipport the available transfer c a p a b i ~ (ATC) i~ calcu~a~~ons. ~ ~ e r mar t e ~ ~ ~ beL used ) to present i n f o ~ a t i o nto customers. ge ~ ~ T will for customers to use to request d by the transmission services i col ( ~ T T Pi s ~used for data ~urchasesfrom a provider. The

- x

Distribution in a ~ e r ~ ~ l a t e d

then: IED on the n e ~ o ~ k .

increa~inglyc o ~ ~ e t i ~

ovide m e a ~ o~f de~~a m i security c to allow the system

A will allow ~ t a b i ~ i t y - c o n s ~utilities a ~ e d safely to inc~easethe loadin will also allow ~ i ~ e l y can also i ~ p r o v es y s t e ~reliability, essment of the security impact of ~ a n s ~ ~ ~ ~ o n s lt af open axess to the

that a p r e ~ i ~ i dn ea~~o n s ~ a ~ usi ion has resulted in a p o t e n ~ savings ~l of ove enefit of the p ~ o d ~ c t ~version on ~ i be a~5 %~inc~eas~ t in t r a ~ s ~ i s s capacity io~ across a cons grease^ capacity could be used on

B

integrat~onprocess should drive all utilit~estowards the standardisatioi~of data ~ ~ o d e l s c o ~ m ~ i c a protocols. ~~on ay an ~ncrea§inglyimportant role in the daily o ~ ~ a of ~ ~ o n ust a means of ~rovidingconnectivity b e ~ ~ one e nperson and t e l e c o ~ u n i c a t ~systems ~n are the c o ~ i e r ~ ~ o fn e ing b e ~ coin~unication, ~ r not only b e ~ e e ne ~ p ~ o ~ e e s recent ~ h a ~ g in e sthe teleco

r~l~~ionsh~~s. ~ o ~ ~ u n ~requires ~ a t ~~ aon sn~ i § s i channels, on which

services from the carriers who

c o r n ~ a ~found ~ ~ sthemselves with campus-wide EANs capabl~o f ~ f ~ c i e nh ~a n~ ~y ~ ~ n y com~arison,the data and not well s u i ~ eto~computer~~o-co

e early 199Os, ~

a began ~

e

~

~

s, lower delay and lower CO ment at each end, much as is s users, because they concen~atetraffic fkom m links from their premises to the carrier’s centra the c a ~ a c of i ~fiber. Fiber not o sion of c a p a c i ~s ~ m p ~byy i n s ~ l l i ~more g by businesses for ~ber-basedaccess has i and an a ~ ~ e ~ a t e

ase s ~ t i o nto a cell served by another, the wire~essaccess link is automatical~y to the new base station.

~

~

s

~in ab Deregulated u t ~ Market ~ ~

services digital networ~swhicb bring common channel si alling right to the ~ t h e ~ was e t chQsen as the d a b link layer because of its predominance in the e subsequent availability of low-cost imp~ementationsand assoc~ated s b ~ d g e sand routers). In addition, the scalability o i i ~ p ~ e m e n t a ~being o n s fairly common and 1Gb Eth its way. ~ r o c e § are ~ o av~i~abIe ~ tooday with multiple 10 Mb Ethernet the chip, and next-generation d e s i ~ are s planned with 100 device, two solutions As it was d e e m e ~desirab~eto be able to access data : Transmission Control P r o ~ o c o l / I n t ~ ~ PrQtOcQi et n t e r c o ~ e c (OS). t TCPlIP is a networ ational ~ ~ d ~ d ~ s a[l4,15], t i o n it was e I S 0 network layers. These layers have robust flow control ery useful on a busy sub§ta~ionLAN. th network layers s u ~ p the o~ to hear. This f e a ~ r eis very ng a message for all devices on the s such as data capture triggering, time s ~ c h ~ o n i s a t iand o ~ , even eel models [ 16,171 are used because they can easily ay makes measurements of voltage, current and The m e a § ~ m e n t smade by the re~aycan be conta~ingall the elements mentioned above. If r quality and power factor are added at a later date, the ori~inalmodel i s easily expanded to ~ c o ~ m o d athis t e data. Fully in~~ractive co~municationsystems provide a full range of voiceldata transfer, remote access and controj, entertainment and e d u ~ a ~ ~ o~ nd. ~ e d c o ~ ~ u n ~ c a t ~no~n s~, o capabili~ r k and services may also have an impact on the s t a f ~ n Qr~anis~tional needs of the organisation. An evaluation of the existing staff, roles r~spo~s~bilitie§ is necessary to determine if the required capabilities exist or if new s staff are requir~d. ith the con§~ntlychangin ~e~ecommunications ~ n v i r o ~ e and nt almost any ~ ~ ~ a n i s ato~ o nffer t e ~ e c o ~ u n i c a t i o n scapa~iiities c o ~ ~ ~ r ~ the i a llegal ~ y issues , related to a system must be evaluated at the organisation can address these early in the development process. power distribution systems requires the use of an effective to trmsniit control and data signals between control c e n ~ e sand a c o ~ u n i c a ~ i osystem n large number o f ~ e ~ o t elocated ~ y devices. Since there are a wide range of available c o m ~ ~ ~ i ~ ctae tc~h o~ on ~ ~ gcapable ies o f performing this task, selecting tbe appro~r~ate co~municationsystem requi~esa thorough understand~ngof the s ~ e n ~ and h s weaknesses of each com~unicationt ~ ~ o l o g ~resently, y. no single communication techno~ogyhas been de~onstrated as being best suited for all distribution auto~ationneeds. Each d i s ~ b u t i oautoma~iQn ~ scheme has ~ n i q u ec o ~ u n i c a ~ requirements, on and t h ~ r ~ f o ar e ~ o ~ m u n ~ c a t i ot encs ~ i q u efor d~stributionautomation must be chosen based on those unique r~quirements. e shared with others. Fo owes on a monitored hr

~~~~~t~~ eats The c o ~ m u n ~ c ~ t iroenq ~ ~ r e m efor ~ ~ distribution s automation depend on the size, complexity and d e ~ r e eof au~omationof the d ~ s ~ b u t i osystem. n In general, it is de§irab~e

ower System ~ e s ~ c ~ ~ n

nt and

re data rate r ~ q ~ i ~ e r n ~ ~ ~ s .

ishibution in a ~ ~ r e ~ l a t e ~

There is no i ~ ~ e r ~ ~ t

1

to the further dev ations. It can be

the isolate^ area.

tec~olo~ provides y near-instantaneous ~ n f o ~ a t i o n of a single household. It supports rapid, report netwo d to optimise network loading for reduced e~ergylosses in effec~, . ~ecause A facilities for the lower levels of the d ~ s ~ b u ~iiioe n~ o r k And ogy can monitor deviations from establishe f possib~et a ~ p e ~ n Variable g. rates can be f debts ( t ~ o u g flexible h p r e p a ~ e n can t ~ be e an empty b ~ i l ~ i nremote g, disco~ectioncan take place with complete c e ~ a i ~ ~ , ~ u ~ hnew e r tariffs can be quickly and easily p r o ~ ~ into ~ eany d ~ ~ s t 0 ~ eete r 9ers down the wires whenever required. The data can be ~ a n s f e ~ einte~activel~ d throu~hthe e l e c ~ ~dc~is~~ i b u t ~system o n to the dis~ibutor’so supplier and custo~er.~ i n a ~ c iapp~ications, al suc b e ~ ~ stores e n and finance organisations, can bec could use the existing e l e c ~ ~ c i ~ txnd supplier. As a result, no s ~ p ~ acarrier t e media. Two-way data ~ ~ s ~ i s s i o n not only on LV n e ~ o r ~ s ,

-

2233 that a circularly $7 linder in free space. And a e, by the early 1960s ther ion in what was then c

and ~~~k~ 124 ised that the losses ties. They ~ o a paper ~ e tlz into accoun~repe~tercosts [25]. This set the stage for the commercial devel and i n d i ~ a to t ~the ~ telepho~ecompanies in ~ a r t i c u ~the a ~way to b e n e ~ ~ c oratio. s t For a ~elephoneappiica~on,the economics are most two conditions:

e b e ~ ~ rep~aters e n should be maximised~ eat~rsare expen$ive nd the fewer the better. idth of the channel should be m a x ~ ~ i s eIn d . this way the ~ a x i ~ u be routed on a given channel, and the cost shared ~ ~ o ~n g~ s n y Because of these factors, the first widespread use of fiber optics for c o ~ u n ~ c a t i o was ns s. s the ~ong”distancetrunk lines of the telepho~ee o ~ p ~ i eThe to ~ ~ ~ r o v e ~ine the n t ~s e ~ o ~ a ofn the c efiber, to the exte rs are usable over large distances. At the same t h e , the cost to the point where it is co~parable,on a l e n ~ h - ~ o r ~ ~ or. The i ~ ~ ~ ~ cofa these ~ i odeve~opments n is that fo er optics can be used to replace copper trunk cables, ne optic~lcables could c

le of an ~ u ~ s ~ n~a ~ i c ~~ bm-way c Q ~ ~ u n i ~ a t i Q Innthis s . case fault detectors must c o ~ m ~ i c wi a~e

cisntrol centre so that the fault location cm be d e t ~ ~ ithen ~ e signals ~ , must

automation will be e ~ ~ a s to u ~adverse e

flashes, ~aultsor switc

~ ~q

~c

1

Free Space Free Space Optical Fiber

External Yes Yes

o f the ~ u r n obf us ~~

lation

I

The scheme uses local i n t e l ~ ~ g ~tonec ~~ a m i ~local e d ~ tot see~ if there is ~

~

h

roblern, b e c a u ~each ~ local set of nctions such as feeder d e ~ ~ ~ y r nse n t

for less than $3000 per km. i s ~ ~ ~§yst~m, ~ ~ twi~Q e r etr ~ energy is very close, ~~r~ is

n is r e q u ~ ~ dit , is usually n from Scratch, For example,

to add ~

o

w feeder monito

~

~

K

Dis~butionin a Deregu~ated

1

Q V any ~ Qb~tacles that might be p r e s e ~ t ~byd the c o ~ v ~ ~ t i o n a l media. Fiber allows the c Q ~ u n ~ c engineer a t i ~ ~to design a s ~ that will s meet ~ all~the ~ worst-case require ents, that can acce as many locations as n e c e s s a ~and can handle the

ee ~ e p a ~ a t i oofn bMsinesses: proposals and cons~tation.Office of Gas and E l e c ~ i c i ~

.N. Allan, ‘A~sessmen~ of customer outage costs due to ~ l e c ~ ~ c s e ~ ~ c e i ~ t ~r e~s ip~~e ~i t~isector’, a~~ s : IEE Proceedings - ~eneration,Transmi.~sion and I 1996, pp.163-170. N.Allan, ‘Eva~uationof reliability worth and value of lost load’, ZEE ~ r o c e e ~ i n g ~eneration, s ~ r a n s m i . ~ s ai on~~ ~ ~ s t r i b u t iVol. o n ,143, 1996, pp. 171-180, ‘ ~ ~ f o ~ aand t iIncentives o~ Project: Defining output measures and incentive regimes for PES dis~~ihutioi~ businesses Up~ate’,Office of Gas and E l c ~ ~~~~e~~ ~ c i (~O f g e ~ ) arc^ 2000. , ‘ ~ a ~ y sand i s evaluation of five s h o r t - t e ~load f o ~ c a ~ t i n g asactinns on Power Systems, vo1.4, No.$, October 1989, ~ ~ . ~ 4 8 4 - ~ 4 9 ~ P. Van Olinda, ‘ N o n p ~ a m e t r~ ec ~ e s s based ~o~~ ~ o ~ - ~ e r ons on Power Systems, Vo1.13, No.3, August 1998, ~

i. T.L.Lu, A. Abaye, M. Davis, and D.J. M ‘ANNSTL~:A neu~ai-network-basedelectric load forecasting system’, IEEE T r ~ n s a c ~ ~ on ons Neural Networks, Vo1.8, No.4, July 1997, pp.835-846. .L.Ring and R. Luck, ‘Very short term load forecasting a l g o ~ ~ ~E~~ s ’ , E l e c f ~ ~c ~ i l i orec casting in an Era of Deregulation Conference, Dallas, TX, November 1996. Wiktor Gharytoniuk and MO-Shing Chen, ‘Very short-term load ~ o r e c ~ s t using in~ a ~ ~ ~ c i a l ~ pp.263a ~ neural n e ~ o r k IEEE s ~ ~ Transact~onson Power Systems, Vol.15, No.1, ~ e b 2000, 268. ED Ad-Hoc W o r ~ Group ~ g 2, ‘Distribution Automation: hnctions and data’, CI

structuring and ~ e r e ~ l ~ ~ ~

IEG 61968 System Interfaces for ~istribution anagenient - Part 1: Interface A r c ~ i t e c ~and r e General R~quiremen~, IEG 1999. Cauley, Peter Hirsch, Ali Vojdani, Terry Saxton, and Frames C l e v e ~ ~ ‘dI, n ~ o ~ a ~ i o n n e ~ o sru~p ~ oopen ~ s access’, IEEE Camputer ~ ~ ~ l i c ain tPower, i o ~ 1996, ~ ~ . 1 2 - ~ 9 . [131 Peter Hirsch and Stephen Lee, ‘Security applications and a r c h ~ t e c ~ for r e an open mar~et’, E~ ~A p p l ~ c~a t i oin~Power, ~ July t1999, pp.2~~31. ~ r A ~ ~ and ~ ~ ai l l ki a mPremerlani, ‘ ata c o ~ u ~ i c a t y oin n sa d ~ ~ e ~ ~nvironrnen~’, u ~ a ~ ~ d ~ p ~ l i c u t i ain n sPower, July 1999, pp.36-39. 11188-3: I994 ~ f o ~ a ~~i eocnh n o ~ ~o n~t e ~ a ~ i o Stan nal ~ o ~ TJpper o nLayer equirernents Part 3: ~ ~ n i n iOSI al aha and W. P r ~ ~ e ~Object l a ~ ,Oriented Mode [I61 tions, Prentice Wall, 1998. est M ~ t ~ ~ o d o i o ~Setup, i e s , and Result ~ o c u ~ e n ~ a t i oEPRI n ’ ~ S~onsor~d hernet for Protection CO ersion 1.O,May 1997. EE ~ o r ~ i i Group ig on ution Auto~ation(Edited) Tutorial Course 8 8 ~ ~ 0 2 8 0 - 81988. -~~~,

-

~

Vo1.82, 1910, pp.4

e on Electric Utility Power Lines

Power System Restruc

eals to ~ h ~ s i c a~ansfers, 1 this risk i s e x ~ e ~ e hl y f ~ a n c i atools l that can be of help. of b i ~ a ~ contracts ra~ (and various other ~nancialdeals on the P faces not only an increase in o~era~ional d ~ f ~ c u l with ~~es ~ ~ as the market need far ore also a c o n ~ indplannin~ ission system can evolve. This has serious ences in ~ e ~ i as~ eviden~ed b ~ ~ by i ~recent system-wi~eblackouts. Tn the subse~uent~ections below, we present a p a ~ ~ c u l amarket r structure that equips the TP with ~ a r k e t ~ , b a s e ~ solutions to conducti~ as energy m a ~ k ewith ~ a large umber of bilateral ~ a ~ ~ a c t i o n s . to become actively involv allowing the TP t~ pupsue em can also be solved in an ~ f ~ c i eway n t as inten with the in~Qduction of compet~tion.

i s ~ s isi one e~ The e ~ ~ ~~ s ~~ system ~~ of~ the ~ most c complex c o n s ~ c t system§. to tbe e ~ t e ~ as lt ie~~ i from n ~ the opera~ionof the ~ a n s m i s ~ ~system, o n imple the market mechanisni to the ~ n d u re~uires s ~ a fair level of u n d e r s ~ n d ~o fnnot ~ cial and r e ~ u l a t oaspects ~ but also the ~ n g i n e ~ r iconsequences ~g of ~ i ~ u5.1 r e shows the evo~utionof the role of the TP in the industry (as at the time of writing).

~ ~ o ~ofuthe~role i of o tbe ~ TP

In the dependent phase the TP functions as a part of the vertically integrated utility. In the ~ u phase ~ the TP ~ stands ~ alone Y and ~ oversees overall market activities, The ark et pa~icipantsare ~equiredto submit their intended use of the system to the TP and based on that i n f o ~ a t i o nthe TF allocates transmission capacities foIlowing strict ru'les set e TP assumes no ~ n a n c ~respo~sibilities al and has mi~imalinterac~~Q

ants. As shown in Figure 5.1 there are three differ nt s ~ c ~ rofe the s TP phase the TP ~ a ~ c i p a t eins every ~ ~ of amarket ~ e nction§ of the TP in this phase can be c a t e ~ o ~ s easd that of marke~m s e ~ i ~~r oev ~ d eOf r . ~ e s two e only the function of market As a service ~ r o v ~ dthe e r TP assumes full financial l i ~ b i l ~ ~ We will disc~ssthe role of the TP in each phase in deta

the TP exists only as a part of a v e ~ i c a ~~nte~rated ~y u ~ i l iThe ~ ~ ity owns and operates a ~ ~ ~ 5 i d e r ~ ~ ~

~

e utility is a an teed to ~ e ~ o v e ning of the system by lem of short-term generation sc n to b ~ l a ~load ~ edemand devi~tio and to do this at the lowes 0 of this problem ~ is given ~ in [ 11:~ ~ ~

~

0

~

I57

Trans~issionExpansion in the New E n v i r o ~ e n t

where e a m o u ~ot ~ i n s ~ l l generation ed capacity at node i and tec the ~ ~ o uofn ~t n s t a transm~ssion ~~e~ capacity for line 1

,? : 1;'

a

the rate of i n v e s ~ e nin t generation capacity using t e c ~ o ~ o (gI y

1 fa

( i ) ,( i~) ,I~) :~the cost of invest men^ using techno~ogya at node i

: the cost i n v e s ~ e in n ~line I

logy a at node i, exc~u

,

(t),PL~ t ) )the : flow on line 1 as a function of system ) : the ~ a x i m ~a~~owable m flow on line 1 as a function of the a ~ o u nof t ~ns~~le ~ t s <<~K, ~ i s § ~ co anp a c ~owing ~ ; to secure c o n s ~ ~&max rresponding cons~ain~§. The o p t ~ ~ ~ ~p ~ e r§~ oa dtT~~ino ~~ r o b l (%I)* e ~ is the longer of two time ~ n ~ e ~ov aIs lion or ~ ~ s ~ i i s~ sv e~ ~o~ ~earen ~valued. § As the syste ides the level of production and the rate of investment on ~ e ~ e ~ a ~ ~ and tra~sm~ssion, P,Jt), d a r serve as control variables in this fo for the status of the system o ~ ~ r a tcm i ~ be n accw

ciated by ~ x a ~ ~ i nthese i n g variables.

This ~ o ~ ~captures ~ ~many t iwell-&own o ~ trade-offs relevant for the e ~ ~ c of~the~ n c the r e ~ a t ~ ~ n sbemeen hlp the i n v e s t ~ timing e ~ ~ and the balance o f the costs er time, the value of different ~ e c ~ o l o g i at es produce powerl and C o i ~ ~ ~ e m e no ~~a~K e~ ~ ~e capacity r a t i ~and ~~ There are two n ~ ~ ~ c features e a b ~~~o n s i ~ ethe ~ i operation ~g by the TP (as a p a t o f the v e ~ c ia~ t~e g~ ~~~utility) t e d as the problem: the apparent compiexity of the problem (5.1) and the G on ~ ~ based on ~ costs e .,ag ~ e,,',, Owing ~ 6J,Tand to ~the ~ o n~ ~ lthe~ x j t~o~ s o ~ u t i to ~ nthe p r o b l e ~is not readily available, and thus the actual o p ~ ~ aand m are performed s u b ~ p ~ i min~ lmany ~ y cases. Further, since the rate nt is d e ~ ~ r based ~ ~ on n ecosts, ~ the o p t ~condition ~ ~ ~of ithe~ fom limited to concern [fa 1: and e:(t). Nevertheless, problem (5.1) is B ark in studying the e f ~ c i oef the ~ ~industry ~ as the r e s ~ c takes ~ ~place. r ~ ~

structuring and Dere

15

ln the passive phase the TP exists as final authority in dist~~bution sectors act~v~ties s e ~ ~ afiom t e the ~eneration market env~ronmen~. A newly cre , manages the system in order to ens cai~iedout by the TP are tailor^

, both existing and eveloping, is highly ~ o n - ~ i f o ~ . reg~onalcharacteri§~icssome markets admit cen~alise for whoIesa~etrad~ngand a r e a ~ - t i ~ ene e ne or two centralised markets and still a~icipantswith no centra~isedmarke e USA can be represented by one multilateral transaction model, the mandatory systena operatQr model and as shown in Figure 5.1. model is based on bi~ateral~ ~ s a c t i o namong $ arke et the proposed structure of the Midw n model. The model consists of transactions. Firstly, ~ d i v i d ~buyers al and sellers make bil ~ i t h disclosing o ~ ~ the i ~ ~ l e m ~ ~ The t i o ~ . ~ h ~ t h or e rnot to allo constra~ts. If the proposed transactions do not violate any cons~aints,then they are without any modi~cations. This is the most esired case. If the ns result in violation of co~straints,then the T accepts none or a p ansactions and suggests necessary mo ion called ‘loading vector’ [2]. a new set of trades to satisfy the limits. Figure 5.2 shows the interaction among various market p ~ ~ c i pfor ~ the t s ~odel. In this model, the function of the TP is limited to ~ e w ~ ~e ~roposed ~ e~ ~ ~ a n~s a c ~ ~ go n s will result in violation of system limits.

~ ~ i l a t ~ansaction e~a~ model

Tr~smis$ionExpansion in the New ~nvironment

Mandatory system operator model

~ i t i a ~m l y~~k ept a ~ i c i p bid ~ ~ supply ~ s curves to the TP, although a ~ the rest of the be made to include elastic d e ~ a n din the f o ~ u l a t i o nfor the consumers' demand is inelastic since not much is lost in term the chapter^ The TP then simul~neo~$ly dispatches generators an capacity using an optimal power flow p r o g ~which ~ , determines the most ~conomica~ mix of g ~ n e ~ t i ofor n s given load. The voluntary system operator model ports a m~1t~"~iered s ~ c ~ that r e min~misesthe TP's ~ ~ ~ u e on n c profits e by m acc~ptablelevels of reliability. Figure 5.4 shows the basic schematic o model.

UlU

Y ~ o ~ usystem ~ ~ operator r y model

bilat~raland c e ~ ~ a l i s e r e s e ~ cof~ spot market ~ansact~ons is desired b a l a ~ ~o fe i ~ s ~ n t ~ supply e ~ u swith u i n d ~ swhile ~ , direct access and custo~er

~ a n d a t osystem ~ operator model lead to an e q u i l ~ b ~ usolution m of the f o ~ l o woi ~ ~ t i ~ ~ s a~ tr io~bni e ~ :

(5.1 1) i,o

i

The ~ ~ ~ ~ ~ defined i $ a tini (5.8) o ~ . The result of solving the (5.12)

~ r ~ § ~ i ~Expansion § i o n in the New Environment

tive economic e n t in ~ ~ e energy market before such a prw

tion sector so that any r e ~ u l a t o ~ pendence based on the reasons given. of the ~ s m i s f ~ c i ~opera~iQn nt than ~ u l t i p lgrids ~ serving c to the high degree of the econ

ng a wide area o f the

v ~ ~~ aon s~a ~st i owhich ns c as well as system users ben tamers w h o s ~~ e n e r a ~ i o ~ t a m Q u ~of t syste~-w For instance, it is easy v ~ o loads ~ ~ issfar s ~ a l l e r

on of the TP, especially at the time of scarci ent to say that the s u ~ ~ e s s ss work9 so that the tm

141

1

Power System Restruc

TP ta ~ d d .Therefore, the ~ a n § ~ ~ § § i level the rate of re^^ reg~la~iQn, m o ~ ~the l , TP sets the bundled energy and tran§mission price that ~ ~ n ~ ~thei so e § ~ ~ i sthe ~ system i n ~ load at each given instant. The ~ a n s ~ ~ s srevenue iQn i first cut is specifie to consumers and to sup and the c o ~ ~ u t e u§age charges a § s i ~ e d In the ~ o ~ u n system t a ~ operator model the

t market for the rest.

market and is subject to ng the ~ e c e s ~§ ~ s ~ s as a service provider, As a service p e charged to each bilate

§

s

~

under strict r e g ~ a ~ othere n , is

act more custQ~ers.

ssion line in (5.9, For ~ ~ a r n ~b l e , etter control d e ~ the ~ ~1TC n can e

c o ~ i ~ i ton any g major transmiss~onprQjec~ the

e 5.1. In the wti

~ i s s ~ o$ yn $ t e cong~~tion ~ only when th e ~ ~ cost gof the ~investment. ~ Asa

~

1

structuring and Dere

ystern because of the ineffici high, some users will choose

EQ

total cost of ~ i s ~ b ~ t ~ d

TP and the TTC lies in the clus us ion of the ~ ~ n ~ ~ n v e s ~ $inn the t s h o ~ - ~ ime-scale ~ n c t i o n sbecom C in i ~ p l ~ ~bilatera $ n ~ i ~ ~ es are create^ BS part of

maker.

5.3.I

I ~ c e ~ RGal ~ ~te Des ve

ows and thus r

Transmissioii Expansion in the New ~ n v i r o ~ ~ ~ t

lator of the ITC’s profit

ITC for a higher ~ ~ c in~efficiency. ~ a ~ e

The ark et- base^ usage charges are commonly referred to as conge§t~o~ charges, The zonal pricing met~odsare two widely used metho~s s. The nodal pricing method computes the ~ansm~§sio (5.8). For a given time instant 1, the problem gian function of the form

L'

(5.13)

i

where ,u,f 0 if and only if FXP,, PL) = F;"'". For si u t ~ gthe flows on each line in the system. ix notation are written as

icity, we use DG power Raw in e DC power flow equations in (5.14)

where &isthe voltage angle vector, Taking the first derivative of L with respect to Pi,, and it equal to zero yields 4-

&(t)= A ( t ) t

i

(5.15)

I

Suppo§e the generation cost of supplier i, ciPa,is a quadratic function ofthe ou (5.16) Then, under the perfectly competitive market condition, the optimal s ~ p p l y i, b,,, is the marginal cost bid given by

(5.17) ~ a t c h the ~ gsolution in (5.15) and the supply bid in (5.17) the system price at node i, p, and the dispatch amount, P,r,,as

(5.18) Finally, the ~ a ~ s ~ i srate s i iso set ~ by the difference in the ,q, i.e. =,q -4. The zonal pricing method consists of two steps: (I) aggregation of i n d i v i d ~nodes ~~ into zones and (2) compu~tionof zonal prices. The system is first divide^ i of smaller markets by a ~ g r e g a t i ~individual g nodes into zones w ~ e n ectat~Q of~c ~ n g e s ~ within ~ o n each market. The ~ a n s m i s s ~ rate a n is so~vinga similar optlmisation prQblem as given in (5.8); the cost C,,@'~,~) n the average cost of generation in zone i. The line Row constraints are now ~ i t e ~ a flow c e limit c ~ n s ~ a i ni,e. ~ sthe , ower flow on any line I along on^^ the ~ ~ ~ g e s t

Transmission Expansion in the Mew Environment

interfaces is within the m imum rating o f the line. The transmission rate is ,U] resent zones rat he^ than nodes. ophistication may be requ~redin order to i ~ p ~ e m e ective, a sign^^^^^ reduction in computat~Qn rice~capr e ~ ~ a t i osince n only a small rather than many nodal prices as is ricing. Further, there is a greater advanta~eto be gained in ~ m p l e m e ~ i ~ d su~sequentsection. ccQm~odat~ng b~latera~ ~ a d as~issi ~ ~ u s ~ aint ethe The a6cess fees are intended to recover the fixed part of the costs and are thus ~ndependentof actual usage. However, usage-ind~penden~ charging for the access fees is impractical and may result in improp~rincentives for the ETC. In order to s t ~ ~ ua l ~ t ~ m e a n i n g ~ ~l h m e c h~~ i § m~ some , measure ~ gof base-load capacity needs to be ~ v e n . practical approach is to compute the access charges based on a coincidental peak The ‘12-CP’ method [3] i s one such approach. The p o ~ i o nof o foloads. nsu~p~~ n a1 access fees is computed as

i

where S,(t> is the load i’s share of system coincident peak, and LXt} is the load in month I at peak load~ngcon~itionof each day. As the total r ~ v e n u charge is equal to the product of access charge and the coincidental peak o approach provides the ITC with incentives to increase individual base-load ca heref fore, price-cap r e ~ ~ a t i oand n the rate design consisting of es and regulator-approved access fees o ver the ~ v e s ~ ewith n t some incen~iv~s for improvement in ~ ~ c i e n c y , H o ~ e v e r 9the resul~ingrate structure does not i ~ e ~ ~ a tyield e l y prop~rincen~~ves for transmis§ion ex~anding.In the subsequent section, a market mechanism called the pr~ority is d ~ s c u s s ein~terns of compl~mentingprice-cap regulation in o i ~ s u r a ~ service ce ~rovidethe right set o f incent~vesto e ~ a n c the e t r ~ s m i ~ s i so yn s ~ e ~ ,

~ v ~ in The driving forces of dere~lationaim to establish a more c o m p e t ~ ~market achieve lower rates for consumers and higher ~ f ~ c i e for n c suppliers. ~ T~ough trades, consume^ can establi§h various service ntracts with any supplier in order to obtain the lowest rate and most desirable service. lateral contracts s p e c i ~ i the n~a~ount of power9 the time and duration of the servic d the associated rate and compensation are n ed and agreed upon between the suppliers and 6ons~mer tition is directly related to the bilateral trades which allow e, the success of the market is dependent on the ETG’s abi

Since the t r a n ~ m i s s ~grid o ~ is a physical system, the 1TC is able to honour and e x ~ ~ ~ I ~ these bilateral c o n ~ a cas~ far as the system design and opera tin^ condit~on§ Unlike in the spot m 1: the ITG is not allowed to pa~icipatedirectly in re~dispatchin Thus, the ITG relieves ~ ~ n s m ~ 1 trades or by 6reating counter ~ Q W Sin

Power System R e s ~ c ~ and ~ nDeregulation g

168

systematically adjusting the rate structure. All bilateral on^^^^ are the replacement resources in case of ~ n t e ~ p ~ die o n to s either the contingencies, ent or ~e~erator-re~ated W ~ t ~loss o ~oft genemlity all bilateral contracts consist ofthe fofbwing s q ~ n t ~oft ye ~ e r g ytransfer ~~g injection point and the c o ~ e s p o n zone

J’,z : the w i ~ h ~ point a w ~and ~ the c o ~ e ~ p o n d zone ~ng enalty payment by generator i for generator~relatedcontin ante.

e ~ a t purc~ased ~o~ in the spot market of zon

ovcr the p ~ r ~ ot dE T = it,, 4. c ~ n g ~ s t ithe ~ nload , is respons

ex ante for each time the tr

robability of be in^ c u ~ a ~ l ewith d the upper bound g~venas ~~~.

In r e a l - ~ i ~oper~tion~ e the ITG determi relieve ~ a n ~ ~ i s §con~estion ion along with

s creates an a ~ a c t i v eincentive far as long as the market can take it. ss~on~ y s t as e ~a subs~ntialeffort by the ITC is expected in order to i T base for priority insurance service. The advantage of this method is tha of the spot arke et^ which is under strict regulation, the w~liin~ness to take the ITC’s profit is well capped, Over time, the sho the spot market or bi~a~eral trade is e x ~ e c ~ etodlev meet the c ~ ~ g needs i n ~of the market. The I t tn mark~tevolution is likely to have a ~ e ~ a ~ v~e l y g level~ priority insurance service and to enjoy profits from b u s ~ e s § we discuss the effect of reduced r e g u l a t o ~unce insurance services on transmiss~one x p ~ s i o n .

5.3.3

T r a ~ ~ ~ ~i s~ sp ia~n~s i ~ ~

The ~ e w ~ a r k eorg ~ e s c ~ ~ ~ate the d ~ e g i ~ of ~ ~thn g ~ ~ n d a n ~ $s ~e t~a~l for n g s ~ s t e m a t~i ~~ s r n ~ §expansion. s ~ o n This ta Qf ~ s e dITG § ~ c ~ r A e .fo~ard-look in^ ~ a n s ~ s s i o n S t of its c u s ~ o ~ ebased r § on ~re¶uencyand r n a ~ i ~ of de its ~ a ~ ~ ~ ~system, s s i o n tools n e c e s s ~ yfor e rovides a basis for su e n ~ a ~ c and e ~ ex~ ~ t of these ~ n h ~ c e ~ e ~ t

in

e

elopment of new market tools for operating the ~ansm~ssion s y s t e ~b as the ITC moves into the active phase of m ~ a g e ~ e nInt this ~ phase to make complex business decisions over a wide r a n ~ eof time scales: long-te deal with ex~ansi d to ~omputingthe i c o ~ s ~ a i nand t s makin~system reinforceme~tsin order to meet this roposed ~ a n s ~ ~ s rate s ~ odesign n that the i n v e s ~ e cost ~ ~ tis on rent on the spot market owing to the r i n v e s t ~ e ndecision ~ is required for the Even when there is s ~ ~ i ~c oc na~ ~~s tti osustain n needed for relieving this congestion is ing the performance of the ITC, since rate lies wit^ the r e ~ i a t o and r not the ITC. The a ed prioriv insurance s

on the h i s t ~ ~ i c pa l a ~ ~ of e ~users s ority insurance services nt in a new efficie~tgen t for i ~ p ~ e m e nbilateral t ~ g rrad The better the projection that the new market the ITC makes and subs The s h o ~ ~dec~sions t e ~ deal with the most ~ i f ~ task c ufor ~ the ~ ITC s on the ITC's ability to funct three aspects to consider in the pr~cing.The first is ~ e ~ ~ i n g s from the long-term m v~l~m is ei r n ~ o ~ in ~ nmaking t investrne proj~ctiona€ the lo~ationaland temp~ral n, Over time, the market f the costs by ~ x ~ a ~ ~ l a t ~

s the vahation of insurance services given the s p e c ~ ~ c a t i o nasb~~f ~ a t ~ r a l d in Section 5.3. The ITC has a menu o f p r In this ~ o ~ u l a ~the o nre~iability , is explic i n ~ ~ ~ ~ p xq t i by o nthc s ITC over the contra valuation and may be solved us spect of pricing is ~ e l a t i ~the g activities. Because the mount of ~ c t ~ bcl e, as ~ well as the rices at the spot ~ ~ kthere e ~ ,

es

Transmission Qpen Access

173

e either set or ~egulatedin ac

a

~ r e a s o ~ ~ borI e‘fair’ q prooft ~ a r gabove i ~ its cost. r e ~ u to~ operate r ~ in a manner that m i n ~ m i s ~odv ~ r a rl ~e v e ~ u ~

~~~e~c

~ the ~~~~~~0~ ~ ~ to serve~was often s m concept was usually r e p ~ a c eby ~ g

174 -----._.__

-

Power System Restructuring and Deregulation _ _ I

~ove~ment-owned e ~ e ~indus~y ~ c i ~ n~ c o u r a ~ e es not have to be part of a r e s ~ c ~ r i effort. ng I early 3990s several Western governments were e rof ~ n n i n gthe power ~ n d u s t ~ , do a ~ e ~job r e d u c t i ~in~ l a b o ~ rcould be achieved by p ise rates and have a greater interest in e l i n ~ ~ power ~a~~~g lace i n e ~ ~ c ~ e n cIn~ eother s . countries either Qwnersh~por to cooperative or to p r i v a ~o~r ~ ~ ~ s a tori oto~new s ~ types of ions or ~uasi-gov~rn~ental entities ts, ~ w n e r s hand ~ ~~nctionalre5 1 invo~~enient of private capit d as private sector p ~ ~ ~ c i p ut b e c o ~ more e ~ol~~ile. also a serious conb: . Compe~i~on breeds imov ism. A competitive powe use of new technologi ~ o n o ~ o sl cy h e ~ ewas unable to ~rovidei n c e n ~ tiva~ionto use new ideas and ~ e c ~ o l o g ~ e ework. Lack of compe~~tion also gave in c~untriessuch as India and Chin very low. A more b : o ~ ~ e r cethos i a ~ could be h ~u~to~~rs.

.1.3

~ n ~ ~ n d~~ ei nng e ~T ~~ at ni s~~ ~ihznd ~~s is^^^^^^^^ ~~n

the variatio~sdis~ussedab amms in certain respects: (I) s e l e c t ~ oof~~ n e r sources, ~y rangi cost r ~ s o u r ~to~others s with low capital and hi econo~iesof scale or n ~ ~ monopo~y r a ~ fea~res~ electric ~ o t e n are t ~ not ~ ~an~ i ~ p ~ to~ ~ e n ~ clear e c o n Q ~ i eof~ scale, but there is a ref fore some form ofregula her unbundled into (a) a wires ides facilities fox e l e ~ ~d c ~i ~ ~andi (b)v ~ ~ ~ lectric e n e r ~to end consu~ers,(3). The tr in the ~ c o n o ~the ~ cgeographic , and the b c ~ h ~ r e f o must r e ~ o n ~ to~ f~~u~ ei c ~asoan n i n ~ e ~ r aand t $ ~re dion in the gen~rationand ret mission system and ensure e q ~ ~ t a bbasis ~ e to all power ~ ~ ~theofocus ~ ofe attention s field’, and the rules for managing access by ~ ~ s ~ r i m This i ~ acha~ter t ~ ~ focuses . on this c ~ subject, ~ ~ ~ a

~

nal, state, provincial or independent generators coexist. In these cases the ~ e c h n i inte~e~ationshi~s c~ are murky and are in a process of rapid ev~lution.

~

~

rate and transfer BOT; {or build, operate and own) plant of ing, ~ s u a ~ I~ya t ~ o n a~l , ~ ~ i t~i Iea€yan on in many f a s t - ~ ~ nsystems. g T r a c~ ~e e~are~ eoften e ~in force ~ as an economic incentive to investors.

n s i b i i i ~on the dis~butionside as in a ~ a ~ i t i o ns a ~ ~ ~ r e g u l is a t that ~ ~ Discos ~ may now be restrict scos a d ~ ~ o ~ ~ l j ed~ ~~ nt sales ~ e~ e ~ o countries p ~ ~ is~to sell to an ~nvestor~ or to coxp em so that ~ ~ v e s ~ m for e n t reinforceme~tcan be o p ~ r ~ tprac~ices i~g ~ m ~ ~ e ~ e ~ ~ e ~ .

~

~

~

netwo was state owned before r e s ~ ~ ~ o~ n g , Where the trans~~ssion i n t ~ g will r ~ ~be r ~ ~ n and e d a d ~ s ~ i n e ~between on owner and operator is r SA, f o ~ e er ~ e c ~ ~t c ~may ~ sell i off ~ ~ their~ other s assets ~ Q r n h~ basic ~ ~pr~mise ~ s ~o f open ~ a n s ~ ~ ~ ~ccess s i o n i s that t r a n s ~ i s s ~operat~rs o~ treat lalt users on a non-discriminato~basis i use of services. This re~uirementcannot be ensured if transmission y generation or supply. A requirement, therefore, is ~~~~~~~~e~~ $~ste~ at^^ to Q ~ e r athe t ~t r ~ s ~ ~ ssystem, si~n

markets are e s s ~ n ~ a lsl yh o ~ - t forward e~ ~arkeEsi serve toad and loads bid for the opportu c ~ two ~ ~ s t i n ~c ~~~ k eday ~~ sahead : ~ t ~ e ~ ~of s24~s ~~~ a r~adt se ~ u a ~~ ~i t~~ o n ~ , ansmission costs, constraint management or congestio bidders may bid. as a single generator or as a ~ o ~ f ~ofl ~i oe n ~ ~hQwe~er, ~ t Q a~unit ; may Iy one p ~ t h o ~ a~~hj d~ ~~~y e r ~s ~m o ~b l ~~ i~ ~ ~ ~~ebids.~ ~ f n there is conges~i5nthe CalPX sets a ~ a x ~ price ~ u m and buyers may su rnand bids for ~ Q n - c ~ i loads; ~ a b ~ ~ auction ~ o ~ a single d ~i c ~ ~ d~~~~~~~~ ~ a ~ ~This e g occas~on , p r o v i ~ ~a sc ~ ~ ~ ~ t i t i ~ ~ g~n$ra~ors ta adjust their day~aheadsched~Iesin the lig s ~ ~load~forecasts - ~md ~unit status. ~ n

as well as the o ~ i ~ r s h la the new ~ ~ k

151.

e

~

o

Transmission Open Access

1 the new ~ a d i n g§ ~ c ~ rinto e s a few alte

ion system is ~ ~ n ~ a ~ l y

I

a

T r a ~ s ~ i s s ~Open o n Access

7

1

ower System Restruc

lation

ostage stamp ra~e,and ~ h e are ~ e no i ~ ~ e ~charges ~ ~ ~if npower a l is ~ in the ~ n g e s ~ zone e d when con E" usem to sei1 or buy ~~~~sare a d j ~ ~ ~ b

~

s

the flow ~ a in a ~ n e ~e owill r ~ de s a c ~ o at ~ sv ~ times ~. ~ cb line I;a meet the n, the load of tine j

p

~

o

~

~

~

s

For each transaction, the rel~ab~Iity of ~ ~ ~ s m i swith s ~ oalln lines but linej in $ e ~ ~ care e , first calculate^. The r e ~ i a b i bl ~e n~e ~RJ:Ti t ~ a n § ~ c T, ~ ~isoc~lculated n as the i n ~ r c ~ e in n t ~ a n s ~ fai~ure ~ ~ i o ~ a ~ s e n of c ~the line. Similar ~ r o ~ ~ ~ are ~ ~c ia t~i ce~s ~ina respect te~ o ect to line j . The e ~ b e d d e dcost of the ~ r ~ s r n i s line s ~ jo, ~all based on ~eliabilityb e n e ~ oniy ~ s is now d e ~ n e das:

where P, is rRe m a ~ n i ~ od fe~ n s a c ~ ~ oand n E;. is the same as in (6.1).The s ~ e c i ~ c is in itself not c e n ~ ato~this r n e ~ ~ oused d to c o ~ p u t ep r o ~ a b i l of i ~ ~ansaction ction may simply be taken as example, the failure probability o f ility that a path will not exist from the t r ~ s a ~ ~si eo~nd ~ n This ~ ~ p r o a ctoh failure c ~ l c u l a t iis~ ~r e a ~ ~ l ~ ut becomes c o m ~ ~ ~ c aand t e d v i ~ a l unworkable ~y i e ~ ~ r circuits, ~ e d a cut set eth hod [Is] or B ~ o ~ d ~ prt ~ o ~ a ~ le for ca~culat~ng this a v a ~ l a b ~ ~ i ~ .

efits in a ratio which has to on the j u ~ ~ e m eofn the ~ ~ ~ ~ s r np i~ as n~n ~~r ,~ n specified e x ~ ~ e n ~ and u § ~depends y ace, the c o r n ~ o § charge ~ ~ e is set at

a i

1

It § ~ ~ ubel dnoted that e ~ b ~ ~ ~d ae ~d § ~ ~ costs i s s iRave o ~ to ~ n c ~ u d~ea c ~ ~ isue ~ie§ ers, § ~ ~ s ~ a tei o~ nu i p m switch ~ ~ t ~ear and shun~§eriesc acitors in additio~to ~ a n s ~ i s s i olines n and cables. and d~~ferent m ~ ~ ~ rno d s ed

us^^ ~ i f ~ e r ern nt ~ t h hence, o~~~

ewices, as i o the case in s d with time of use, g ~ o g ~ p h ~uantj~ative ~ c ~ ~ , and customers ~ c ~v ~i ~~a ct i o in n s~ ~ n d .

rovide~f17] in this A brief ~ ~ s of the c ~ansm~ssion ~ ~ ~p r ~~c ~ nin~ g the nNCC, subsection. to cover the costs of its assets but the ~ o s t sof The NCC i s altow cO~§~m~~§~ is the price quoted by e nlo§~e ~ p ~ n s ~ ~ dis~atchduring each half-hourly time slot w ~ e n generator which is ac - simple u n c o ~ s t r a ~ e dispatch, d transm~ss~on constraints dc ~f“lassof ~ load p~ r o b~ a ~ ~i ~ ~ ~~ 9 ~ is c a ~ c ~ i l aby t ~ cd o ~ n ~ ~eixn~ge c t ~d d e ~ with ~ the d c a ~ a expected c ~ ~ to be a order to encourage capcity offers from generators the pool purchase pmce y ” w ~ IF ~gh~~d ~ ~ by a u~g ~ e n ~t ~ nSg ~ with ~this ~ ~e ~ b ad b ~ ~ ~ t ~verage. ~ s ~ e n~ ~ ~is~a alseady t o ~ s e tl e ~ ~ ~eind the u ~ c o ~ s ~~ i as ~ anbut t ~c preve ~ enerating owing to € u ~ n § ~ i s sconstraints io~ or other factors the generator , This p a ~ e n consists t o f the ~ ~ ~ ~ $ s e n c e n compe~lsato~ pay

hand, same ~ ~ c o n goe n~e r~a ~cowho ~ are n called u ~ o nto g ~ ~ ~~swa ~~to n eg~ a n s r n i $ ~ ~ o ‘bid’ price, which is h~gherthan the puev a . This is, e~ect~vely, of-merit g e ~ e ~ a t o~pera~ion. r Constrain ts, out-of-merit paym ges, ~ ~ n s ~ i losses, s ~ ~start o nup ~ ~ s ~ s and ancillary sewices c~argesare passed on to c o n s ~ ~in~ the e ~~~~~. s The ~ ~ s t o m side e u of the market is simpler: a11 energy is puscha$ed at the pool selling SP), All of the extra costs of energy above the PPP are simply 1 EzjY and s ~ oversall ~ ~ is by ~ ~ u s t o t~ e~ ~o~the ~ calcar. g ~ h a l ~ - ~coo ~ s~ul ~~e r ~ c e ~ ~

§

~

~

~

~

~

+

~

~

l

~ ~ c t in^^ l y t ~ e ~ e ~the o ~PSP e , is fixed on a r e v e n ~r ~e ~c o ~ ~ basis; ~ ~ t~~ a is, t~ Po ~ the e l e c ~ sold ~ c is~ made ~ equal to payment to generators p ~ other ~ s 60 ce, the costs o ~ t r ~ s m i s slosses i o ~ ~are also rolled into

poiiy p o § ~ ~albeit ~ o ~regu~a~ed, , the ta ~ r o v i ~aaequat~ e srnission access are ~ o ~ s ~ ~Fur e rthis s . ~ea§onmd since average ~ 1 ~ ~ c a t of ~ Qupl ns ~h~~~~to d ~ ~ users, e ~ ~ 11% n been ~ the prima^ focus of con share the costs of o ~ e ~ a the t ~ system g and bear not but an ~ ~ e ofr the~ OS& g ~i n c ~ by ~ aall r is ~ ~ - e these x ~r o b~ l ~~and ~ s ~r ~ n~ ~~ n t ~ y ~~~~~

l

2

T~ansmissionOpen Access

is ap~roachcan work in a market s ~ c where ~ ethe organ~sa~on e for pro cur in^ a n ~ i l l aservices ~ is also respoxisible for opera~~ng the e n e r ~ amples of this include New Zealand and New England. llary s e ~ ~ c separate es~ from the energy m ~ ~ eist ,b ~ ~ e r n l ~ s ~ ~~ ~cwhere t ~ the ~ IS0 s is se~aratefrom the PX, as is the case i ~ c services i which~the ~ a ~ l ~ f o r nIS0 i~a is responsible for p r Q c ~ ~ incl ng inning reserves, AGC, replacement reserves, voltage supp ompet~tiveauc~ons start. The first four services can be procured by the ESQ through d to ~ a n d a ~or # ~ or be s e ~ ~ - ~ r o by ~ i ~users, e d ~ ~ w e s~lf~provision ~ e r ~ is li c o ~ ~ c~ ~~ na gl e m kThe e ~use ~ ~of”all. anci~~ary services, incfud the exclus~ver e s p o n s ~ b of ~ ~the i ~ TSQ. ~ l a c k ~ §capabili~ ta~ and reac must be provided or purchased by tbe ISO. In some e ~ e c ~ markets, c ~ t ~ both the mandatory and m a r k e ~ ~ b a ~approach~s ed are . For e ~ ~ ~in Spain, l e , the ~ Q ~ c a~ ~ l e ~~ reserves~ ~ service ~ ism~ ~a ~for ~ ~ ~ rs, All plant must be ~ ~ u i with p p a ~governor ~ and there is no r ~ ~ ~ e r ~ of frequency deviations ~ ~ ~o~n g associated. This is intended to reduce e the required primary r e s e it~ m~ ~ s t power i ~ ~ a ~ a n cIfeas generator . can procure~entof AGC service is from other ge~~era~ors. On the othke ~ o ~ a ~ c t~m ~ o~n ~ p~ e t ~ ~ ~ ~ ~ ~ ~

.7 ~ o n g e s ~ i o[21] n is not a new prob~emin power s y ~ o~p ~e r ~~ t ~and o n was a r o ~ t ~ e ~rob~em for the ~ y § t eo ~p e r a ~ oin ~ the t r a d i ~ o nsystem. ~ In e n v ~ r o ~ ~ ehowever, n ~ s , pre~~ously established practices for dea~ingw no longer be relied on since coo~erationbetween market p ~ i ~ i pcannot ~ t s be ~ a ~ n ~ e e d . Any control ~ e a s u r e sadopted by the system o~eratorta eliminate c o n g ~ s t i omust ~ not ody be t e c ~ ~ ~u s ~t ~a~ but a~ b also ~ 1 ~~be ~ fair to users and c o ~ e r c~ia n~ s ~p ~ ~ some e ~ ~ ~markets i c with i ~bilateral and multi~a~era~ c o n ~ a c~t ~ s ~ c t ithis o nps ~ o b is~ e ~ more d ~ f ~ c utol tsolve since these contract transactions introduce additional c~nstra~nts on the system o ~ e ~ t o For r . example, c u ~ a i l ~of~ an ~~ilateral ~ ~ s a c t i o~equires n s i ~ u l ~ e o and u s equal reduction at the entry and exit points, All this makes CO ~ a n a ~ e ~ae ~nhta l l e n g ~p ~r ol ~b ~ eand ~ ~equiresa comb~ationof ~ o pri ~ ~ c o p ~ ~ ~ respo~se$. o n a ~ It i s perhaps the t h o r n ~issue ~ ~ t in t r ~ ~ ~ i sope~ation. s~on

Boot md c o ~ ~ amodels ct are s e ~ ~ r a t addressed e~y first and an app~oachto reconcile h t h ma~agement issues without C o n s ~ d e ~ ~ tof ~Qn models i s then explored. Conges~i~n c o ~ ~ n g e n c y / s e c uproblems ri~ are discussed in this subsection. A fbller ~ e a ~ ofe n ~ topics can be found in [22], tion given beIow assumes ~ ~ c e - b a s edispatch d built on spot p ~ theory ~ [I I] and in its simplest terms, ~ e g l e c t i nprice ~ elasticity effects md the signific location, the dispatch a ~ g omay ~ be t ~stated ~ as:

~

subject to:

ere i andj are the set of ~ r o d ~ cand e~~ s

~ c G and ~ D atheir ~ ~ ~s~ e ~n es r ~ ~ ~~ c o n s u ~ ~ and ~ ~ ~ o rn ~o offer ~ (bid) ~ cprise ~ and ~ ~ r c h~a s ~e ~ un t ~~ ~~ i ~~ we g~venby C and B, respectively. Tltc single load balance constraint will later be ~ e ~ ~ ~ a ltoi sa eset d of a ~ g ~ ~ nload t e dflow equations. L is a ~ ~ s m i s s loss ~ o n~ ~ ~ t i o n , Gi.ma g ~ n ~ a tioer a ~ and a 2, ~ the ~ ktfi ~ o~erat~ng c o ~ s ~ ~~ n~t .o b(6.2) l e leads ~ to the s o l u t ~ and o ~ ~ ~ ~ ~e o n- d ~Tt i o~~ ~ :c ~ e ~

Mi

af y ~ i a bon l ~the c nt above is the $tipu~ation ty. The demand-price elasticity issue, which was

I [25-271 is d e s i g ~ ~ aas t e power ~ ~ i ~in ~a s a t ~ b ~ l ~ and ~ emultilateral ~ a ~ ~ a n s ~ i s s i oc no ~ ~A cbilateral ~ . ~~sastion Disco pair while a m ~ l ~ i ~transaction a t e ~ ~ ~is 2ibp extens s, such as broken or ~ o ~ a K CO d ~ a ~ s ~ be ~ ~ provided. § ~ o I nf there is no con d i s ~ a t c h all ~ s r e q ~ e ~ansactions s ~ ~ ~ and makes

~

Transmission Open Access

As

~ ~ o ~ ~ t a ~ ~~ n ~ ~ . ~ o ~ s i ad p~ or ~ system ~ r ~Qsit~Q n r e s e ~ ~ford 1 is ission losses. (There ISQ pur~hasespower to ed with any one or my aenG0 and Disc

ik

subject ta:

z' E I,; i f

i s the desired or target YaIue of

a u g ~ e n ~ eby d a set of contracted tram ~ n e ~ ~ a lci toyn s ~ i ~ in t s (6.4) are an extension of the c o n s ~ ~ ~augmented nts by additional inequalities for the up led e~amplesof the method with d ~ ~ fc eu r~ ~a i~~ ~e n ~ ewed as a p o i ~ t - t o - ~ otransfer ~ n ~ s ~ ~to a~b iil a lt e ~ont ~ tract^

is the total number of bilatera~muitilat~r~i ~ ~ s a c t and ~ o ~Tks the Mh

bifaterallmultilaterl t ~ n s a c ~ i o nPpL,i , and

,j

are bus i pool g e n ~ a t ~ and o n bus j pool

~ o n s u ~ p t ~ro~n~, ~ e c t iPT., v ~ land ~ . BTkJare, r ~ ~ ~ c ~power i v e~~~ yj e~c tat i obus ~ i Etnd owe^ ~

x

~at bas ~ j ~nu er ~ ~ ai ~ soa c t~iTk ~,n PLT, ,; is

bilaterallmultilat~~al pa~icipantsto make good transmissio~ Section 6.6.4). In actual ope~a~ion of power systems the ~esponsibili~ d ~ s t r i ~between u ~ ~ ~ all dispatched ~ansactions. Twa ~d~~~~~ in [29] but for s i m ~ ~oni ~ i ~ by pool ~ ~ nis c ~~ ~ s~ ~ dhere, t e r ~~ d o ~ T*,i

=o

idG; kE

and the IS0 will dispatch pool power to make good t r ~ s ~ ~ slosses, s i o ~kncludin

Transmission Open Access

-

:The n o ~ condi~ion a ~ is when all pool demand and all bila ns me dispatched without system security violations. ~ a n s ~ c t ~will o n sbe s ~ ~at ~ ircdes~red e ~ value and the ISSO only ~ i e e to ~ sop ~ i s ~ aand t ~ ah ~ ~ ice^. ~ l ~ ceaa s ~~ ~ g~~ to ~ dy ~ ~a d~ ~eOPF ~ ~o~ n0 1~ ~ PPL

subject to:

T PPPL

(6.6)

o vector ~ with~typical ~ element ~ pppL,i ,which is the

is a c

where

a

0 ~ ~ :

and bid price for this pool power;

i s a vector of

pool powers with e ~ e ~ ePPLI ~ts

~ v ~ na b ~of ~e sthis~p ~~ ~ ~~ ~ e ~ ; values of pool ~ o n s ~ n ~ and ~ ~ ibilaterailm on~ arrd e x ~ ~ c ~ i~~h on tors o f r e a ~ t ~ ~o e~ a l elements DpL,i, Pq ,i and ,?Iq ,j , respectively~ ~ e a ~ ~ and ~ ~~~e~~ ~ e sr~s~ectiyely. The ftrst ~ o ~ s ~ ~ n t ~ ~ w ebus r, v o l ~m (6.6) is the c o n v e ~ t i o ~load a ~ flow equation set plus the set of n o ~ power a~ second c o n $ ~ a i nis~a set o f ineq~lities,incl~idinglimits on pool p~~~~ an rating cons^^^^ such as bus voltage levels and line overloads. ~~~~~~

we the c

~ ~ s a c t i in o ~full s would result in the violat~onof operational cons~a~nts. The f o ~ l ~ w i ~ dispatch ~ r o b is~now e ~f ~ ~ u i a ~ e d :

subject to:

T)=

e first tern, within brackets, in the above objective r$pres~ntsthe net pool w ~ ~ f a r ~ ,the ~ a ~ s m i s charge s ~ o ~for d e ~ ~ v PTkj e r; ~ ~

pPTis a columi vector of elements yo,, i s a c

~ vector ~ of e~I e ~ ~nn t s ~ pL,j

~ e s ~value r e ~of (w,,,,,

AP<,~1

=(D:L,i

. ~ ~ L ,where i ~

woPLJ is a w ~ l ~ i n ~ ~ ~ s ~

- D P L J ) ,the pool customer shortfall, w h ~ D~ &e j

and s a t i s ~ ~DPL,] s 5 B& ; y4. is a c wher~ whs

i s the

o v ~ ~~ of~ e~ o ~r e ~~ ~ ~

is also a willingness-to~pa~factor

,i

-

-

,[

and satisfies F$;,,! 5 P;

.

,i

It is worth mentioning that willin~ess"to-payfactors, which have been ~ n ~ 5 d u c eind the above c o n ~ a cmodel t a ~ ~ ~ m the ~ interests o d a of~ ~ of and bila~erallmulsitaterd p ~ c ~~~~n~ ~ cpo ~ g~e s t ~~That ~ n . is, any p ~ ~ ~ ~ j ~bea w n ~t l ~ i ntog make ex^^ ~ ato avoid ~ c u ~ ~~ ~ m eThis n~ t .a ~ ~ a ~ g ehas t ~ to e ~bet ~ g r e with e ~ the IS0 in a d v ~ c e and the 1S0 will d e t e ~ i m ~ ~e g n i ~ d for e s the w i l l ~ ~ e s s - tay o ~factors in order to ration ~ a n s ~ ~ s saccess i o n accordingly. The first c o n s ~ in~ rob^^^ t (6.7) is ~ ~to the~ 5rst ~ in ~ a~ ~but6r with ) 9

,r ~ s ~ ~ ~The t ~second v ~ c~ oy .~ ~in~(6.7) ~ n t b a ~ a equations ~~e and ntract model. Th bilat~r~l/~ultilatera~ ~ ~ ~ c ~ pina advance. nts The i ~ ~ t e ~ of s ~the o in n e ~ u a ~~ xi ~p r e s ~in~(6.6), ~ n obtai

die c o ~ ~ e s ~ ~i o~n ~ a g e problem ~ $ n tin d ~ ~ e g u l power a ~ e ~ systems whic pool and ~ o n ~ at cr ~t ~ a c t i o n s ~

.%3

~~~~~~~~~~~~

~~~~~

a

~~~~~

The e bus system o f Figure 6.9 is used here to ~ l ~ u sthe ~ airat ~ discussed above. For simplicity only the i €actor is examined. ~ y s t data e ~ inc~~d found in the A p ~ ~ n d ~ x ~ e g ~ ~ ~ rat abuses t ~ 1r and ~ 2 bid into the pool and the 80 51 at pool prices. The 200 load at bus 4 is ~ i ~ into ~ two d ~ d ne-half takes power from the pool and the other enters into a g e n e ~ a t at ~ rbus 3. The re~uitsof a pow buses 4 and S are c o ~ ~ ~ e~~~e pl yp l ~ e ~ fuft, line 2-5 it; ~ ~ e r ~ o ~~ ~ ehd . ~ ~ ~ o od of ~ r o b l (6.7). e ~ The s o ~ u ~ o in n sw~ ~ i c ~ wit hi^ ~ ~are giv$n ~ ini Table~ 6.1. 9 ~~

Example system

Case 1 ~ s § u ~~e ~§ i n ~ i e s s - tfactors o - ~ aof~the pool d e at b ~ same as that of the ~ ~ ~ a~t ne s~~ afrom e ~r 3 to 4? namely 20 $1 ~ ~ ~ - €actors ~ o of - ~all apool ~ ~ e n ~ we ~ d~ so u b while ~e~ unchanged. As expected, the pool d ~ ~ a n at d §buses 4 m d 5 case 2 ~ h a ncase I and the bilateral transfer from 3 to 4 was curta consumers were willing to pay more.

Pool ~ e n ~ r a ~at~bus o n1 Pool generation at bus 2 Pool demand at bus 4 Pool demand at bus 5 Transfer from bus 3 to 4

54.6

56.1

119.1

119.5 95.2 1'

94.5 73.8

~

~

75.2 1' 94.3 -k

96-8

The ~ e ~ a ~~ o ~~i ~ entsvand w~ ~ ~ ~e i n ~~~ sr s -n t oa - ~~ ~ ~ ~ ~ ~ i 6.10gis ~ b~t a i n~eas~ t ~ ~ i n ~ ~ s ~ - t of o the - ~ ba ~i ~ a transfer varied fiom 0.0 to 60.0 $ /MW2 h In 10 $ /MW2 h steps while other factors are r e t ~ the n ~ ~ same as case 1. The n o n - l i ~ ~curve a r in Figure 6.10 shows that the more the w i ~ I i n ~ ~ s ~ - t ~the - pless ay, the c ~ ~ ~ ~andr that n e when ~ t it becomes larger the bilateral transfer tends to tb . It is i m ~ and~ ~ ing t to ~ ~ ~ h athat § iwilli ~ e Iment of its own ~ s a c t ~ i ~ ~~ m r ~e u~on e the ~ ~~ e ~of other loads i and ~ aln s ~ c t i ~ R ~ s. ~ ~ I

105

100

95 90

-

85 80 0.0

10.0

20.0

30.0

40.0

50.0

W ~I iI ~ ~ ~ s s - ~( o -~~ a ~y _.

__

_ I _ _

~

60.0 ~

~ I

1

)

~

67.4

Static $ ~ c u ~ i ~ - c o ~ ~ t ~ ~ i ~ ~ d

doIogy E301 to reschedu~epool and b i ~ a ~ e~ansact~ons ra~ t inp: account of s y s ~ e ~ which may be helpful in ~ ~ o v i d i nang insight into the security c ~ a l I e ~faced ~ e s by the context of system der~gulatio~ i s l resented next.

~esched~ling [3 1,321 is the ~ r e v e n ~ ~ v ~ ~ a ~ g e r oop~es~ a t ~c on ~~~ i t i o and n s b~inginga w state. This is an ~ d t5 on-line ~ ~~ c~ ~ o n~u~ g Qand r~ n GO g ~ ~ i m ~ ~ ~ eonly n twhen e ~ the s y s t e ~ is found to be in a ~ ~ n e r a state. b~e It is r e c o ~ ~ s from e d [33] that transaction o ~ ~ - c ~ n t ~c~o~~ e cn~ci cyv ae p a b ~will l ~ ~be ~ e l p f5r ~ le l ~ ~ a t i cn og ~ s ~ a~i no ~~ a tBoth ~ o ~~ o. ~ t - c o ~ ~ ~~ n~ ~ ~~ necontrol c cy t md ~ v ~ p r e ~ e n t control ~ ~ e are, therefo~,taken into accoun~here. The aim o f the method is to m i n ~ ~ i sd e v ~ a ~ ~ ofrom n s ~ a n s ~ s~c ih ~e ~n uw ~h ~ ~~ e es to enswe s~~~~ tive and ~ ~ s t - c Q n ~ n gcorrce;ctive e~cy for the case ofpre~entivec ~ i ~ on ~ the o~ s ~ u ~ ~ t ~ o n d ~ l e d~~~ s a c ~ is i o the n most c ~ ~ control, e ~the ~conisidertition ~ ~ $is to reduce benefit loss of the ~ e v i o ~ s$ ~c yl ~ e ~ ~ ~ ~ ~ s ~ cwhen ~ i ao~ on ~~t ~ g e on c yc ~ . Cd

ion to the case where the may be out o f sus der ~ i i ~ e ~ con o ~ t ons ~ ~ and e deals wi reschedu~~ng o f pool ~ ~ ~ randa demand t i ~ as~well as bilateral transactions. A new nota~ionis i ~ ~ o d here. ~ ~ e d

ieNG

!END

ifNT

and AFT{ we: ~ r e ~ e nc~~ ia~n e~ine s~ a ~ ~ .PG~ ~ c ~ ~ o n where QG,, pool to ~ u s t o sales ~ e ~~ a n s a c t ~ oPDi n md b~latera~ ~ ~ s a c PE, ~ ~ resp o n and NT are the total ~ ~ ~ o fbt r ~e~ s~a c ~si o nf the s pGi and pw are ~

a prices rof pool~~ ~ s ~ a cPGi t~ ~ ~ ~~ §

~ ~ ~ eprice ~ i en f ~o ~ta tsi ~innrespect of bilateral transacti The Q ~ j e c ~ i~v e~ c t i for o nthe klh line-outage con~i~gency is taken to be:

ieNG

ieND

ieNT

the ~

~ pcj, ~ p& c armd~ pTi$ U1 s u b ~ i~t i f f e prices ~ e ~ for ~ n o ~ states a ~ and for con~ingen~ c o n ~ i ~ 5 depen ns aversion to occasional s h ~ ~ ~ i ~n tie ~ mp et i o n or s c~ai~ments. ltiple o b ~ e c ~ ~ v~e h e dprobliern ~ l ~isnnow ~ f o ~ u l aast folfows: ~

(6.8)

Min F = W O.Fo f RER

where "0 is a weight attached to PO and satisfies

pdce at bus m (the calculation of pool price i s outside the scope of this c [233/); 2) pTj,nis set equal to pTj,m plus the ~ a n 5 ~ ~ §price 5 i oof~ later^^ ~ ~ a c Pn~, ~ o The prices p;f under ~on~ingency can be obtained in a simiiar way to pR. ~~~~~$~ ~reventiveand pos~-c~n~ing~ncy G onse~ationof power in the pool, A s s ~ that ~ ~only ~ transact ~ g ide system regulation the associated generatQrbus can be ch to say, the slack bus power a ~ j u s ~ emust n t balance the c h ~ g e in s ~ e and load power with due regard for ~ ~ s ~ i fosses. § s A ~ linear Q ~ pool power b a ~ ~ c e equ~tionin the or^^^ state can then be written as

where rGi, rDi and rn are sensitivities o ~ ~ a n s a c ~PO, ~@o. nI), ~ PDi and BTi with ~ ~ ~ ~ e

to slack bus power

in the n ~ state.~ a ~ any ~ Q contingency S a d~~ i t i o nAteratio ~~ ~ ~ ~ ~ ~ y ~os~~con~~ c oe~nec~yt i Q nthe § linear power b a ~ ~ cequatiQ~ e ~nde~ ~can be witten ~ as ~ ~ n ~ ~ ~

~

w: The flow of power in a line c

since c u ~ ~c r ai ~~ be o allo ~

limits,

0 2 1; I ~ I

~ I E, ~ ~ ~ z

The c h ~ of~linee flow ~ in the normal state ~ ~ u~s t ~

t

~

§

~

where sGj,sni and sTI are sensitivities of the c o ~ e ~ p o ~n rda ~n ~ s ~~~ to ~ ~theQsn §q ~ of~ e the c ~ ~ in~line e Enin ~the normal state.

Transmission Open Access

Similarly, the changes of line flow under contingencies must satisfy

5

2 0 ~

Power System R e s ~ c t ~ and n gD ~ ~ ~ l a ~ i o n

as a h e out^^^^ could c a ~ s generator§ e to lose ~ ~ c ~ rthen ~ the n o~ s ~ , reventiive action to modify the operating state, Hence, d y n ~ se ~ c resc~edu~ing in the context of both pool and lateral dispa~chis also a very i ~ p o ~ a n t issue. An erati tin^ state can be ~ ~ d in~many ~ ~e ~ fdf e r e ways, nt and the operator ~ h ~ the o a~tion ~ e which will not only ensure s ~ s t e ms ~ b ~ ~ ~ ~ e q ~ in an ~ a ~~e n - ~a ~ cenviron~ent. e~s s ~ ~ A ~~ r ~ s~i eener nt described in this section. The TEF is a Lyapunov-like huncti oint (UEP) €or a particular fault is the most c ~ pointt on ~ ~ ~ ~ state spase of generator angles. The transient energy m ~ g i (TE n d i ~ ~ r e between n ~ e the ~ n § energy ~ ~of ~the tsys ~ ~ of ~s ~ c~ u~ ~~ and t~n its ~ value ~ n g at the ~ o n ~ r o UEP, l ~ ~ %e n~ ~ o ~ e s ~ oto~ the ~ dfind i n p~~~-disturbance ~ system and t o ~ l o ~ , the tran~~ent energy is less than the potential energy Cone the system possesses transient stability for the fault in que The chief a ~ a c ~ i aofn a TEN method Is that it lends itself very ~ o n v e ~ e n t tloy a s e n s i t i ~ ~ ~ - b~a s~e ~p r [34]. o ~The c ~sensitivities~ ~ ~ r ~~ a ~t e~ s ~are ithe~c ~h ~ af~n ~c s EM with r e s ~ e to ~ t ~ ~ d ~ g ve ~ i e d~ a~~ t a o~~ w~0~ r ~ In the~ eve ~ ~ . d i s ~ a t cc ~o n ~ ~ u r a po§sesses t~o~ d ~ ~ a security ~ i c riskscs,the most available to the IS0 is to ~ ~ d ~ generator § ~ a power ~ ~ houtputs so as ~ e ~ s~ i~ tf o~ ~~provi~es a ~~ o~ na clear signal o f the most he~e,The a ~ p d e~s c ~ o b ehere ~~ ~is to use the sensi pe in ~enera~ion from critical generators to non-c~tica~ genera to^. -based methods we available, ine~~iding ethod [XI, the c o n ~ o l ~ i nugn s ~ b l eequ ~ e t h can ~ ~ be sused to ~ o ~ pthe u UEP, t ~ and thw EM and the s~~sitivities can be obtained then. A eth hod E39 c Q ~ § ~ ar ~ s~c he edd u l ~isn ~resented next in which the ~ o s s ~ ~many s e § ~ v ~ n ~[40]. a ~ e s

P,,,, Gii Ei

MY,w

= ~echanicalpower input = dr~vingpoint c o n d u c ~ c e I-

y s

-cl -cl @cr ' m8ys

consta~tv o l ~ behind ~ e direct axis transient reactance

of the critical generators and the r e m a ~ n i ~gene~a~ors, g res~ective~y = speed o f inertial centres of the critical generators and the r e ~ ~ n i n g generators, respectively, at the instant when the fault is cleare = inertia constants

~ u p e r ~ c rpr'i pstands ~ for tbe values of variables in the final post-fault system confi~ration. The a p ~ r o x i ~contro~~ing at~ unstable e q u i l i b r ~point ~ ~ (8') is c a ~ ~ u ~ ausing ted a method which is used as s ~ r t ioint ~ ~for solving the post-fault system equ ~ e r i v a t ~ ofthese on results can be found in [42$

where AEM = Eit4""" - EM'. The sign of qi-+j~ n ~ ~ cthe a ~direction es in which g e n ~ ~ ~ is tar be s ~ to en~ance ~ ~ the eEM. The ~ m a g n i ~ ~ofe s e n s ~ t ci o~ ~~ e ~ ~ o ~ d i n change in the ~enerationfrom the most a d v ~ critical c ~ generator to tbe least ilical) ~ ~ ~ e ~will a tbeo high r ~ and, hence, the best c~ndidatefor the ~ ~ c h e d u ~ ~ ~ l g er ~ ~ ~ e ~of tthei ~a r ni t i c aand ~ ~ o n ~ c r ~ t i~enerators cai ~ ~ o ube l dset as

-

Transmission O ~ Access ~ n

a, ~ h o o s e;a contingency from the given set. btain the optima^ dispatch using equation (6.2) (enskng post-fault system static b, of this p ~ o b l e ~ ) . corre§~ondingEM (eq~ation(6.23)). c. d. If EM is positive, then go to step a and select the next c o n ~ i n g e ~Ifc ~ . go to step e. ute the n u m e ~ ~sens~t;vities, al qiY9for the set o f system g e. di§cussed above. For computation of new energy margin for ch g e n e ~ t ~ ofrom n ~eneratori toj, the base loading 6r has been used to get the new UEP c o ~ e s p o n d i nto~the change in the gen~ation.It is fa rbatiQnin power gives the best result. als; c o ~ p u t ethe c o n s ~ ~ n ton s f. the sens~~~vities by m ~ k e tprice ~en~ratQrs; red~spatc~ the ~ ~ n e r a t in o ~pairs s a m g If all critical contin encies we tested, then stop, Otherwise:go to step a. ~ e ~ i l nu~erical ed s ~ d i reveal ~ § that: f cr~~ica~non-cri~cal generator buses is r e d u c e ~ i n ~ r e as ei g~ n ; ~ c ~ ~ l y

~ y n a msec~rity ~ ~ incre~sesspot prices at the buyer’s bus but the ~ n c r e ~ si se not e p r e s ~ c of e b~latera~ contracts

than discus§ed s

er increases the spot prices at the

systems become more 11 entail a more comple S S U ~ Sin relation to con

d price e i a s ~ i issue c ~ ~ [23,2 oraneo~sprice but also ral price d e p e n d ~ ~ of ce ory and is central to pool d ~ s ~ a t cThe h . fo of elasticity are ~ ~ ~ o d u c e ~ :

the conve:ntio~~~ price elasticity o f demand and 2 e ~ ~ sof~ d ~ mc a in The ~~ . a ~ ~ i t i o nsubsc~pt ;a~ j on e

10

Power System Resestructuring and

elasticity and the pool demand, denotes each p ~ c h a sepa~ately~ s~~ p t is an e ~ e m oe f~a ~ me-dependent price vector p (for e x ~ 24~~0~~~ ~ ~podep r i c ~ ) . The follow~n~ ~ u ~ ~ ~ T u conditions c k e r can be d e ~ from ~ ethe~pool

where: C md C me ~ r ~ offer d price ~ cand~power ~ sets of matrices of dual variables, at: time b, on the set of

respect~ve~y, as

- Eet r., G$r.2

Ic

tr

-3

)* * 9

et,,

I

where J i s the n ~ ~o f~poole loads r (load busbars). The ~ ~ s p a t c~hr o ~ ~ dwould u r e begin with a fore~asEof the d ~ y ’ sprices, say a%halfh~~~~~~ n ~and~the c~ ~a ~ ~ expecte~ ,§ ~ setoof ~~~~d ~ d ~~e c t~oThe ~ . s ~ ~of the ~ t ~ lem as well as (6.25) will provide an a I t ~ ~ a ~ set i v eof rchatxaser expec~tio~s, If these match the m i s solved. Tf they do nrot, the exp ~ Qusing the ~ eia§t~~~ties i in~(6.24) and ~ the p~ ~ ~ ec v i~a t ~ o ~ § ain. The p ~ o c e is~ repeated ~ ~ e till c~nver~ence o b ~ ~ i ~ ~ ~ p r ~ ~ ~ d are ~ K~r er o ~ini[23]. ~ e ~

Transmission Open Access

under a normal cQndi~~on is d e ~ e ~byi defini ~ ~ d a La~rangianA and fin its derivativ~with res ect Eo all the v ~ a b l e s : ~~~~~~

1

is the matrix whose elements are s e n s ~ refations ~ ~ v ~ ~~ e ~ e e n and when ~ ~ r ~ ~3 ~e 8for ~ a~nset o § f i n c r e ~ e transactions n~~~ A ,this also satis~es ~ ~ . where the e ~ e ~of~vector ~ t s er than or eqml to zero, whose values d ~ p ~ on n d the state A. Inequ dition that an ~ d ~ t~ i ~o s~ ~set ~~ t AT as^^ at state for the next interval, is in a feasible d i ~ e c t ~ ~ n . nce E257 ~ ~ ~ andv ~ ~e t€ [QJ~~ theo~ ~ n~ i linear ~ ~ ca ~~ an ~~~~~

subset o€ Z ( A ) of constraints in state A , w ich are on limit, If

(6%)

The e ~ a ~ ~~ ~~ Isee~t ~e s ~~ dhere e r eis~&-ke same as Figure 6.9 line 2-3 now has its square o f the set shown as ~ ~ x e d + A d ~ i t ~~ansact~on W at bus 5 is Genco at bus 3, for c o ~ ~ ~ rreasons, c ~ a l is ~ e s ~ Q ~ s for i b l sup e ~ e ~ r e s e by ~ ~~ansaction ed TA (78 MW initially fiom bus 3 to 5), atch. How ~ o o ~ ~ n a ~t ~ o n e ~ a~ r p ~ ~~a t ~~

~ n ~

~

Power System ~

~

s

~;;andcDere~

i

n

s ~ ~ ~ ~a ~~ e c~c i~~s ~t a lis~ now yc h i ~ ~ u s ~ aThe ~ e de .x ~ s pool ~ ~ ~n ~e ~ a n and the ~ i ~ a ~~ansaction e r a ~ fkom 3 to 4 are given oricy in the fWo they are ~ s to have ~ ~ b ~u i prior n e~~~ ~ ofrom ~ the ISO. ~ i

~

~

us 3 Genco s u ~ ~ the ~ ~~eds i t ~ ~ er n a ~ TA alone md only bus I i s ~ ~ ~r e ~ i~ s pna ~to c ~hmake e ~ l up the ~ ~ ~ ~ s ~losses ~ c~ ~s ~by ~ sthe oe d~n 3 ~ e n c o~ v ~ t bus e $ 2 Genco to join in the ~ran§~ction r making good extra lasses. ~ e is w n ~ ~ not ~ ~only n gto ~ o o ~ d ~wn ~ t e s 2 ~ e ~ but c odso to pool ~ ~ ~ s a c~~~~o ou the ng shISO. ispatch results of cases 1-3 are given in Table 6.2, Case 1 is the *nomase md new ~ ~ s a T", c which ~ ~ i so~ i~i a ~ e(r far~Qbus ~ 3 tdlt 5) in is h ~ c u~ ~ ~Case ~ ~ ~2yshows e ~ c. o Q ~ d i n a twitbin ~ o ~ ~ ~ s ~Td,~w~~~~ t i ~ oo w~ ~ u l t i l a ~ ~since r a l both t r a ~3 and 2 supply the ex s 5. Since the l o a d ~ ~ g s ~ n s i t of ~ vline ~ ~2-3 with respect to the ~ a ~ s ffrom er the ~ e ~ $with i r~e ~ ip e~cto~ the ~ ~~ s ~EFom e 3r to 5 is ~ ~ some ~ o o r ~ noa t i o( ne ~q u a ~ ~(6.38)) ~Q ~ where ~ both e ~ a n s f ewe~

case I, not only is the ~ ~ sf?om ~ 3e tor5 ~ u less in~ cme ~ ~ ~ d TA is c o r n p ~ ~ ~ e ~ y transfer is s a ~ i s ~ eind fbll. That is, ~ansacE~on of c ~ ~ r ~ i n aist ii l~l n~ s ~ ain t ecase ~ 3. hen the a d d ~ 70 ~ ~ ~ -5 is ~0~~ to be close to its c a ~ ~ c i a s i the ~ ~~ a n s f e ~ s eref fore inore expensive power at bus I can be s h ~ ~ to e dbus 2 ~ h to bus ~ 2, the ~ less e the~~ o a ~ofi pro n~ bus 3 ~ e n c oi s to p ~ o a smaller ~ i ~ share ~ a i l ~ e nto t ~an§ac~ion TA,Overall ~ e ~aree improve^. ~ ~ s

Tramxtion

Fixed+ Case 1 ~ d d ~ ~ i o ~ a ~

Case 2

GaEe 3

Poo1 gen. at bus1 Pool gen. at bus2 001 demand at bus4

38.8

39.8

t

40,6 t

75.0

75.0 ~00.0 10.0 80.0 38.1

29.2 146.8 f ~00.0

100.0

10.0

~ i i a t ~ rfrom a l 3 to 4 from 3 to 5 Lateral from 2 to 5 ~~~~~

P

80.0 70.0

""

75.0 100.0 10.0 80-0 47.3 22.7

10.0

80.0 49.0

2n.o 1

-- No data apprQachestowards ~ ~ s m i s s i osystem n operation in power ~ ~ r k ewt s~ r ~ in ~ nangopea

skied ~

u u ~ Prmedikre d i ~ ~ ~ ~ ~ ~

price ~ ~ ~issues d i §t c u~~ ~ ec Section ~ ~6.8.1 i s based on the ass able to respond p r o p e ~ ~and y ~ i ~ aare ~ ~ t es r ~ e c trational ly ~ o w e v ed~ , e m elastic~~~e$ ~ ~ of c ~ s t o ~ c r s range from h ~ ~ elastic h ~ yto h ~ ~ h ~l y e ~~ ~~s ~ ~o mwith ~e r ~shighc elastic~~~es . will b s e ~ stoi power ~ ~ ~prices while to tom er^ who are more ~ n ~ ~ awill s t be i ~inert to prices and fail to react in time. ~ r a n § a ~~.Qt o~r~d ~ n a ias i ~dn e ~ a n s ~ a t e&dQ V ~ explores an a ~ t e ~ a t a~ vp e~ r o to ~c~~ ameliorate c o ~ g e s ~ oand n provides useful ~ u i d e l ~ nfor e ~market pa~ic~pants to n ~ ~ ~ puwer e x c h ~ n ~that e s avoid ~ ~ n ~ e sand t i learn o ~ to d e ~ e on n ~geo h ~ c da ~ ~~ ~~ ~ s ~~~~rs u ~ ~(and l y d ~ m ~ ~n od ~~ f o ~~ ~~ ~o §t ~ a d . An ~ n ~ ~ o~ n ~g e m as ~~a~ g~ee r ~~n ~~ ~o~ c [45] e das~$ ~ ~ in ~i~~ o ~ orates the ~ b o v etwo issues into the wiil~ngnes$~to-~a~ ~ ~ $ ~ a ate tch describe^. ~ # ~ i ~ e~0~~~ s~~on

t ~ ~ e ~ tby i othe n IS

R

e U

arket response ~ ~ r ~ procedure ~ n afor~congestion ~ o ~ relief sent a short time ~ e ~ which ~ o ids divided ~ in e~ E need ~not be ~equal to ~ each ~ ~ y s t ee er ~ at ion conge§~ionis found i n ~ o ~ ~ a i iwhich o n , may in~lu~~n o p ~ r a t i nstatus ~

~

$

(6.7) during i n t e ~ a l

operation time ~omain. procedure must iterate in the whole ~ansi~ission This i n t e g r a t ~c~o ~ r d ~ n a ~procedure ion can be seen as a practical aliemat~~e to c a rig~ts,which are defined in a airwise manner l46] or as rights on i n ~ i v i ~lines ~ a lt47J.

~

~

e s ~ c of ~the rpower ~ ~in& last two decades and up to now about 20 countries have r e s ~ c ~ r their e d systems and others are ac~~vely p ~ s u i n gsimilar paths. One of the mo c~~~~ efforts is the t r ~ s ~ open ~ saccess ~ ~ and o this ~ has c ~ a p ~ ~ ~ , Firstly, this chapte~d ~ s ~ ~ ~b he $a r ~ ~ t of e ~Uie$ ~~ ~ ~ s~ d ~ ~

and the necessity of open ~ ~ ~ ~access i s to sf a~~ ~o ~ ~ t ~ t e disc~ssionof e l e c ~ i market c ~ ~ s ~ c t and ~ asso ~ e ~ c ~ ~ p has~ been ~ provided. ~ e ~ The ~ fiinc~ion~ s and r e s ~ ~ n s i b ~ ~ i ~ ~ ~ rns have been d ~ s c ~and i ~disc e ~ sion open ~ c ~ e si,e. s , costs of ~ ~ ~ s r n i s s ~ o n ~ ~ e briefly. ~ ~~ n~ n~this~ec l ~ l ~ ~ nal issues in the e ~ e raz~ ~ ~ g ment and effects of security c o n s ~ d ~ ~ ao t i ~ ~ s h ~ all~been e isc cussed from the o~e~-access v i e w ~ o and ~~~, of this c h a p t ~ rMost ~ o f the discussion is based

, a s~~~~

T~nsmissionOpen Access

17

80.0 Transfer MW

Bilateral Contract

T

100.0

IO.OCI,L,,

20.0

Delivery Price $k 4.WF .3

__

I

1.02

2

1-04

_-

3

1 .OS

--

4 4

** **

20.0 20.0

** Voltages are kept within the range of0.95-f.05. -“ No data

~ i l ~ ~ n (w) gne~ $/MW 2h

80.0

21

Power System ~ e s ~ c ~andr Di enr e~~ I a ~ i o n

Transmission Open Access

239

.David and R.S.Pang, ‘ S e c u ~ ~ ” ~ ar~cheduling sed o € ~ r ~ ~ a c ~ini ao dn s~ r e ~ l pQwer a~~d system’, IEE Proceedings - Generstion, Transmission and Distribution, Vol. 146> No. 1, January 1999, pp.13-18, J.G. ~ a ~ t e n b a cand h L.P. Hajdu, ‘ Q t i i a l corrective re-scheduling for power system security’, r E E Trsnsac#~ons ~ on Power ~ p ~ a r and a r~~ y ~ V~o l ~eP A~S -sNo.2, ~ , ~ 1971, p p , $ ~ 3 - 1. a~ A. T ~ ~ ~ k a c an ~ l a m Tudor, ‘ ~ p r e~ s c~~ ~~ d ~of~power 1~ n g for s y ~ ~~ ~e ~~ a ~ j l IEEE ~ r a ~ s a c ~ ~on o F i s ~ ~ ~ aundsvs~ems, r a ~ s Vol.PAS 71, 1971, pp.2~86-2~92. A. Monticelli, M.V.F, Pcreira and S. Granville, ‘ S e c u r i ~ ” ~ o n soptimal ~ ~ n ~ dpower flow with post-con~~n~ency co~ectivereschedu~ing~, IEEE T r Q ~ # c ~ i oon n sPower S y ~ ~ e mV01.2, s , No, 1, February 1987, pp.175-182. Davivid and R.S. Fang, ~ ~ i ~ ~operational c a n $issues in open access ~ y s ~ ~~ s ’ , ~ ~ srch C o ~ hop, n ~ ~ paper ~ ~resen~ed at U n i ~ e ros f~ ~~ e s t Ae ~~ s ~Jufy l ~~ a9 9~ ~ . M,Kakimoto, Y. Ohsawa and M. Hayasbi, ‘Transient stability anal~sisof electric power Lyapunov hnction, Part I and 11’, IEE Japan, VoI.98, 1978, pp.62-79. re and S. Virmani, ‘A practical ~e~~~ of direct analysis o f ~ransient stab^^^^', iEEE Tram. on Power ~ p p a r s and ~ ~ Systems, s Vo1.98, 1979, ~ ~ . 5 7 ~ ~ 5 $ 4 . Y. Xue, T.V. GuEsen and M. Pavella, ‘Reai time analytic s ~ ~ s ~ i ~t ~~e v~ ~for ~ ot dy sec^^^ a ~ s e s s i ~and e ~ ~t r ~ v e n t i control’, ve IEE P ~ o c e e ~ iPart ~ g ,@, Vol. 135,No.2, I H.D, Chiang, F.F. Wu and P.P. Varaiya, ‘Foundations of direct methods of power $ y $ ~ ~ ~ ~analysis’, a ZEEE ~ Transffctions i ~on Circuits ~ and~ Systems, Vu1.34, No.2, 1987. S.N. S~nghand A.K. David, ‘Dynamic security constrained congestion ~ a ~ a g c ~ in ent c o m ~ ~~ ~ e~ cv~ ~~ ~~ ri Proceedings €~ ~e ~ ’af ~the IEEE PES ZOO0 ~~~~e~ S~ngapore,January 2000, , On-line A l g o ~ for ~ sTransient S~abilityassess men^ and Security Control, P Thesis, Hong Kong Polytechnic University, 1995. B.M. Anderson and A.A. Fouad, Power System Control and ~ t u Iowa ~ State ~ ~ n i ~v e r~s i ~, Press, 1977. P.W. Sauer md MA. Pai, POMW~ y s g ~ e ~ y n and a ~~ ~~ u~Prentice b ~ Hall, ~ Mew ~ ~Jersey, , 1998. S. Sterling, &LA, Pai and P.W. Sauer, ‘A ~ e t h o d o l oof~secure and 0 ~ opera~on ~ of ~ a ~ power system for dynamic contingencies’, Electric Machines and Power Systems, Vol. 19, 1991, p~.639”6~5. H. ~ l and F. aAlvarado, ~ ~ ~ a~n a ~ ofe~ ~~ e n u~~c ~ n g~~~s r ec~ do n d ip~ ~in~~n~ ~ ~ ~n d ~ ration of a power system’, IEEE ~ ~ u n s a c5n t ~Power o ~ ~ y s Vo1.13, ~ e No.3, ~ ~~ ~ ~ 2998, p p . ~ ~ ~ ~ - ~ O 1 9 . .S. Fang and A.K. David, ‘&I i ~ ~ e congestion ~ a t ~manage men^ strategy for real-time system operation’,IEEE Power ~ n g i n e e r ~Review, ~ g Vo1.19, No.5, May 1999, pp.52-54. W,W, Nogan, ‘ ~ o n ~ r anetworks ct for electric power ~ ~ ~ s s iJ oo i ~~of~’ ~~a ~ e ~ ~ oV01.4, ~ 1992, ~ p@f1-242, ~ ~ ~ s , .P. Chao aid S.G. Peck, ‘A market m ~ h a n i for s ~ electsic power ~ ~ ~ ~ ~ Ji os ~s ~oif of f~ ~~ La8~0ryECOPZOP~~CS, Vol. 10, 1996, pp.25-59. ~~~~~~~~

XYan ~ i a o T o U n ~~ ~ v e r s i ~ China

Dr Loi Lei Lai City Universi~Lon

UK

Since China ~ n ~ its~ first ~ eac o~n oe~ ~r~ce ~ o in ~ sjate 1978, e ~ c oe ~ s~ u has ~~ ~ ~ ~ oss don~e$~ic ~ r ~ d ~ c t e ~ e r c~ y o ~ s ~ ~ ~ ~ ~ n prices have p ~ ~ y e ~ ears or so. Before thaf strict but e ~ f ~ c ~ ~ h e l ~ hold e ~ ~ n e d~ e~ y~ in~check, ~ d and until r e ~ ~ te~l ye, c ~s ~h io ~~ ~ e s rion. Now that the market larg~lyd e t e ~ i n e ~e~~ s price ~ e n planing ~ r ~has ~ matured, economic forces are I

~

~with a~~ o ~ c~ i ~oeif grants a ~ ~~and o ns u ~ s ~ d ~ s e from provincial and local utilities. In 1985, the d to set tariffs to recover inve ed ~ o ~ s ia c~ ~ ~s ~s ~r oa v ~~ ~l c~ e s ~ ~ ~new ~ ~policies s h e to~ c r ~ ~at e

. Power ~

in the power sector d u ~ n gthe late 1 ~ 9 ~ s ~ r a ~c r ~e e a~w~~~~i define ~c s~ , how ~ u p c ~ er plant, are not being ~ o ~ o u r~e ~ . ~ ~

c

~

c

cording to ~ a r g i n cost; a ~ that is, old plants usually sell the most power sin ady paid off their c ~ casts p and~need~to cover ~ only fuel,

ent of ~ h i ~electric a ’ ~ power i n ~ u has s ~gone t ~ o u g ha ~ e r y that is, the power i n d u s ~is gradua~lyc et economy c h ~ a c ~ e ~ s tand i c §the i s also ~ h a from ~ the ~ stle n side ~ as the r~~~~ of the electric power ~ n d u goes s ~ on, the de will have a great in~uence01p the f u ~ r e ectriciy sector i s cha~ginga great ~ v ~ ~ oofpChina’s ~ ~ nelectric t in effect and the e a v ~ o costs ~ ~of e ~ ~ d for. These factors must energy use are not yet even p ~ i a l l y ricing s c ~ to~ promote m ~ the ~ s ~ ~ $ ~ a ~ nuse a bof l eenergy 13-61. tran§pare~cy and legal r ~ ~ will ~ eu n s~~ ethat e c o ~ ~ a are: c ~ sh ~ n o ~ ~ ~ d g is s ~ ~ a i ~ and ~ s the e dde~i§~oa-mar~ing ~ a u ~ isoclear. ~ ~ e c o ~ o ~cost i c of pollutio~needs to be con§i~eredso that true ‘kast 60 made. China’s e c a n ~ ~eifc~ ~ i e n and c y e~vironmentalquality ~ e p e n don ~ ~ ~ o ~ s . The ~ l ~ ~ co w ~ ~~c r~ din ~ h ~~ has n§ aalread~ t ~ gone ~ h a t a ~~yc ~ ~ g past ~ e ~ ~Atd ~e .r e ~its~~ ta n, s i t ~ o ~ a p l a ~~c o~n o~~ toyn under way. On 16th 19 of Electric Power w s e m ~ ~ State ~ ~u ~o C ~ ~ n e~ ~ e~r e s~~ o~n s ~ ~ ~w~ i t i e s search. ~ i K grid e man ~~~~

reale a m e n ~ cred~tab~e ~ e service a ard bgal r ~ ~of~~ tn sv e s t ~ ~ ~ The ~ s t a b ~ ~ $of ~ ~ e n t Power ~ o ~ o r a ~ (SP) ~ omn ~ the ~r er i n ~ u ass it~ entered a new stage [7-161, Pilot bi ~ u n ~ c i p a~l i~~ e~ ~~ ih aa ~n d~~o rno~v ~ n and, c~s s in ~~~~, and thea ~ a ~ i o by n 2005. ~ ~ eAn and order1~ ~eneratm ~ g ~ ~will e be t full ~ o ~h ~~d reopower s plant will allow a wi The reform of China’s e l e c powe ~~~ of r e f o and ~ a satooth

~ %hat ~will e e

~~ of

~

~

e ~ e c ~~ c Q i W ~ d ~u o§~ ~ economy presents problems that demand further exploration and confro~ta~ion; many s ~ e to ~ be sa~s~actorily a ~ ~reso~ved~ from basic theory to ~ o n ~~r a ~c ~~i ~~a

This c h a ~ will t ~ ~ ~ ~ o the d u c ~

and ~ a ~ a g e m system ~ t of plan. The p r o b l e ~ and s obstac of e l ~ c ~~ ~ c c ~and ~ ~n g ~ e v ~case r a ~$ t ~ ~ ~ e s

Electric Power Industry ~ e s ~ c ~ inn China n g

The ~ ~ s ~ i of~power u ~ network o n service areas and their installed g e ~ ~ r a ~ ~ 7.1 and are shown in Table e 7.2. A c ~ a ~ l the y , f i t four ~ ~ t e r ~ ~ ~ o ~ i n c shown in Table 7.1 an ins~a~led capacity in excess of 30 CW e ina I n ~ e r c ~ n ~ Nefwork e c ~ e ~ with capacity of over 45 GW has be i n ~ e r c o ~ e the ct~~ uangxi, Gu~zhouand Y u ~ four ~ n HNPG is just below and has not been shown in Figure 7.2. ~~~~~

~~~~

~is€ribution of power network sewice areas

Installed capacity Network &c ~ e ~ i o n

Total

Hydro

Electricity ~ e n ~ r ~ ~ ~ o n

Total

(m) 4"w

(TWh)

(%)

North China Power Net. (NCPN) Northeast Power Net. ("JEPN) East China Power Net. (EC Central China Power Net. (CCPN)

34312.1 37186.6 46121.0 40749.3

15.96 5.94 9.62 30.60

Northwest Power Net. (NWPN) ~ ~ ~ n d Provincial ong Grid (SDPC) Fujian Provincial Grid (PJPG) ~ ~ a n g d Provincial on~ G) Guangxi P r o ~ ~ n c i a i G ~ Chongqing Power Grid (CQPG) S ~ ~ h u Power a n Grid (SCPG)

19275.1 17380.1 8008.0 29027.7 5645.0 315'7.0 11942.3

36.28 0.27 58.28 18.99

141.15 178.93 211.45 160.37 69.60 84.06 32.19

5.62 1.36 5.40 28.70 28.33 0.09 57.41 0.09 56.94 9.37 49.94

58.32

9.93 57.14

103.85

22.78 12.64 44.37

2

Year

Society

Share o f ~ n d u (%) s ~ SIiare

tot&

~~~) ~

Of

~ h Heavy ~ Light l ~ ture

Share of Share of Share of ~ ~ ~ s ~ ~ ~ ~urban c~ k;C, nual i p r ~ tation ~ etc. ~ c ~ ~~ ~ 1r ~ c ”e ~ @.$)

<%I

~

o

C%) I987 49Q,27 8f.Q 1988 535.87 80.3 1989 576.20 79.8 1990 6 ~ ~ . 678.7 0 1991 669.63 77.8 1992 745.54 77.1 1993 820.11 76.7 1994 904.65 75.4 1995 988.64 74.8 1996 i0~7.03 74.1 1997 1103.91 73.0 1993 1134.73 72.0

64.5 64.1 64.0 62.6 61.8 61.2 61.2 60.3 59.8 59.3 58.3 58.0

16.5

7.1

16.2 15.8 16.1 16.0 15.9

7.0

15.5 15.1

15.0 14.8 14.6 14.0

7.0 6.8 63 6.8 6.3 6.3 6.2 6.1 62 4.0

1.6 1.6 1.7 I .7 I .7 1.8 1.8 1.9 1.8 1.9 1.9 2.0

4.8 5.1 5.1 5.3

5.6 5.8 6.3 63 6.9 7.2 7.6 10.0

5.5 6.0

6.4 7.5 7.9 8.5 8.9 9.7 10.2 10.7 113

12.0

~

s

~ ~~

s ~

~

Municipal and Coinrnercial 10%

\Heavy

Chemical Products

Others 10%

~~u~~~ 58% Coal

an ~

x large ~~ ~ ~ e~ -e o ~ ~ ~~r ~hase~assets ~~s eof ~,x 8 . 2 ~ yi ~ ~ ~ o n

t role in the fbture ~ e v e l o p ~ of e n~hina’s ~ ~ o w e in r ng and ~ a ~ a g i nitsg assets well or not. In two y e w 9 pract~cemd e SP has set its deveiopment objective o f creating a ~ r s ~ ~e ~ tl ea~~rins~ s e the wor~din terns of h o ~ d ~ nstock g and g r o ~rn ~ agement, and this is a ~~~e~ de~elopmentof the policy of ~ o ~ o ~ t irs e~d s ~ c ~ ~ g ~ ~ ~ g a l i §~~a~~~ ~d Thus, the SP has made a strategic §~~~~ for fhe fiime d~velo~ment of ~ h i n a electric ~s ~ o w e in r (1) The Erst step, from January 1997 to

~~~~

1998

~ s ~ ~the~ SP~ands the~ d i s ~n~ n~~ l the i n g ini is^ of Ekctr transfer of government functions and p~ofessiona1~ a n a g e ~ e n t c o ~ ~ ~ r u cat inew ~ g ~ e c i s ~ o ~ system ~ ~ ~ a~ k~~ ~~ ge in w coo ~m ~k l ~ with a n ~a ~ o c i a l ~ s ~ m ~ r ~~e ct o ~ ~ ~ n ~ y ;

(2) The second step, from 1998 to ~ 0 0 0 Insisting on the policy of separating the gov etions and taking the ~ ~ o v i n c eass entities ~ o ~ p l the ~ ~~ ~e n sg ~ in the c SP ~ system; ~ n ~

the genera~o~¶ rn

) The fourth step, after 2010

Upon the es~~b~ishment of the SP, the ~ n c t i o nof~ m i n g the s ise ~ a ~ a g e m formerly e~t ~ ~ d e ~ by k e n SP. It ~ o $ s ~ sno s e~~ o v e ~ ~ e n s u p e ~ ~ i ~ from i o n related gov

~

~

e

its ~ n a budget ~ cis a~located ~ ~ directly from n The ~ r g a n ~ ~is~ shorn t ~ o nin Figure 7.4 belo , n a ~ e l y~ o ~ h e aNorth s ~ , China a Power ~ ~ o and u p~ o ~ h w ~ s t Gs5up and Gezhouba ~ n ~ ~ n ~Group, e r i as n ~w d s u ~ s i ~ i o~ ~ e s

~

~ Power e J u ~i n ~ -~ ~ ~eo~c~ ~o r a~t iwo ~h ,~ are ~ h all exclusive~yowned s ~ ~ s i d i a r i e ~

Electric Power Industry Restructuringin China

1

ese state~owne~ assets, held by the Armed Police Hy under the SPC. C o ~ s t ~ c t i oTroops n (also as Anneng Corp.), belong to the SP’s ~ ~ a g e m e n t . Other c o m p ~under i ~ ~the ~ i of Electric ~ Power ~ are the ~ SP’s swholly ow ~ s~bsidiari~s, holding or jointly shared companies according to their property n s ~ c ~ r eThese s . cQrporatiQnsand ins~i~tions under the SP include the fol~Qwing:

8

(I

*

China ~ u a d i a nPower Plant En~~neering General Corp. China Anneng Construction Corp. L Q n ~ ~Electric a n Power Group Corp. ~ h o n ~ e Power n g Tech. ~evelopmentCo. Ltd China Fllecbnc Power Trust & Investment Co. Ltd China ~ ~ e cPower ~ i cTechnology Import & Export Corp. China Fulin Wind Energy ~ e v ~ l o p ~Corp. ent China Power Investment Co. Led China Power Investment Holding Corp. ~ h o n g ~ Electric ~ n g Power Industry ~evelopmentCorp. National E1ectrh.i~ Power China Extra High Voltage Trans, & ~ ~ b s t a t i oConstruction n Corp. lectric Power Fuel Corp. Other e n t e ~ ~under s ~ sthe m a n a g e ~ of e ~the ~ SP.

Engi~eerin~ Institute ing & Design General Institute Natio~alPower Control Centre of China China Electric Power Information Centre Electric Power Research Institute Thermal Power Research Institute Najing A u ~ o ~ a t i oResearch n Institute Wuhan High Voltage Research Institute North China Electric Power University China Electric Power News China Electric Power Press

Power System ~ e ~ ~ ~ c t uand r i~fel~~~ ~ M l a t i o f l

22

The o p ~ r a ~ i o nand a ~manager~alfunctions of the SPC niainly include: ~ u n n ~ nthe g e x ~ l u s ~ owrted v e ~ ~ subsidia~compan~esand Ehe h ~ l ~ori jno~~ n §t h~a~~ d ~ o ~ p aand ~ the i e sia~~"owned ~ stock r ~ ~ inh their ~ s a f ~ l i a ~ units e d ~ ~ ~to state s u law, a ~ ~ ~ ~ g u l a t policy ~ o n ~and s ~ t ~ ~ y . ~ a i s i n gfunds within the financing scope approved by the state to finance and invest in power projects and related enterprises; the income from investment and assets property transfer will be used for capital reinvestment ~ u ~ u toa the n ~regulation; taking charge of national power network int~~connect~ons. Running and managing the large power stations connected to regional i ~ e ~ o r kors t r a n s ~ ~ butk t ~ ~ power ~g across regions and the n e c e s s 8 ~peak~ngand f r ~ u ~ n ~ y r e ~ u ~ a ~power i n g stations. plan^^^^^ and ~ ~ s p a tthe c ~~ ~ € a~power ~ ~network ~ s u p~e ~ ~ s ~i nsafe, g stable, a e~~ o n o m ~ c and h ~ g ~ ~ ~ uo ~ a ie ~r at ~yoft ~ oalf n ~ o ~ w ~ ~in~the cormtry. ~ ~ r ~ s ~ x e r c ~ ~power i n g n e ~ o dr~~s ~ a t c h ~m~nagemen~ ng on the na~iona~ power network and the related generation, ~ a n s m ~ s s ~ and o n~ i s t r i b u ~e on ~~ e ~ ~based ~ ~ son e sthe ~ e g u l a ~ i o n of Power System ~ ~ s p a t c l ~ i n g . The restructuring within the SP, separating generation from transmission and dis~ribution, promoting the na~ionwidepower network interconnec~ionand speeding up rural power ~ n s t ~ ~ t i oreform n a l are the current focuses of electric power industry r e f o and ~ are listed as foflows:

(1) To ~ ~ othe ~r ~ a~ a~ ~~~0~~ ~i ~s nc ~~ r nh ~~a of ~n d~ i~ ~s e r ievels e ~ ~ in the In ~ ~ c ~ ~with d 8the n r~ ee ¶ u ~ ~of~s e ~~ ~~up n sga mode~isede n ~ e ~~~ s e ~the 5Psis ~ g a m u ~ ~ ~ -el ~~v e~ ~ ~ ~ entity r i s~ a~ n/ a ~~ e~system. ~~e an ~ The ~ r ~ ~ a t i o n ~ and its s ~ b ~ i isdan~ equal. a ~ one ~ in law among ~ n ~ ~ p e npersons, d e n ~ and it is a capital link ~ e ~ a ~ ~ o nins hproperty ip between the investor and the e ~ i t e ~ r i invested. se The pilot project was ~ ~ i i ~ ~by a t ethe d Northeast Power Group Company WPGC). The NPGC was ~ e ~ r g a n ~8s s e an d af~li8tedentity of the SE",being an agency of the SP in Northeast China. The three provincial power c o ~ p 8 n ~ eIns Liao~ing, Jilm and Heilongjiang provinces, formerly affiliated to the NPGC were reorganised as companies with ~ n ~ e p e ~ dlegal e n ~qual~~cation, having c ~ ~ p ~~ ee ~~ ~e ~ rights. ~ n nThe i nbasic ~ p ~ n c i for ~ ~r e o r ~ a ~ ~is~tos iset n ~up one p r o ~ c power ~ a ~c o ~ p ~ for n yedc its r ~ s p ~ n s j b i sl ito~ ~ ~ p l ~ rthe ne~t and ~ ~ ao ~ ~ ~ ~ we^ n e t w o ~ The ~ . a ~ m ~ n i § ~ ~~a tn~ vc et ~ ~ n ferred ~ to the s m ~ ~ a g e m edn et p a ~ e onf ~the local g o v e ~ ~ ~ ~ ~ . ~~~~~~

(2) To promote s e ~ a r a of ~ ~g ~~n~ ~ ~ afiom t i o~na n ~ ~ ~and § sd i s~ ~n ~ ~ ~~ntroduce t i o n ~the compe~ition~ e ~ ~anda build n ai n ~o ~~ a l ipso ~w ~ market. r The launch on the ~ ~ i l d ~power n g market was d ~ ~ in eDecember ~ ~ 1997. n For ~ ~ establishing a ~ o ~ ower~ market, l ai step-by-step ~ ~ method was adopted. ~ c c o r d i nto~ the policy, 'power plants can be run by multi~atera~s, power networks must be managed by the State'; the current objective is to separate genera~ionfrom ~ a n s m ~ $ s and ~ o nd ~ s ~ i b u ~ o and to build the & e n e ~ t ~ o n - spower i ~ e market. It has been d e ~ e ~ ~ to n einitiate d pilot n ~ ~ ~ projects in five ~ r o ~ ~ and n c ~one s city, Le, ~ h e j ~ a~ n~~e~n d ~o ~ ~a , o Jifin, ~ e ~ ~ pro~inces ~ n ~and ~~ h~ ~a g~ne~c ~ aau s~eof .the ~ ~ ~ ps ~~~ ~i ~and ~i o~~n st ' ~ d ~ issues, the c o ~ c r a~p~~ er o ~ c hof e sthese power companies are d ~ f f e ~Ine a~ ~~. ~ owith ~ ~ ~ the r e q u i ~ e ~ c of n ~the § SP, the follow in^ ~rinciplesshould bc ~ o m p l with: ~~d

Electric Power Industry ~ e s t ~ cin~China r ~ n ~

0

e

equaI competition high ~ansparency sharing benefit lowest cost opera~ionby laws and reg~la~ions subject to supe~ision.

i o n reorganise several The concre~epractice is to separate generation from t r ~ n s ~ ~ i s s first, generation group companies, and adopt a hid price ~ e c h a n i sin~ genera~ionfor the generating companies, but a few power plants, such as peak regulating units rhsrmal units mainly used for s u p p ~ y i nheat ~ to the local area, are temporarily not included. For the sake of transition, the electricity genera~ioncould be divided into two categories: one is the basic part o f electricity generation, the account of which is sertled according to the current electricity ~~neratioii price cons~deringthe repayment of principal with interest far newly built power plants; and the other is the competitive part of electricity genera~io~~, which is detemiined by the bidding price. As time goes on, the bidding part should be increased gradually. Finally, the principle o f ‘an equal electricity price for the same network and the same quality of electricity’ should be carried out. ( 3 ) To promote the implementation of the nationwide power network interconnection and realise the o p t i ~ a disposition l of resources. Owing to the distribution of energy ~ources and loads in ~ h i n a ,implemen~ingthe nationwide power network interconnec~~~n and realising the optimal disposition of resources is an inevitable option. The construction o f the e x ~ r e m ~large ~ y Three Gorges hydro power station and its ~ a n s m system ~ s ~ ~ ~ motes the f o ~ a t ~ of o nthe nat~~nwide power network interco~nection.It is p~aff~ed that inte~cannec~~on between the Northeast and North China power networks will be acc~mplis~ed in 2 ~ 0 0 the , in~erco~nec~ion between the Fujian provincial power network and the East China power n e ~ o r kwill bc accompl~shedin 200 1 , and the inF~rconn~c~ion between the Shantong provincial power network and the North China power network will be accomp~i~hed in 2003, the ~ n ~ e r c o ~ e c ~between i o n the Sichuan p r o v ~ n cpower ~~~ network and the N o ~ h w ~power st network will be accomplished in 2004. Three crossregional. interconnected power networks in northern, middls and southern China will be basically ~ Q around ~ 2010. ~ The d ~ ~ i interconnec~ed ~ e d power network of the whole c o u n ~will ~ be achieved between 2010 arid 2020. The decisions for the above large engineer~ngprojects are all made on the basis of detailed preliminary feasibility studies of the ~ e i i e f i ~and s effect~vene~s of ~n~ereonn~ction. The f o ~ n a ~ i of o ~the i nat~onwi~e power network in~erconnec~~on will ~ ~ f i n ~accelera~e ~ e l y &hefuture develo industry more e~Qnomica1 and effective way. (4) To s~~~~ up mra

r n a ~ ai ~~s t~i ~~~ ~~~ o ~n at~l ~ Q ~ . I ~ p i e ~ e n rural ~in~ ~ a n a ~ ~ m~enns tt i ~ ~ i oreform, nal technically r e n ~ v ~ ~rural ~ng power nc’nvorks and reali~inga unified electricity price for urban and rural areas in the same power n e ~ o r kwith the same q u a l i ~of electricity arc the current objectives of rural power system deveiopmen~.It will take three to five years. The task o f this r e f o ~i s mainly to s i m ~ ~ i fthe y ~ ~ a n a g e structure n ~ e ~ ~ and ~ to solve the chaos in rural e l e c ~ ~ c i ~ pricing, targeted at realising unified maIiagement, unified a c ~ o u n t i nand ~ ~ a uni~ed electricity price for urban and rural power networks. Technical renovation of the rural power network aim to e the losses of lines and transformers. The estimated inves~men~ is 180 bill~on art. The line loss rate will be reduced to below IS % from

,

Q~~anisation ofthe SP

gy of power ~ d de~e~oprnent, ~ sthe SP~will o ~ s ~ ~ e le d e v e l o p ~ ~byn relying ~ on technical progress, mher d e e p e n ~ n ~ r ~ f and o ~~ d e n ~ open n g policy. In additio~to focusin n ~ d ~ eresearch n ~ and l staff t r a i n ~ nthe ~ ~SP ha echnology pilot projects, namely clean coal power g y c o i i s ~ ~ a t i oand n e ~ e c t saving, ~ c ~ e~ as well as a ~ o ~ ~ u ~ e informati ~ised the $16” has focused on bath international omestic ~ n ~ c i sources. ng ’s power i n d u s ~is still m ~ u o task. ~ s~~e r n ~ t i g a tin i~~ d ~ ~ ~due c to ~ the l t increase in electri in 1997 in China was only 0.21 kW, e i e c ~ i consu~ption c~~ acco~te~ world average. It i s planned that the nation’s total installed capaci W in 2010, the na~ionwidepower ne ect being at the centre. In order to achieve the goals, SP will ~ u8 ~ policy of

change that must take place is that the electric power sector ented to the market rnechanism. The pr~vides8 good o p p o ~ for n ~the~power sector to make itself. These ~nclude: should be worked out in accQ~d~nce wit course of economic d e v e ~ o p ~ e inste nt

e x ~ a n s i o f~s r n a l ~ ~ sci ~o e~ d ~ n s i ~n h~ e~ ~ppower a~ i plants. ~ ~ r ~ ~ g the ~ hc o~ nn~ ~t ~nc g~ofothe n ~ e ~ With o rthe ~rapid dev ~1~~~~ the c o ~ ~ s ~ coft the i o ~ e ~ oincluding r ~ , the facilities from lines to rnediurn~and l o w - v o l ~ ad~i s~~ b u t i no ~e ~ o r lags ~ s beh~nd ~

.4 o ~ ~ ee ~~ists o ~~ ~ ~ la e~ ~~ ~~ ~and e oau d ~ i s~ ~ ~ %awhere conflicts are enc~unteredbetween on and d ~ r e ~ l ~ ~ b e ~ ~ e n value and controlled profit, and b e ~ e %~n o v e ~ ~ e n local and p ~ v a t ~e ~ ~ t i a t iThe v ~ so. f ~ e i a~ d~ ~anager§ § at the vikpiaus levels U necessity and i ~ ~ of transition o ~ to~market e ~ C Q ~ Q but ~ Y they , are not well i about the ~ e c ~ and ~ approache$ ~ s ~ tos realise the ~ ~ s ~7%t ~ o ~ . ~~~~~

cing reform lies at the heart of China’s response to

Electric Power Industry ~ ~ ~ ~ ~in cChina ~ r i n g

33

~ ~ w e v none ~ r , of these has been st~dardisedas ~ ~ t i o np ao~ ~ ~ c ~ inev~tab~e~ however9 there is a gr~wingrealisation that the establis p ~ i c ~ n~g ~ has become c vital ~ to the e ~eve~opment of a s ~ ~ t a ~ ene ab~e Since Ehe mid ~ ~ an ~ ~ncreasin~ 0 s n u ~ b of ~ enterprises, r ~ ~ ~ c ~joint l av r ~ y a d o ~ t ecost-pI~~$ ~ ~ r ~ c si n~ ~c ~that e sbase the price of e n e product ~ ~ ~ on ~ ~ d ~ c t i o n costs ( i ~ c l u d i ~ the i ~recovery of c o n § t ~ c t i ~capital n and interest, o p e ~ t i o ncosts and labour costs), tax aid to the government, and profit. This is a considerable j ~ p ~ v ~ over the a ~ i n j $ ~ r a t i v efixed l y price, but it still results in several amb how to c a l c ~ ~ costs a ~ e in an e n v ~ ~where ~ n ~~ ~ ~~ist often ~ t~ oi ~ob l~ regulate the profits of enterprises. pricing was i n ~ o d ~ c eind 1987, along with $ e a s o ~ a ~ er is a mjar compo~ientof base load. H o w e ~ e the ~, to d e c o~w ~ to n~~the~ ~nabi~ity ~ of rates to cover ri ~ c ~ to ecap^^ s pricing differences, i ~ v e s ~ e ~n t~ v ~ rtos si m o ~a ~ ~ ~ lm capacity~and the inability to collect user fees. The regional and ectldcity tariffs, which are jointly fixed by the state, inre a ~ ~ i n by~ the s ~ ~ ~ ofdPower%The ~ ~o f united ~ ~p r ~f c c ~§n~~ ~of ~th ~ s t grid prices and outes which arc: managed by the ~ ~ o f Power i $ ~ ~ r o v i respectively. ~ c ~ ~ The prices of mid- to small-size power plants man and counties are fixed by local ~ o v ~ ~ m echecked nt, and r a t i ~ e dby r e s ~ o n § i for ~ ~ pe ~ as welf ~ as ~ the, p~~~~ ~ u r e a Tbe ~ . w ~ o ~ e s prices a ~ e of r p ~ which ~ are ~ priced $ by the state are checked and ~ a t ~ f i e d d p r o v ~ c ~ power al b ~ ~ a respec~~vely. ux The base price reflects rice, md has no relationship to ~onsum~tion. The c ~ r c ~ ~ aprice t ~ n g~ ~ ~ o r the v ~ ~ r ~~~ ~ ccost. ~t i oThe n~ ~~~~c~~~ ~ tariff e s ~ c is ~d ~ i einto aix ~ c a ~t e ~ o ~ ~ $ , ~ b w d on uses and v o ~ ~ ~ gThe e s .cate~or~es include: ~~~

electricity rates for ~ ~ g h ~ n g ; md o r ~ ~ ~ n n d ~ ~ s ~ ~ e ~ e c ~ i crates ~ t y for the larger ~ n d u $ ~ ; electricity rates for agriculture ~ r Q d u c ~ ~ o n ;

rirnarily becaus~nationalis~dand § ~ ~ d ~ d ielectricity sed pricing po~ic~es have not been $ ~ e n t ethe ~ , c l a s s ~ ~ c a ~of i oelectricity n tariffs does not reflect the ~ ~ a r a c ~ e ~ s ~ i c t e ~ e ~ ~~ ~~ i ~t y u ~ pfor t ~example, o n ; e ~ ~ c rates ~ cin i s~~~ ~ ~~e~ (e.g. ~ o ~ ~~~~~~~s e r and c ~~ o~t e ~ontinue ~ s ~ to be subsidised~as are the p r e f ~ r e n t ~i ~ ~ sent, tariffs fixed by the .e. electricity prices are b be r a ~ i o ~ a l ~and ~ e d~ner~ased 9 power es have offset revenue s ~ e froni a ~the~~ m ~ ~ h with s y s ~ e ~ - weffects i ~ e of the resulting ~ n a n c ~ a ~

~ s t a b ~ i ~ h iRn g~ o w market ~ r in China will in~oducea m ~ ~ e t - o ~ eec~o tn e~ ~ y , promot~ngs ~ b s t ~ t development ia~ in the power i n d u s ~Tt~is ~ x p ~ to c ~sollve ~ d the 1. Power resource location problem the electric power the s ~ ~ a t i oo nf an electric power shortage>the main issue ~ n d u s must t ~ face is to speed up c o n s ~ c t ~ofo new ~ power plants. Divversi ~ n v e ~ charnels ~ e n ~md ownership of power p h t s will help to achieve the e ~ o a ~Ats .the same time, however, it will also bring ~ r o ~such l eas~the~i n a p ~ we^ mix § ~ c air~~ oe~ ~~u and t ~ onno ~ i ~ s y n c ~ r o ncoo~ s~ s ~ c ot f ~theo ~ ~ w e r ne~orks. In tbc past, the above p ~ b ~ were e ~ cs o n ~ ~bya lthe ~ power s ~ o ~ sg~ e~ a t i o ~ . hen power supply exceeds d e m a n ~these ~ p r o b ~ e b~ es c o ~ ethe ~ a i cons~d~~ation. n Afier establishing the power market, power projects are to be decided a c c o ~ i n gto ~~~e~ ema and, not adm~nis~ative order. Under r e ~ l a t i o no f the ~~~e~ m e ~ h a n i s ~ p ~ w eresources r allocation will be more e f ~ c ~ and e n stable. ~

2. Low a ~ i n i s ~ a t i efficiency on ~ u ~ the n gpast 20 years, r e f Q of ~ the Chinese economy has u ~ ~ ~ r ag rapi o ~ e ~ e ~become a b ~ d while ~ t prices ~ de~e~opment. Supplies o f c o ~ o d i t have no effort to i To compete in the market, manu de~~eased. ever, Chi~a's a ~ ~ n ~ sand ~ ~a ~t oi ~ their o~ t es e ~ ~ ~ e i n d ~ does s ~ not face such p~ssure.lt still o p ~ a ~ aecsc o r d ~ nto~ ~ l a ~ ~ n ~ ~ ~0 ~ o d u ~and e s has a ~ ~ n oelectric ~ ~ pl ~y win~selling. r The gene ratio^ cost has been ~ ~ c ~ e a year s i ~ by g year. The g ~ e r a ~ oand n ~ r ~ ~ ~ i sindices s i o nare very low: for ~ n s ~ n the c ~ na~~onal , net ~ o n s u ~ rate ~ ~ is~ o n about 400 ~ W (standard h coal); the line loss rate is a b o 7%. ~ ~ These ~ n ~ i c are e s fw b e h i n ~the world's average level.

3. ~ c i p r~~ bg~ e m ~ ~ a ~ of the power i n ~ u tos gain ~ ~ more b e ~ e ~ t the s, Under the ~ a d ~ t i omonopoli~s o p e ~ ~ ~strived o r s to ~ a i n t aaj higher ~ rate o f e ~ e ~ ~ At i c the i ~ same . time, the cen ~ o v e ~ ~encourag~d ent ~ ~ vin the e power ~ plant ~ by ~ s ~~ o t ~ sgoiices i ~ s~u ~t has~ ~ 'anew p~~~~l ~ m rate', 'QW p ~ ~ one n t rate'. "bus the e ~ e ratec ~ ~o ~~ ~ ~~ n~ u i n ~ ~ ~every a s ~year d with new power plants pat into average rate in an areas was about 0.47 ~a~~~ 0.67 ~ ~ ~ o r n ~p with ~ d the~average ~ncome h of .~ h ~ n epeop se other c o m ~ ~ d ~ t the i e sprice , of electricity in China is If the power shortage was an obstacle to develop the electricity rate gradually became a new barrier to the growth of China's ~ c Q ~ o ~ To maintain a sustainable development of the national economy thorough reform ofthe electric power i n d ~issurgently ~ needed.

To ~mp~emen~ing reform, the SP has set forth a ‘four~s~ep’ r e s ~ c ~ r ~i na ~m e w The ~ ~ ~ . period from the e s t a b l i s ~ e n ot f the SP in 1997 to the t e ~ ~ ~ a of~ the ion Electric Power was the first step in ~ e a l corporate ~ ~ ~ nr ~~ s ~ c ~From i n 19 ~ , the SP will c o n t ~ n ~toe i n t e n s i ~~~ s ~ ~ ~ w ~ n~ period gc ,~the v e ~ m e n~t n c t ~ o from n s those of e n t e ~ ~ sas e sf o ~ ~ o ~ s ~ ~~~~~

1

2.

3.

4.

5.

In to

ower plants, and a w e l l - r e ~ l a t ~ dt , e c ~ i ~ a ~ ~ ~ wifl be open to all power plants. The SP and a111 r~tionswill run the power n ~ ~ o r k r i s ~ / l e person ~ ~ l and e c o n o ~ i e after 2020, the f o ~ step h of the reform, the Chinese power indus will t~~~ a p p r ~ ~ i the ~ a t ~ ational a ~ v a n e elevel, ~ ~o~ing ~ a t i ~ top n alevel. ~

7.4.3

~bstaclesip2 ~ ~ t ~ b l the i ~Power h i ~ ~

ow to acceIe~tethe pace of reform and smoothly make the ~ ~ s ~ t fkom i o ne ~ i s t ~ n g ~ o n d ~tot a~ ~~ ~~ ks e ~ - o ereetric ~ e ~ power ~ e d i n ~ u are s ~the q u ~ s t of ~ otoday. ~ ~ The ~ dachieve the ~~0~ goals are as foI~ows" ~ b s ~ athat c ~must e ~ be r e ~ o v to plants are not real comp~n~es. In fact, they me just s of their necessary powers held by other higher e ~ o n g ~togthe SP+Thus many of the key ~ n c t ~ w o~s run d i r ~ c t ~ory ~ ~ ~ ~ ebycthe t l SP y and its s ~ b s i d i ~ ~ e t cannot be estab~~shed* beca~ it ~ ~ e ~ ~ ~ e d ~ ~ ~j ~n §g ~,and c er ~ ~ o n a b l ~ n e ~ § . Here we face two major problems. The first one is the rope^ right' issue i s based on the observable fact that un the e n t e ~ r i § ~ins the power i n a ~ must s ~ be ~ ~ v o ~ v aIloc~t~ and o ~~ ~ e ~of their t i property. ~ n In fact, it is a p~oblemof how the ~n~~~~~~will be able to manage and operate itself. The second one i s the ~ r Q b of~ §eparatin e ~ ad~inis~~t~ve t and e n t e ~ ~~s e~ c ~Upi to o now, n ~ the ~ t~~aitional s basically u n c h ~ ~ e although d, the SP was establish wer was t e ~ ~ n a t e C d . ~ e ~ with l y ~such p e~ ~ ~ t o i ~m ~~ ~ ~ s a ~ ~ the s t ~ c of~ ~e ~ o rights, 1also r e ~ unr~solv~d ~ n [I]. the ~ l e c ~ price c i ~ is quite c ~ n ~ s e . ~ ~e$onomic c e reform began in ~ ~ i ~ a , s has been ~plementea.But the e ~ e c ~ ~ ~ ~ e ~ o v e ~ b~~ ce an~the s~e electric i n ~ u i ss consi ~ a ~ ~ the ~ cOv~rall t e c o ~ and ~ ~ syt ~ d a ~ofd living. T h e ~ e ~ ~ r ~ ~ r e ~ o ~ a t i oofn the power market is rigorously l ~ ~ to~the~ gene e a f a ~ v e r s i ~~ni v~e ~s ~ e ndifferent t, hnts can be economicallycla

4. Small hydro power or thermal power p ~ a n constructed ~s by focal g o v e ~ m e n t They s ~ are u s u a ~ ~mn y by local power ~ o m p ~and ~ esell s electricity a c c o ~ d i ~to~the g price audite by local gove~ments.

5,

a1 power p ~ a sold n ~ to f o r e ~ ~e n~ ~ e ~ rTa~ get s efurads ~ . to ~ o ~ §n ~ c t in solme regions, several t h e ~ apower ~ plant8 have been sold to the agreement of sale regional power companies tee that this kind o f ~ e ~power a l plant will sell a certain amount of e i e c ~ ~ cto ~ tthe y grid each year at a 6 e ~ i price. n

Prices for e ~ e c power ~ ~ c varied co~siderablyacross p r o v i ~ cand ~ ~even withk $ ~ a chase agreements, which define how much e ~ e thatc the~ ~ ~ m the power plant, are not being honoured. Old pXants ~ s u ~se1 ~ly most power since they need to cover only 5x1, operatio~and m a ~ n ~ costs ~ n and ~ ~neee not pay their c a ~ cost, ~ ~ a ~ From the above classification of power plants one can see that the price system 0x1 ge~ierationside is very omplica~edand it is very difficult to form a n o ~ a ~ i s ceodm ~ e t i ~ v e n ~28 ower Law ap~rovcdby the People’s Cong~essof ~ h j on f date. It i s almost useless in r e s ~ c ~~h n~ n~~a ’ ~J power ~ ~ l~shmen~ and i ~ ~ r o v e m eofn ~the electric power m ~ k e t~ u § tbe ed by a complete legal f r a ~ ~ w oThere~ore, r~. new ~ e ~ i s l almust t~o~ dition €or the neces reformation of the electric power. be dealt wilh are: 1. prop^^ owne~hip:This subject occupies a very important and critical p o s i t ~ oin~ China’s electric power ~ n d reform. ~ s Without ~ a thorough ~ ~ a r ~ fand i cfia ~ ~ ~ ~ e f i n ~~ o~n~c oe ~the i nproperty ~ right of regional, p r o v ~ cpower i~~ c i n ~ ~ ~ ~e o ~w d~~~~§~ ~ er ~ the ~ ~ a n s ~ to~ a~ true s ~power o ~ m ~ ~ e t

l a ~ i ~ nIns ~today’s SP lmanagement system, temp a d m ~ ~ ~ s ~orders a t i vare ~ still the main measures used to manage ~ ~ ~This a ~s r~ ~~ .shows a ~ the~ ~o ~ ~ ~ h n ~ ~ ~ ~of t e~ s ~ i ~~ s ~~~0~must be d e f i n as ~ ~the basic p ~ ~ c i p ~ ~ p r o ~ ~ power c i ~ ~ c o ~ o ~ a t i oand n ~ ind they often show more concern a ~ o u their t ing ~ i $ ~ b u t ~ofo social n ben^^^ is an i t of the ~ o ~market. e r In s ~ mChina’s ~ ,e l ~ power ~ ~ industry c is at its initial stage of r e f o ~ a ~ ~ o ~are . many c h a l l ~ ~toe sbe over~ometo establish a fair and efficient power market. ~~~~

The characte~~stics of ~ r n ofo g ~e ~~ ~ ~~ a t s u ~ t a ~for le ~ h ~ n a

ity ~ ~ 6 ~inxChina 1 g originated from the r e ~ u ~ r e moef~a t1 g f ~ and n ~~ ~ v e s ~ eThe n t . Ps used in the uf( are not do~og~es have beea $~ggestedto d e s ~~~c~~~~~ ~ ~ price

~

Power System R e § t ~ c ~ and u ~ ~~ e~ rg e ~ l ~ ~ i o

23

~ § t e ~~ s~ ~ ca two-pm l u p~r ~ c~~ ns g~y s ~~eThis ~ ~" section will discuss a one ~ ~that cm § cope ~ with e the above ~ rob^^^. ~ o § t of s e l ~ c t r ~include c i ~ o ~ e ~ costs ~ o md n inves ~ ~ of e ~~ ~costs ~ involves ~ c i n~ a ~ ause operation optimisation is the basis of r e ~ ~ a ibs ~the~ basis i ~ ~ of determining capacity i R v e s ~ e n t p r o d ~ c ~ osimulation n of the power system becomes one o p r e ~ i e ~l e c ~ ccost i ~ "71.

~

In order to analyse e l e c ~ i ~ i costs, t y we should mn a ~ r o b a b ~ l ~~s tr~ Q c d u s ci r~n ~~ ~ ~~ ~ Q R y hour. Then we can obtain fuel costs F(d) and loss of load ~ r o b a b i l i td;~ (t = ~,2;.*,~760) - This data is the basis for cdculati Variable costs of electricity consist of fuel costs be re~re§en~ed by

where

e7ti., : ~ e ~ ee ar~ a c i~for i ~~base ~ load ~ ofthe s y s ~ e ~ : ~ e n e ~ ~ci ao n~ a for c ~ peak ~ load of the s y s ~ e ~

K, : a n ~ rate ~ aper-unit ~ capacity for base load K , : annual rate per-unit capacity for peak load

~ $

dicting generat~oncosts for each hour, the annual rate of evenly d i s ~ ~ bamong u t ~ 8760 h, while, the ~ u rateaof ~ ~ h o u~~~ h e (r e fthe~~G~ O~S ~~of) h be o~ ~~ s~ ~ ~~ cb o tou~ ~~~ i n ~~€or e&ch ~ for hour~t is c ~ ~ ~ ~ ~ ~

where

F ( t ) : fuel cost ofthe power system in hour t ( t ):~ ~ ~ the ~ risk ( tlevel ) in, lmur 8

R A : the risk level in the investigated year 8760

where P(2) is the system load at hour t. The ~ a ~ cost ~ p(t) ~ of n e a ~ ~e c for ~ hour c ~t is~

Electric Power Industry ~

e

s in China ~ c

~

~

~

~

39

~ u ~ s~ ~~u a ~t i (72) o ~n tinto~thenabove ~ equation, we have

in equation (7.5) can be found by running a probabilistic production s i ~ ~ l a ~To ~on,

~ ( 0

find the second tern of equation (7.51, we can use the follow in^ two methQds. I. ~ ~ i ~ tgae ~i ~n e r a ~capaciv ~ o n ky, u ~ ~ h a and ~ gincrease ~ , I unit load for each hour, ~ ~ ~ o b a ~ i ~ i ps triocd ~ ~ t i os~mulation n in this situation, ~

~

~ ~ c ~peak ~ aload s e c a p a ~ kt/, i ~. Because it is d i ~ ~ utol get t the cost of loss load, the second way is ~ r e ~ e ~Ue d . above ~ o n ~ e ~~ u~ a o~ (7.5) on n can be r e ~ ~ a asn g ~

5can be ap

~ i ~ a t e found l y by the following ~ ~ o c e as d~e in Fi ~ ~ ( t ) LDC in the ~~~~e is the load ~ u r curve ~ t formed ~ ~ from ~ ~ ~After~~ ) n. a n ~~~~

The risk level o f ~ whole ~ e year, LOLPA,i s determined by tbe abscissa W, i~ ~ 7.5 is g the ~load duration ~ e curve formed by P(t) -E- hp ;here AP is an inc duration cuwe become C , With the same L O P A we can nd a point in ELDC, the abscissa of which is erefore, the section of line A r e ~ r e ~ e the ~ t sc a ~ ~ f~ ~t e~ I equa~~on o ~ ~ g ~ ~ ~ ~ eAW i ~ Thus e n we t can s ~ ~ s tthe I

~

i

a ~ ~ to e r~ a ~ the~ n e~ c ei s s n~c ~ a c reserve i ~ an in[: ~ ~ ~ ~ acapaci~, r i o nwe can draw up ~ ~ n ~with a c ~ s ~ a p aof~ the ~ system t ~ is not eno ; in ~e~~ these ~ o n s will u ~ s ~ ~ af c o n ~ ~ r n ethe, ~ s ~ e n e ~ cost a ~ s~ ~o ~onot ~ inch l ~ cost b e ~ ~ ~ e ,

ises have their own power units and ~ o i n ~ u s er ~ a l they do not GO respective reserve, their electricity e q (7.8) ~ ~ so~ ~~ e ~~a q~(7.6). ~ ~ ~ ~ ~o n

Electric Power Industry Restructuring in China

0.9 1

11

*

0 1

(a)

2

3

4

5

6

7

B

9 10 11 12 13 14 15 16 I7 18 19 20 21 22 23 24

I

1

September

Xn order to supervise the market price o f electricity for the ~ o v e ~ m easily, e n ~ we can calculate several characteristic costs (or prices) for an interval o f a certain time, say one week, one month of: one year. the f ~ to ~ ~e ~the n c~ ~ ~~~ ai ~~casts ~ ~ eefor~raat~i n~ ~t ~e of ~~ one ~c ~ is 27, and the set of s h o u ~ load ~ e ~time i s T, . The n u ~ b of e ~h

~ are~ t , ,r t, ~an

r e ~ ~ e c tand ~~~~y, 1, -t- I,, -t t, = 8760

~

l

~~ oc n ~s in~ the i ~ peak, ~ e ~ ~s h o ~~ l ~and ~ f valley : load ~ ~ r i are o ~A,,s A, an

r ~ s p e c ~ i ~and e ~ yare , calculated as follows:

242

Power System Restructuring and Deregulation ___-

._.

I _ I -

I

If the cost of e ~ e c ~ igenera~~on ci~ in the peak, s ~ o ~ i and ~ d evalley ~ load p e r i ~ is ~ sCp, Cs

and C, ,res~ect~vely, then they can be found from equation (7.2) as

(7.10) Mp

id,

tETV

ence we can d ~ t ~ ~the i naverage e cost of electricity for the peak, s h ~ u l ~and e r valley load p e ~ ~ as o~s

The ~ v e cost ~ of ~ ~~ le

e for ~the year ~ isc thus~

~

(4.13) ra

and is,, are average marginal costs €or the p

s respectively, and are c a l ~ ~from l a e~ q~~ a~ ~ ~(7.6). o n The a ~ e time ~ used ~ e costs u ~ ~ ~ eforca real ~ cpower i ~ system are s ~ ~ in w Table n 7.3.

Peak load 0.402 1

7.5.3

Shoulder load 0,2014

Valley load 0,1133

~ ~ e c ~~ ~ i~ ~ ~ ~i i ~ ~ ~ ~ ~~w~~ i~ ~Q~~~~ e~ r ~ -

~

~

~

7 the ~ ~ t i ~ electric ~ w ~ ~~~e~ d e shortage which last

As a c ~ ~ s e q ~many ~ ~ power c e , plants suffered a 1 first t h e . T h e ~ f ~ the r e income and ~ ~ of the n plants e were~ ~ of m a ~ the n ~p ~ v ~ n cp~oa~l e cr o ~ ~ r ~ ~ ~ o s and ~ ~dispat~h, i ihe ~ p ~ e ~ flows e ~r alon lly died off. Owing cap the emph

This operation obviously The cause of such a p e ~ o ~ is~due c to e the pr a m ~ ~~r Q g vi~c~s. Let us i ~ ~ u s this ~ ~~t reo b ~ eby m a real ex^^^^ of the ~ o ~ hp w ~ e ~r s sy ts ~ In e~~ resource^.

~

~

~ China, ~ w four ~~rov~ncial s t power s y § ~ e ~are s ~n~erco~nected. These ~ ~ o v i n c e s

Electiic Power Industry ~

~

s

~in China ~ c

~

r

~

n

~

i n ~ l u d~~n~~~ ~ ~ ~ n s~ui ,n ~ xand i a i n ~ ~The ~ ~i i. ~ j ~ a ~~~~~c n g ~ o ~~ ey sr ~ eisman isolated system. The ~ o source w mix ~ of ~ the: n o ~ h power w ~ ~system ~ iat the end of 19 shown in Table 7.4 and F i ~ 7.7. ~ ~Wee can see that in Shaanxi a d ~ i ~ ~ ~r ~ x~ ~ ~n cae s electricity is mainly s~~~~~~~ by c o a ~ - ~power r e ~ plants; in Gansu and ~ ~ n ~~ hr a~ i v ~ n ~ more than half the electricity is supplied by hydro power plants. Therefore9 u ~ ~ disp~~chiRg in the n ~ ~ power w esystem ~ can ~ make a significant profit.

i

I

I

40%

(a) ~ e r ~power a l

(b) Hydro power

re 7.7 Weight of we^ ~

ai n s ~ ~a in ~ ~four ~~dprovinces ~

i

~

la 7,%Power source mix in the northwest power system (~~~ Province

Thermal power

Hydro power

Total

~ ~ a a 4025 ~ i

44.6%

988

17.5%

5013

34.2%

Gansu

2668

29.7%

2285

40.4%

4953

33.7%

400

4.4%

2080

3 ~ * ~ % 2480

I&9%

5.3%

2224

15.2%

100%

14671

100%

In@&

~

~

Total e 7.5 ~

n ~ 21.3% x 1924 9017

o

100%

S ~ a a Exchange ~ ~ Energy ~~~~~~0~ cost

~

i

ia

Total

5455

nbenefits ~ ~ o f interco~ection~ i ~

Load ~~e~~

Gamu

~ 302 ~

Separate Operation 63568 ~ ~ 6 D O 32 3,26887

~ ~ ~ ~Energy i a n g ~ 0 ~ ~ @OSc e ~3 . 4 ~~ ~ 7 ~ ~ 9.71239 Fixed Cost Lmd E n e r ~ I8747 ~ e n ~ r a tEnergy in~ x 8747 ~~ x~Energy~ a~ 0 ~ ~ ~ ~ ~ e r ~ tCost ;on 1,04226 3.37173 Fixed Cost Load Energy 30079 3095 I G e ~ e ~Energy a t ~ ~ E x ~ h a ~ Energy ge 872 ~ ~ e rCost ~ t ~ o ~ I. ~ ~ ~ Fixed Cost 4.27124 Totat ~ ~ ~ ~Cost t i o 35,59297 n ~ n ~ ~Benefits ~ ~ o ~0 e c ~

Interconnected Operation ~ 3 5 ~ 8 8 ~ 4 5 ~ 17890 4.34856

39318 -24821 ~1.78287 Q ~ ~ ~ ~ ~ f 8747 8~49 e-10078 0.19944 1,19632 30079 47088 17009 ~~ ~ 36 D ~ ? ~ ~ . ~ 9 8 ~ 4 33.294~4 2.29884-

-5

-

8

Electric Power Industry R e ~ t ~ incChina ~ ~ n ~

2

Effects of prices on interconnecting benefit d i s ~ i b ~ ~ i ~ n

Gansu of change shaanx~ ~ ~ W h ~ 0.35 1.82291 -1.47891 0.34 1.64401 -1.23070 0.33 3.46511 -0.98249 Q.32 1,28621 -0.73428 0.31 1.10731 -0.48607 0.30 0.92841 -0.23786 0.29 0.7495 1 0,01035 0.28 0.57061 0.25856 0.39171 0.50677 0.27 0.2~ 0.21281 0.75498 0.25 0.03391 1.00319 0. 1.25140 0. 1.49961 0. -0.502~9 1.74782 0. -0.68169 1.99603 0.20 -0.86059 2.24424

Qinghai

~ingxia

-0.005 17 0.09561 0.19639 0.29717 0.39795 0.49873 0.5995 1 ~.70029 0.80107 0,90185 1.00263

1.96001 1,78992 1.61983 1.44974 1.27965 1.10956 0.93947 0.76938 0.59929 0.42~20 0.25911 0.08902 -0.08 107 -0.25 1 16 -0.42125 -0.59134

1.30497 1.40575

1.50653

Transmission of electricity is becoming a separate industry ayer. A viable ~ ~ s ~ i bus~nessis critical to a s ~ ~ c c e s compet~t~ve s~l electric ~ ~ k Ine the t ~past, the i n a ~ p r o ~ ~ends a ~ e of ‘~rnphasisinggeneration, i ~ o ~ansmissjon~ ~ n ~ in the el power industry ~f China made adequacy ~ ~ s r n i s very s ~ o oor. ~ Since the basic business of the SP and its ~ubsidiariesis ~ansmissionthey hav a duty to prov~de~ ~ o u g h ~ a n s ~ ~ s s capacity ion to satisfy the requirernen~of the power t ark et, This is a massive ~ n d e ~ that ~ ~will ~ cost n gbillions of yuans. Where will the money come from? The r e f o ~of the power industry brings ~ ~ s ~ i s pricing s ~ o into n a new there is a g r o ~ ~ need n g to identify the costs of ~ a n s r n ~ s ssieQ~~i c e sIn , such a one s h ~ answer ~ ~ dquestions such as how much i s this ‘g~neratoror load’ making um of this ~ a n s ~ ~ s line? s ~ o nOr ‘what p~o~ortion of the n ~ ~ losses o r i s~ allocate^ to this gene~a~or (or load) ? olutions of these prob~emsare very i services provided by transmission systems, and hm a direct hfluen This section presents a comprehensive inves~igationof load fl c o s ~ i Two ~ ~ . current d e ~ o ~ ~ o s i t axioms i o n are first i n ~ o ~ u c easd the f ~ d a m ~ ofn ~ ~ s load flow analysis in wheeling costing. Then rigorous math cal models of ihe ~ e s ~ b l i s h e ~To. solve these dist~~bution faGtor problem and loss allocation p ~ o b l eare p r o ~ l e ~we s , i ~ ~ o d u ac series e of theorems based on graph theory and a very simple and e f ~ c a~l eg o~ r ~i ~is~developed. Finally, case studies are introduced to ~ ~ l u s ~ the ate u § ~ ~ ~ l nofe the s s ~ r o ~ otheory s e ~ and algorithrn [ 181.

Electric Power industry Restructuring in China

7.6.I

Current Decomposition Axioms

~ a r ~ e t - d r i v etransactions n have become the new independent decision variables d e ~ n i n g the behaviour of the power system. Understanding the impact of bilateral transactions on system losses is important in order to aliocate a c o r r ~ s p o n ~ nloss g co~ponentto each ~ndividua~~ ~ n ~ a and c ~ improve i o ~ econom~cefficiency. One essentia~ piece of ~ n f o ~ a t ~that o nthe biI~tera~ market needs in order to improve economic e ~ c i e n cis ~ k n o w ~ e o~ f~the e ~ ~ n s m ~ s 5losses ~ o nassociated with each proposed b ~ ~ a~~~ ~n s~a ac l~ i o ~ . This k ~ o w ~ e pernits d ~ e buyers and $ e ~to~~ ~ c~o rs ~ othe r ~level t e and cost of Iosses into their n e ~ o t i a i i o The ~ ~ , essence o f the pro~osedloss aI~ocati~n ~ e t ishthat~given ~ a path, along which the ~ ~ a i ~ s a ~ €vary i o nwith s time, it is p o ~ s i b ~toe find for each ~ ~ ~ ~ ~ ~ e s i n ~ ~ e m e n~t ar~a n ~ ~anc ~~ st s~onc ~~~~~~e a ~ e ~and 5 e ~ a r foss a ~ ~~ ~o ~ ~This ~ lea o ~ e ~ ~ to a loss ~ l l o c ~ c~ ~oo n~ ~ for o each ~ ~ ~ n~ n~s a c tAi n~ ~~, ~ ofb c ~u r~ ~n r ~ ~ ~ for o 5 a ~ c ~ ~ ca su s ~i ~~~~~~ ot e~nd t~ ~ tosses ~ c ~haye a ~been ~ ~ ~ ~ f 19-24]. o § e ~~~n d ~ o n s ~ ~ ~ we need to identify the power (or c u ~ e n ct ~o ~ ~ o n of en~ the ~ r o ~ I o~f w r nh e~~ I i ncos< each branch and allocate the effects such as losses to its componen~s~ To solve this kind of pro^^^^ i t is not enough to use onIy Kirchhofl’s laws of electric circuits. ~ h e ~ e f o rine ,this section, we introduce two axioms. Assume the current o f branch k, consists of L current compo~ents 1 ~ ~ ) , ( 1 l= ~ , ..,L) . supplied by L generators, L

(7.14) /=I

where I(,) and I(,), are the effective or r.m.s. values of the currents, which can be either ‘active’ or ‘reactive’ components. Similarly, in the following description, the term ‘p~wer’ can also be replaced by either ‘active power’ or ‘reactivepower’ @5]. The coiiipon~ntso f current in a branch are conservative. The axiom states that each co~ponentI{,)[ is the same at the initial and terminal node of a branch,

(7.15)

~ ~ s ~ ~factors u ~ are i othen same at the two nodes of a ~ r ~ n c ~ . is obvious when we define f C k j fby cments as shown In ~~~a~~~~~

~

,

because both I(,)! and IWt ~ a ~ nthe~ same a i values ~ at the two nodes o f ~ ~ k~, c h ~ o w e ~ine power r ~ s ~ s ~ ea m ~ a ~ ywe s ~us~ u a ~ use I y power instead of c u ~ Thas e ~we~ ~ § ~ i op~~l ~~this v es ~ ~ ~is also ~ ~true e when n t we use power to defme d i s ~ i ~ u tfactors. i~n ~ s s u m the e v o l ~ g e at s the initial and terminal nodes of branch k are U, and U , . Thus the r ~ s ~ e c t i powers ve are

47-16) Tlie powers at the W O nodes supplied by source I are

2

Power System Reshucturing and Deregulation

8

This corrcludes our proof.

e ~ r ~ of ~toss c~ ~~ i ~~on~~theaebasis ~ io fo ema ~ an^ s q u a ~ ~was d also sug 271, i.e. the loss a l ~ o c atot ~component ~ current I(,,, should be calculated a6cordi~gto

The current components in the outgoing lines of an i ~ J cunene e ~ at ~ba node ~ ~ are ~ ~ o ~ o to~the o cnu rar ~~n ofthe ~s going ~ ~ n e s . ~ s s u that ~ ~total n ~~ u n ~n~ected e ~ ~ at nodei i s Ii, this a x ~ so ~~ ~ that e swhen the

t ~ ~ n~ ~ ~ o current ~ n ~ bye generator c ~ ~ 1 ~at node i is IiZ its ~ o ~ ~c o ~ e ~~in ~

h e k is

where

"xi#)

is called a ~ l ~ ~faclor a ~ of i #line ~ k, Qi{k)

= ~~k~~~~

~7.~3

The whole loss caused by t r a n s m ~ ~ energy i n ~ from ~eneratorsto a node is called loss ofthe node. We will denote the loss of node i by Pi

rs, the loss of node i, d ! $ is equal to the total loss of these ~ncominglines. To the outgoing lines of node i we have the f o l ~ o w corollary. ~~~

The factor of n o ~ eloss a ~ l Q c atot ~an~ ou~goingline is equal to its ~ ~ ~ o c a t ~ o n factor.

Assume that node i hasLi incoming lines all directly cQnnec~e~ w ~ ~ n ~then ~ the ~ loss ~ o f rn ~s i~~ i se L

i ~ lines g of node i are not all c

~

~to the ~ g ecn e ~t t o~e~ $ ~~

the r ~ ~ s i rve aes o ~ n

As ~ c n ~ above, ~ ~ there n e are ~ rkvo ~ ~ o b l e m re~ated $ to load flow a n ~ ~ y ~ ~ s , namely the ~ i s ~ i bfactor ~ t ~problem ~ n and the loss ~llocatio~ probl the ~ i s ~ ~factor ~ ~~ ~~ i b5 ln e ~ . For a s p ~ ~o i~ ~e e~ c~ao ~n ~~ ~gof Q an ~ Q w e~r y ~one ~ can e o~ , ~the t ~ ~ Ihe or ~ ~ by a loa s ~ ~ e d i s ~ ~ ~ ~factors t i Q nof each g ~ ~ ~ r afor t o er s N nodes, N , gene~at5rsand N, branc ctars d e ~ n by ~ de q ~ t i ~o7~. 1 $ ~ , &E the ~ ~o~erat~ng c ~ ~ o n ~~ i e ~ ~~ ~ ~ ~ e ~1afor~~Q r r k i s ~ ~ c ~ e~ ~ c o ~~~ o nee in n ~o~tgoing ns ~ lines at their c ~ ~ ~asnshown ~ s in , equa~~on (7.22). vo~tageat node i yields

e n e r a 1~can ~ ~be c a l c ~ ~ ~ ~ e dation (7.21). To do

~

~

Power System ~

e

5 and ~ Dere ~

~

(7.26)

where = [PiG,P? >. *. ,P,G f'

is the vector o f ~eneratQrpowers and =E S 9 Pz 9 *.*, P, I'

is the vector o f total node injee elements of which are defined by (~.27) 0

r, (i)

otherwise

is the set o f the outgoin

b r a ~ ise ~~ Q W hence ~ ; the eIeme~ts us i l l ~ ~ ~this a t with e a simple p have the f Q l ~ Q wrelationship: ~g

1

ove ~ a t h e ~ a t ~model c a l is ~ ~ o r Q uand s , does need not to ~ ~ thev ase l~sslessbranch as adopted in [25].

A circuit ~ ~ a g r afor m simple power systems

(7.28)

n

Electric Power Industry R ~ § ~ c in~China ~ n g

equation (7.26~we can obtain the c o n ~ b ~ t i o nf each generator to the total r at each node from the following equation: (~.2~) B", = -1

5

Thus the con~ibutionfactors can be readily calculated by equation (7.25). After getting f ( k ) iwe , can further allocate the loss of the transmission network to each generator by the f o l l o w i ~equation ~ according to equation (7.2 1): (7.30) k=l

is the loss allocated to generator 2. where ~ Q w e v ethe r ~ loss allocation robleni can be an independent problem. ~ h ~ r e f ~ need a ~athematicalmodel for problem to allocate loss to each load or each gen To allocate the loss to each load, the key step is to calculate the losses of ~ j ( j ~ 1 , 2 , . . . , N ) . consists of two parts, as follows:

Firstly, the sum o f loss APg in the incoming line i j E

( j ), where

r- ( j )denotes the

set of the in~ominglines of node j . S~cond~y, the loss of @, allocated "CO line ij

which can be calculated according to equation (7.24). Note that k balance equations are as follows:

($

i j E l?-

(9) . Tbe loss (7.31)

where ai(k)is the al~ocationfactor d e ~ n e din equation (7.23). For cQnve~~ence we can use the following form to determine

:

Here E";:L is the load power at node i . Equation (7.3 1) is a linear equation system including N unknown variables o q , which be solved by a conventiona~ a l g o ~ t ~After . so equa%ion (7.31) for ( i = 1,2,... N) loss allocation to the load at node j is then 13 I.

(7.33~

We can use a similar approach to foimulate the problem of allocating the loss to generators. Based on the discussion above, we may conclude that to solve the di factor or loss allocation pro~lem,we should first build and solve the linear equation (7.26) or equation (7.3 1). However, this approach is not ef~cientand not flexible, We will develop a very simple and efficient algorithm by means o f graph theory in the next section.

Power System ~ e ~ ~ cand~~ er r ei ~nu l ~a t i o ~

25

graph is a directe graph. At this stage, the direction of each the direction of its ctive power flow. Each b r ~ c has h its initial while each node has its outgoing lines and ~ n c Q n ~lines. i n ~ The number of outgoing lines at node i is denoted by d+(i),the n ~~entioned above^ the set of outg~inglines is den lines by r-(i). A directed path is formed along the direction of d terminal node o f a directe are identical, we to denote the resistance, r ~ a ~ ~ a n c e , er flow of branch k, and If the

follow in^ reIationship holds for each branch alon

T in a load flow ~~~h~

then there exists no directed circuit in the graph. e use the methodo~og reduction to absurdi~.If the~eexists a follo~vingrelations dB(k, = 0

(7.~4~

keC

where AB(,, i s the phase angle ~fferenceb e ~ e the e ~two nodes of bran~hk, and can be

i

(~.35)

equation (7.35) take the values at the terminal node of branc k ) , then

b6,k) > 0, and

exist in this situation. ns. In case there i ch is certainly negl When a directed graph has no dir~ctedcircuit, there are at 1 tidy d+(i)= 0,and d- ( j )= 0 respectively. ~ s s d,~ (i) ~> 0e holds for all nodes, i.e. each no e has at least one out out from any node q , we can travel to the n ~ x nt from n2 we can travel further to n3 by similar reasonin Thus there are only two p Q s ~ ~ boutcomes: le one is that we ~ a v in e ~ m p o s s i ~for ~ e a finite graph; the other is that there exist dire the c ~ ~ d i tofi othe ~~eorem.

3

Electric Power Industry R e ~ t ~ c in~ China ~ n g

larly, we can prove the other half o f the t~eorem.~ o m b ~ i n g

h, there exists at least one node without an outgoin and one node without an i n c o ~ line. ~g

e i is a node with d-(i) = 0 on a load flow gra~hs.The ~rocessof d its out go in^ lines r+( i ) is called the e l i ~ ~ ~ ~ afor~noi n ~ p ~ ow graph, a111 branches can be eliminated t ~ o u a~ recu~sive h

raph by Y, and the b r ~ ~ h at least exists one node i, ~ ~ a out an ~ e l i ~~i n a t i nnprocess ~for node i l , we get $ubgraph El'( cted circuit. Hence, there exists at can carry out an el~minationproce so on. Thus we can e ~ ~ m i ~ all a t ebranches by a finite (less than elimination ~rocess. lain the e~iminationpr~cessby a simple example, as raph has no directed circuit, and d - (1) = lines 1, 2 and 3. A ~ eliminatin e ~ 7.9b, in which d"-(2) = 0. The e l i m i n ~ node t ~ 2 and its out and its out go in^ line 5, and thus w the above e~iminat~on proc cessively e l ~ r n ~ ~ m e s s can also be carried out by $uc~ess~ve~y eiim~atingthe node with d+(i)= 0 its ~ n c o lines ~ ~ ~ g The co~esponing d e ~ n i t i o nand ~ t h e o r ~ mare ~ s i m i ~ ato~ the discussion above.

e(i).

~ r o b ~ofe load ~ s Raw anaiysi below, we will use PDF and P location problem respec~ively~

Power System ~ e s ~ c ~and i n~ ge r e ~ ~ ~ t i

i by equation (7.25).

LA: Calculate loss allocation to the load of node i by e ~ ~ a (7.33). ~ ~ o n Do the follow^^ for all j , ij E (i). F: Transfer the power of each generator at node i to nodej

r+

bji is defined by equation (7.27).

the loss of node j , Pj,according to Equation (7.3 1). indicating that the node has been eliminated. . Search for the next node without ~ c o m lines, ~ ~ until g all are ~limina~ed. can §imil~lyin~oducean a l g o ~based t ~ ~on e ~ i ~ i n a t i n~gn ~ o m i nlines. g In this case, cal~ulatedresults for PDF are the d i s ~ i b u t i ~factors n of ~ r a n c ~ used e s by loads; for losses are allocated to generators. ~e~ no~hwestpower n e ~ o r kis calcu~atedby the ~ r o p o aigori cost. The data used in the case study is real exchange p i provinces dated I6 January 1998 as shown in Figure 7.10 a, The n in T ~ ~7.7 the wheeling cost paid by Cansu province as l eand Figure 7.10 b. The calculation results are the wheeling cost paid shown in Table 7.7 and Figure 7.10 b. In Table 7.7, sum of wheel y’ in the table refers to the energy loss cost, ‘Capaci erator capacity to compensate power loss, while ‘Line9refers to ~ a n s ~ i s s i o n , The average wheeling cost of the day is 0.0247 yuanlktlrh.

.7 This chapter has described the Chinese power market that is an embryo which the state retains ownership of the generator§ and some of i n f ~ s ~ cbut ~ is ~ ~pening e , up the market to limited c ~ ~ p e t i t iEle ~n. transmission loss methods have been proposed and examples of a simplified Chinese power system have been used to demons~atethe advantages derived from such ~ e t h o d s .

s §uppo~edby the ey Project of ~ a t i o n a~cience ~ ~oundationof China. The authors would also like to thank IEEE for granting permission to reproduce the ~ a t ~ r i aclo§~ t a ~ n in e dreference [ 181.

Electric Power ~ n d u s structuring ~ in China

2

%e7.7 Wheeling cost for Shaanxi and Qinghai power exchange Hour

Exchange Power

Loss of Wheeling

Line Using Cost

1 2 3

CMW) 273 339 279

WW> 10.29 13.29 7.68 11.63 11.83 14.80 7.52 0.1 1 3.37 5.58 5.25 5.84 1.71 3.34 2.15 2.66 7.46 8.29 0.83 0.00 1.88 3.00 0.00 0.75

(Pan> 4410 4510 2410 3630 4200 5640 2920 1080 2160 2170 1870 3760 2370 2660 2400 3280 4090 3130 420 0 I220 2700 0 1100

12 13 14

33 1 325 335 268 101 “122 -226 -172 -260 -165 -189

15

-143

16

-197 -293 -305 - 30 0 -116 -210 0 154

4 5

6 7 8 9 10 11

17

18 19 20 21 22 23 24

[I] [2] [3] [4]

m e l i n g Cost ( p a n I kWh) Energy 0.01 1 0.012 0.008 0.01 I 0.01 1 0.013 0.008 0.001 0.008 0.007 0.009 0.007 0.003 0.005 0.00s

0.004 0,008 0.008 0.008 0.000 0.005 0.004 0.000 0.001

Line

Sum

0.006 0,005 0.004

0.016 0.013 0.009

0.004 0.005 0.006 0.004

0.011

0.033 0.030 0,021 0.026 0.029 0.036 0.023

Capacity

0.013 0.0 I7

0.001

0.011 0.0 17

0.004 0.003 0,004 0.003 0.002 0.003 0.001 0.002 0.003 0.004 0.004 0.000 0.002 0.002 0.000 0.001

0.018 0.010 0.01I 0.014 0.0 14 0.014 0.017 0.017 0.014 0.010 0.014 0.000 0.01 1 0.013 0.000 0.007

0,019

0.830 0.020 0.024 0.024 0.019

~ . ~ 2 2 0.023 0.023 0.025 0.022 0.026 0.000 0.018 0.019 0.000 0.009

S.Q. Gao and P.L. Chi, Several Issws Arising During the Retracking of the Chinese Economy, Foreign Language Press, 1997, J.P. Sun, Electric Power Industry in China 1999, China Electric Power Information Center. W. Sweet and M. Hood, ‘Can China consume less coal?’, IEEE Spectrum, Vol.36, No.11, November 1999, pp.39-47. M Hood and W Sweet, ‘Energy policy and politics in China’, IEEE Spectrum, “J01.36, No. 1 1, November 1999, pp.34-38.

. .-

0.04

0.03

0.0

0.01

0

9 10 11 12- 13 14 15- 16 17 1 - -..

~

~ e e l ~ ncost g for $ ~ a ~ and xi

[S]

the Electric P o w e ~ I ~ d ~1998, s t ~p ,p . ~ ~ 6 " 1 4 1 ~ Shi Yubo, 'Take vigorous action to promote power industry's reform and ~ ~ v e ~ o p ~ ~ ~ i Power ~ f fE ~ ~ e~ ~Q ~ ~ ~ Qs No.1, ge e 1999, ~ e pp.7-8. ~ ~ ~

Electric Power I n d u s ~ e $ ~ ~ c ~in rChina ing

7

[lO] Zhang Shaoxian, ‘Clear up reform i as and initiate a new chapter o f p ~ f e s s i o n a ~ ~ ~ a g e m e n tChina ’ , Power Enterprise ~ a n a g e ~ e nNo. t , 1 1, 1998, pp.4-5. [ l l ] Ciao Uan, ‘On the second step reform o f the State Power Corporation o f China’, China Enferpp~~e ~ a n a ~ e ~N0.2, e n ~1998, , pp.4-5. [12] Lu Yanchan~,‘A ~ e ~ ~ d e ~ tabout a n ~then simulated g power market practice’, Chinina Ente~prise~ a n a ~ e ~No.4, e n ~1998, , pp.8-9. [I31 Wang Yoii~ian, ‘Strive to accomplish two reform in thee years and basidly feaIize equal China Power Enterprise ~ i n a n a g ~ ~ No. e n t 12, , 1998, pp.16-18. books on energy pricing’, IEEE Spectrum, Vo1.36, N0.12, ~ ~ c ~ m b e r 1999, pp.59-63. sector decision making in China’, IEEE Winter Power ~ e ~ t ~ n

rprise ~ ~ n a g eNo. ~ 1,z 1999, ~ ~ pp. ~ 16-1 , 8.

[21]

pp.~405-1413. CXgiRE Task Force 38.04.03, ‘ ethods and tools for transmission costs’, Elech, No. 174, ~ c t o b e 1997. r services by the end user Case n p r o ~ d i~ngt e r c o ~ e c t eopera~ions ~ tioopral Science ~ o u n ~ a t ~ Workshop, on Nove~ber1996. aliana and Mark Phelan, ‘Al~oca~ion of transmission losses to in a ~mpetitiveenviro~ent’,IEEE Transac~ionson Power stem^, Vol. 15, No. 1, ~ ~ 2000, pp.143-150. Tomas G o ~ a l e zGarc~a,and er losses’, IEEE Xra~sactionon nd load distflb~tionfactors for supple me^^ c h a r ~ e o n Power s S y s t ~ ~Vol. s , 12, No.3, a l l ~ a t i o nin ~ ~ s ~ i s sopen i o naccess’, IEEE ~ r a n s a ~ t ~ on 1997, pp.l189-1193. L.L. Lai, J.T. Ma, N.~ a j ~A.~~ a~ ~dand ,a , to ~ompu~~tional efficient a l g o ~ ~ h mfor s ~ a ~ s m i s s i o10s n s’, ~ n ~ ~ r n a t i ,Iournal o n a ~ afElectric Power and Enerm Systems, Elsevier Science Ltd, Novem~er

b

Prof. Vijay K. Sood Canada

power In recent years, major changes have been introduced into the s ~ c ~ofr electric e utilities all over the world. The reason for this was to improve ef~ciencyin the the power system by means of deregulating industry and opening it e~tion.This is a global trend and similar ctural changes have o c ~ u ~ el~ewhere ed in other industries, i.e. in the teleco~unicationsand air~ine~ ~ s p o r t a ~indus~ies. ion The net effect of such changes will mean that the ~ansmissi5n~ generation and dis~ibution syst~msmust now adapt to a new set of rules dictated by open r n ~ ~ eIn~ s . trans~~ssion sector of the power utility, this adaptation may require th ~ o d i ~ c a ~ iofo ninterconnections between regions and countries. further more^ the ptation to new generation patterns will also necessitate a ~ p ~ t i and o n require in~reased xibility and availability o f the transmission system. Addin to these problems has been the growing env~ronmen~al concern and constraint upon he righ~s-of-way for new i n $ ~ l a t i ~ and n § facil~ties.Yet further d e m ~ d are § c o n t ~ u a lbeing l ~ made upon u t ~ ~t i ~ ~ e supply increased loads, improve reliability, delivery energy at with ~ ~ ~ power ~ ~ quality. v The e dpower industry has respon the ~ e c ~ o l oof~ flexible y AC ~ansmission systems or e n c o ~ ~ ~ sas whole e$ family of ower electronic controll achieved maturity within the industry whilst some others are as yet in the design stage. FACTS have been d e ~ n e dby the IEEE [4] as:

Flexible AC Transmission Systems (FACTS) A power efectroiiic based system and other static equipment that provide control of one or more ac

transmission system parameters to enhance con~ol~abi~ity and increase power transfer capability. For m a ~ ~ u f a c ~ rofe relectrical s equipment, this challenge provides an o p p o ~ n i t yto build equipment that is reliable, flexible and relocatable since planners now d e m ~ dr adaptation to c h a n g ~ gsyste FACTS rely, to a large upon advances made in power electronics (PE) and microprocessors. The PE tec ,well known in low-power industrial applications, has now migrate^ to hi~h~power utility applications because of the economical availability of reliable high-power switching devices (i.e. thyristors, GTOs and IGBTs). Note that developmen~sin other related areas such as communication systems (using fiber-o etc.), super conducting materials for energy storagc and metal oxides for surge arrestors will also play important roles in the continuing growth of FACTS applications. This t e c ~ o l o g ywill impact on all aspects ofpower system operations, for example, in: generation systems (i.e. from hydro, thermal, wind or photovoltaic means), storage systems (Le, by conversion of energy from AC to DC, DC to AC, transmission systems (i.e. by the rapid control of system parameters such as voltage, current, imp~danceand phase angle), dist~butionsystems (i.e. by the rapid circuit or current i ~ t e ~ p t i ofor n purposes), and consumer systems (i.e. by the power conditioning of consumable energy). F a ~ i c u l ~for l y transmission systems, FACTS technology offers the f o ~ l o w i n~ossibilit~es: ~ e

Greater control of power, so that it flows on the prescribed ~ ~ ~ s m i s sroutes. ion Secure loading (but not overloading) of transmission lines to levels nearer their thermal limits. Greater ability to ~ r ~ ~power f e r between controlled areas, so that the reserve margin typically 18% may be reduced to 15% or less. Preve~t~on of cascading outages by limiting the ef3ects of faults and equip men^ f a i l ~ e . ampi in^ of power system oscillations. ~

6p

-

Static var compensators (SVC) is an example of a mature FACTS applica~io~. Other more novel ap~lications(i.e. STATCOM, UPFC) are being developed and tested to provid~ increased flexibility, enhance stability arid transmission capacity in the operation of power systems. The present environment of deregulation and constraints on building of more tra~smiss~on ~aci~ities provide compelling reasons to develop FACTS c o n t r o ~ ~The ~~s~ ~ p r o v e m e n tof a deteriorating power quality will be an additional focus for FACTS controllers of the future,

8.1.1

Benefits of FACTS Technology

The two main objec~vesof FACTS controllers are:

*

to i n c r e a ~the ~ power transfer c a p a b of ~ tr~smission ~~~ networks, and

260

Power System Restmcturing and ~ ~ r ~ ~ l a

ravide direct control of power flow over des~gnated~ r ~ s m i s s routes, ~on ible AC system owes its tighter transmiss~oncontrol to its ability to m ated parameters that constrain today’s AC systems, ~ ~ c l u d i nsg e, phase angle and the ~ c ~ ~of~osc~~~ations n c e at various ~ e q u e n c ~below es the rated ~ e ~ u e n c y .

~ Power flow over a transmission system is limited by one or more o f the f o ~ l o w i n[4]: system stability, loop flows, vol~agelimits, 1 limits of either lines or terminal equi hart circuit level limits. itations on power transfer are primarily inter re~atede ~ e c ~ parameters ca~ including vo reactive and ~ t i v power. e ~igh-speedcontrol of any one or more of these parame~erswith E controllers will enhance the value o f AC transmission assets. liminary studies of several ~ y s t e have ~ s shown that FACTS controllers can provide economic sol~tionsto some of these p ~ o ~ ~ e m A $discussion . of each of the above-~entio~ed iimi~tionsis provide^ next.

This r e q u ~ e sthe power system to retain a margin o ~ystemand still main~ains ~ c ~ o n iSin s~. nd are able to eontrol the for avoiding the addition ofne traints [4] could be hrther spli i ~ i t ~ few a l seconds after a major p rove the performance by the use of, s exci~ationsystems and the implementation t ~ m p i concerns n ~ the ability of a ns once initiated by a small disturb include power system st

er system to ~ a i n t asi y~n c ~ o ~ i sfor m ion. A n u ~ b e rof e~ample§ lled series capacitors, high

it describes the situation when the next i n ~ e ~ ofe load ~ t causes a voltage collapse in the power system. This v o l ~ g ereduc~onis g occu~ingover time periods ranging from many secands that are used to improve VSL, include operator action, a rs ~ o ~ p e n s a t i ogenera n, to^ or sync nous ~ o n d e n s ~and

Flexible AC T~ansmissionSystems (FACTS)

-sync~onousresonance (SSR) is due to interactions b e ~ e e nthe seriescompensat~dAG power ~ansmissionsystem and torsion v i b ~ t i o n g e n e ~ t o runits. This issue is dealt with by cons~ainin that desired for c o ~ ~ ~ n sp ae t~ ~i ~ to e~ disafe limits; usually this level is system security. A p p r o ~ h e sthat are used to improve SSR condi of series capacitors d u r i n ~unsafe operation^ passive series blo tor exci~ationor SVC on the generator bus. Ge lied to c o ~ any e ~u n e ~ p ~ c t contin~enci~s. ed

8.2.2

Loop Flows

ows occur as an unwanted result of the operation of the interconnected ~ a n s m ~ s s ~ o n are dictated by e l e c ~ i c acircuit ~ laws (i.e. Ohm’s and ~ r c h h o f laws). ~ s These at steady state where the undesired loading affects the v n of thermal or stability limits, These effects are address or by series capacitors. The new FACTS c o n ~ o ~ l e r s ver, since speed of opera~~on is not a major c o n c e in ~ this problem, c o n ~ o ~ ~will e r sbe justified onIy if .frequent a d j u s ~ e n are ~ s require

o ~ by a c o ~ b i n a t ~ oofn genera~orreactive V o ~ t a c~ eo ~ is~accomplis~ed ent, fixed or mechanica~ly s w ~ ~ c ~reac~o~s~ca~acitors ed and m e c ~ a ~ ~ c a ~ On ~ a n s f ~ ~ e~~~t r s . reactive equipment is used for coarse control while the ~ e n e ~ ~ ~ o r prov~dev e ~ i con~ol. e~

Thermal limits are inherent in ~ansmissionsystems owing to both line c o ~ d u c ~ o ~ ~ series equipment (i.e. ~ a n s f o ~ e rreactors s, and series capacitors). Trans~issionlines ope~atedbelow these limits to provide s e c u in ~ the ~ event of a role of FACTS c o n ~ o ~ ~will e r sbe to use this inherent thermal capacity in a more e f ~ c ~ $ n t and secure manne~.

8.2.5

High ~ ~ o r t - ~Level i ~ Limits c ~ ~ t

The p r o ~ ~ eofmexce$sive s h o ~ - c i r c ulevel ~ ~ can be quite difficult and expen§ive to c o ~ e c ~ dition is made to the ~ a n s m i s s i ~system. n This can result in sho~-ci~cuit levefs c r ~ e ~ i up n g in sub-~ansmiss~on equipment.

262

Power System Restructuring and

Y The IEEE definition of a FACTS controller is: A power electronic based system and other static equipment that provides control of one or more ac transmission system parameters.

The technology concerning FACTS is well known in the low-power industrial applications field, but is relatively less well known in the utility power field. This technology is intimately concerned with developments in the follow~ngtwo areas [S]: Power electronic switching devices and pulse width modulated (PWM) converters. Control methods using digital signal processor (DSP) and ~icroprocessortechnology. Developments in both areas are advancing rapidly, and need to occur further before a~plicationsin the power utility field appear econo~callyattractive. App~icationsof PE in the power utility field still need further research in the following areas: active harmonic filtering and reactive/active power support, single-node or area-wide application, c~mpensationof non-linear loads, and transient performance of the controller.

8.3.I

Power Switching Devices and P WM Inverter

Of the switching devices presently and potentially available within the near future (next >,the gate turn-off (GTO) thyristor and IGBT are the most promising. However, in the longer future (10 years), competition for these switching devices will occur from o nthe various ~ o w e r - s w ~ ~ h i n g ~ ~ S ” c o n ~ othyristor 1 ~ e d WCT) devices. A ~ m p a ~ s of devices is presented in Table 8.1. wever, owing to the higher switching losses in G devices, the ~ a x ~ e , to ng frequency operation is limited to less than about 1 z. F u ~ e ~ o rowing the switching and drive characteristics of the device, it has been feasible to operate devices in parallel for high power applications. Some limited success in the series op~rationof devices has been reported, but again this remains a l~mitation~ r increasing the rating capability of a FACTS converter on appears to be the use of several converters op hing frequency presente~to the total filter can als shifting the switching functions of individual inverters, and by converter ~ a n s f o ~ e rAs new , possibility exists with the use of ~ulti-levelconverters.

3

Flexible AC Transmission Systems (FACTS)

Max. voltage rating (V) Max. current rating (A) Voltage blocking Gating Conduction drop 0') Switching frequency

8000

Th~~~st~r 6000 1700

4000

6000

Sym./ Asym. pulse 1.2

Sym./ Asym. Current 2.5

ThyristQ~ 2500 3000

800

800

Asym.

Asym.

Voltage 3

Current 4

400 Sym./ Asym. Voltage

1000 100

Asym.

1 .a

Vo~~ge Resistive

1

5

20

20

20

100

10000

10000

3500

5000

5000

2000

8000

8000

2000

2000

2000

200

(kw ~evelopment target max. voltage rating (V)

Development target max. current rating

4

GTO

: Gate Turn-off thyristor

IGBT SI MCT MOSFET

: Insulated Gale Bipolar Transistor : Static Induction thyristor : MOS-controlled Transistor : MQS Field-effect Transistor.

Two versions of switching converters are feasible depend storage device utilised is an inductor or a capacitor. When the storage device is an inductor, the converter is called 8 current source converter (CSC); when the storage device is a capacitor then the conve~eri s called a voltage source conve~er(VSC). A n~ticeable change in converter topology usage will be the increasing use of VSCs instead of CSCs used in traditional HVDC transmission. The VSC will find applications in advanced static var co~pensators(ASVCs), active filters, S T A T C Q ~ etc. ~ , The main reasons for this change are that VSCs are smaller and less expensive than CSCs; ~ ~ h e ~ oVSGs r e ,are expandable in parallel for increased rating. A brief comparison between VSCs and CSGs is given in Table 8.2.

Power System Restructuring and Deregulation

264

le 8.2 Comparison of current source versus voltage source converters

Current source converters Use inductor L for DC-side energy storage

Voltage source converters Use capacitor C for DC-side energy storage

Cons~antcurrent Fast accurate control Higher losses Larger and more expensive More fault tolerant and more reliable Simpler controls Not easily expandable in series

Constant voltage Slower control More efftcient Smaller and less expensive Less fault tolerant and less reliable Complexity of control system is increased Easily expanded in parallel for increased rating

T r a ~ ~ ~ ~power o n a lconverters used line-commuta~edthyristors as their active switch~ng elements, but next-generation converters will exploit self-commutated CTO thyristors in the near-term future, and will probably exploit lGBT and/or MCT devices in the long- re^ future. The basic PE building blocks will comprise either the:

e

anti-paraIIe1 thyristors which will be used to control irtduclivelcapacitive i ~ p e d a n ~ e s , or six-pulse CSC or VSC unit, employing multi-level operation (with or without multihase ~ ~ s ~ ~to increase ~ e r thes pulse ) number (up to 48 pulses), to reduce ~ a ~ o n i c ~eneration.The basic switching elements will be the anti-parallel G ~ ~ ~ ~ori o d e IGBT-diode unit,

Control ~ e ~ and h Do~ P~/ ~ ~ i c r ~ ~ r ~ cTechnology essor Control eth hods based on either the time or freque~cydomain are feasible. These ~ e q ~ i $ e i n s ~ n ~ n e o monitoring us techniques and complex computation of switching ~ n c t i o n sfor the firing of the converter switches. A comparison of the control methods in the two d o ~ a i is n~ made in Table 8-3. Comparison of time domain versus frequency domain comp~nsation

~ r e q u e ~ dc ~y m a i n Fast response

Easy to implement Computa~ionalburden is low Ignores past periodic characteristics

Slower response Complex measu~ementsand analysis Computational burden is high Depends on periodic characteristics of d i 5 ~ 0 ~ i o n

wing to the complex switching functions required and the c o ~ ~ ~ ~ ~ tburden ional necessa~,exten~iveuse of DSPs and microprocessor technology will be required in a power system environment. Utilities have some experience with HVDC ~ e c ~ o l oSVCs g~, rotection relays which use microprocessor-based controls. However, the application of FACTS devices is likely to be at a greater level of complexity than anything

Flexible AC ~ r a ~ ~ ~ i sSystems s i o n (FACTS) known previously within the utility environment. This will require careful considera~ions of r e l ~ a b i land i ~ ease of use within the utility environment.

8.3.3

F ~ e s ~e ~ ~t on ~ACT^~ Activities u ~

EPRI of the USA has been promoting a program (EPH Project 3022) on FACTS for some years [ 11. A number of special conferences on this topic have been o r g ~ i ~ by e dE these conferences comprise, by far, the largest effort on FACTS-related literature. Since the last five years or so, IEEE and CIGRE working groups have also become involved and pub~~cations are being reported in their literature also. FACTS have been with the power industry for many decades in the form of SVC and other applications. However, it is only recently that these applications have become classified under the b ~ ~ a d - bheading a s ~ ~ of FACTS controllers o f the power system. FACTS technology is not a single, high-power electronic controller, but rather a collection of controllers, which can be applied individually or collectively in a specific power system to control the intenelated parameters that constrain today's systems. The thyristor (either line or sclf commutated) is their basic switching element; however, in one particular application called the interphase power controller, no active switching device is used.

8.4.P

~

~

n Concepts ~ ofa~~ansmission ~ e ~

~

~

~

A simplified example of power flow in a loss-less transmission line, with inductive impedance X,, c o ~ n e ~ ~two i n gac systems with voltages V , and V, is shown in F i g ~ r e8.1. The transmi~edpower P is given by equation (8.1) and also shown in the figure. From equation (8.1) it is evident that power flow can be controlled by varying 5, Y,, X, or the angles 6, and 6,: P = (V,. VJX,) sin(&,- 6,) (8.1)

.1 F u ~ ~ d ~ r n eonf AC ~ l spower transmission

n

Transmitted power P can be regulated by control of any system parameter by a FACTS controller, or any combination o f controllers, as indicated in Table 8.4.

Power System Restructuring and Deregulatbn

66

.4 Control of system parametersby FACTS controllers

Voltages V, and V, ~mpedanc~ X, Angles 6 , and 6,

Shunt Series Phase angle regulator

SVC, STATCOM TCSC, IPC

TCPAR

The FACTS app~ica~ion§ have been split into the fo~~owing categories d their mode of operation: Control~erswhich act in shunt to the ~ a n s ~ s s i system. on Controllers which act in series to the tra~smiss~on system. Controllers which act in a s e ~ e § / s ~ ucombination. nt Controllers which alter the phase angle between voltage an A special category which encompasses HVDC controllers and any remaining controllers. Details of these various c a ~ g g o are ~ ~provided s in the following sections.

8.4.2

Shunt Controllers

reactive power com~ensationsince the mid t for arc furnace flicker compensation and then in power ~ansmissionsystems first 40 MVAr SVCs was installed at the Shannon Substation o f the ~ i n n e § o t ystem in 1978. At present some 300 SVGs with an installed capacity of 4~,000 s are in service all over the world. The SVC results in the ~ o ~ ~ o bwe inne ~~ [S]: t~ voltage support, and transient s ~ b ~ lim~rovement, ~ty power system oscillation damping. A l ~ ~ o u gmany h versions of SVCs exist [9] (i.e. variants are TSR, TG common one (Figure 8.2a) usually employs (either thyristor or mec~a~ically) switch capacitors and ~ ~ ~ s t o r - c o nreactors ~ o l ~ e(TCRs). ~ With an appro the capacitor switching and reactor control (Figure 8.2b), the var continuo~sl~ and rapidly ~ e ~ ~apacitive/inductive ~ e n values. It maintains the steady state and dynamic voltage at a bus within bounds, and has some ability to control stabili~,but not much to control active power flow.

267

Flexible AC Transmission Systems (FACTS)

%“=Capacitive

Ratlng

(a)

Conanuous Inductive R%ng

(b)

fig^^^ 8.2 (a) The SVC and (b) its V-I characteristic

S ~ a t ~ cQ n ~ ~ (ST e ~ s ~ t o ~ The S T A ~ aCsolid-state ~ ~ ~ voltage source inverter coupled with a transformer, is tied to a transmission line. A STATCOM injects an almost sinusoidal current, of variable m a ~ i t u d eat , the point of connection. This injected current is almost in quadrarure with the line voltage, thereby e ~ u l a t i ~ang inductive or a capacitive reactance at the point of connection with the transmission line. The hnctionality of the S ~ A T C Omodel ~ is verified by regulating the reactive current flow through it. This is useh1 for r e g u l a ~ ~the n~ fine voltage. An advanced static var compensator (ASVC) [ 101 using a voltage source inverter (VST) is shown in Figure 8.3a and its V-i characteristic is shown in Fi 8.3b. The VSI is storage capacitor to generate an output AC voltage V,. When V, equals AC bus, the VSX draws no current; when Vo > V, the current drawn by the leakage impedance of the transformer is purely capacitive. On the other hand, when V, < V then the c ~ e ndrawn t is purely inductive. The ~ n c t ~ o npael ~ o ~ a n ocf ethe A superior to the t r ~ d ~ t i o nSVC. a~ Tran smmion 1ine

pu Voltage Couptiiig *amformer

DC storage capacitor

(a)

Figure 8.3 (a) STATCOM application and (b) its I/-I characteristic

Tram

ive Rating

Power System ~

2

The ~

e

5

~ and c ~ e~ r ~n ~ ~l a ~ i o n

~ isValso C superior to the conventional SVC for the fo~lowingreasons:

eduction in outdoor area requirement, since it replaces the ~ o l u r n ~ ~ o ~ i s capacitor/reactor banks associated with a conventio~alSVC. roved dynamic p e r ~ o ~ and ~ c enha~ced e s ~ b ~due l ito~its a ~ ~to i~ ni ~~r e a s ~ siently tbe var generation. roved p e r f o ~ a n c eat low operating vo~~ages down to by t r ~ $ f o ~leakage). er Reduced need for AC filters. ve r e f ~ e dto the ~ ~ O - ~ system a s ~ as d functional operation of this device is, howe ous condenser, but without the slow response t i n e ~ i aand so it was briefly own as the ‘Static Sync ent practice is to refer to these as STATC generates a three-phase volt a reactance. When the AC (lower) than the bus voltage, the current flow is cause eS how ~ u c h ~ u ~ flows. e n ~This allows the control of

on and to prod~cepra~ticallysinus0 is s h o ~8.3b.Tlie ~ ~STATC ~ i

~

re effective than the SVC in providing voltage support

~

269

FIexibIe AC ~ ~ s m i s s i Systems on (FACTS)

Acts as a voltage source behind a reactance

Insensitive to transmission system harmonic resonance Was a larger dynamic range Lower generation of harmonics Faster response (within ms) and better performance during transients Both inductive and capacitive regions of operation possible Can maintain a stable voltage even with a very weak AC system Can be used for small amounts o f energy storage Temporary overload capability translates stabilih, into i m u ~ o ~ evoltaae d

Acts as a variable susceptance Sensitive to transmission system harmonic resonance Was a smaller dynamic range Higher generation of harmonics Somewhat slower response Mostly capacitive region of operation Has difficulty operating with a very weak AC system

,based on IGBT switches, and capable of operating at s w i ~ c ~ i n g been developed. The core parts o f the plant, compris~n s, control system and the valve cooling system, are fitted into a of 10x20 rn. The outdoor equipment i s limited to heat utation reactors and the power transformer. A rating o f rl: 100

container with a exchangers, air-c Mvar per converter i s availab~e~ in case of increased rating, multiple units can be operated in parallel. The modular design makes it easily relocatable to another site when desired to meet hanging s y s t e ~needs. ~ The response time of this unit i s very fast (about o n e - ~ u ~ e r cycle). As a result of its high switching frequency, the plant can operate without h a ~ o n ~ c filters, or may only require a small high-pass filter. The risk for resonant condit~onsis heref fore negligible. ~ u r t h e ~ o rthe e , possibility of active filtering of h a ~ o n i c salready p~esei~t on the network makes this an attractive choice.

mat~hingthe ~ ~ r ~ i mechanical power and the generator electrical power during system faults. This can be ~ ~ a series or shunt braking resistor. Shunt resistors are p r ~ ~ e r a b ~ e done by i n t r o d u c ~either because they are less expensiv~and easier to coordinate in a system with any and lines. Moreover, a §hun~-connecte~ thyristor-controlled resistor with a radial transm~ssionline can be used effec~~ve~y to damp power swing oscillations [13] in a transmission system. These systems are esigned to provide post-fault AC system speed control by compensatin~for fault accelerating power by dissipation in a shunt resistor. A pair ofbackto-back ~ y ~(Figure ~ s8.4)~ does o the~ application of the shunt resistor, The application of braking resistors should take place as soon as possible after fault detection and they should not be switched out until the derivative of the swing curve becomes negative. The ~atingof

Power System R ~ s ~ and ~ D c~ ~~l a nt i o~ n

70

the resistor should be such that the kinetic energy injected by the fault sho~ldbe before the generators slip the first pole.

-& .4 Dynamic brake application

The ~eliabilityand effectiveness of braking resistors have been demons different projects:

PA's Chief Joseph substation, 1400 MW, 3 seconds, 230 kV system; C Hydro's G.M. Shrum substation, 600 MW, 20 seconds, 138 kV sys~em; 3. Argentina El Chocon Project. asically LT changers regulate the output voltage when subjected to variations in the input voltage due to c h ~ g i n gsystem conditions. M e c h ~ i c aversions ~ were used widely in the i n d ~ for s ~many years. These mechanically operated load-tap hanging transfo now have t h ~ § ~ o r - o p e r a ~switches ed (Figure 8.5) to do the same function faster [14]. This permits the improvement of system stability and damping of the power system osci~l~tions.

High speed static tap changer

use a super conducting coil acts as a b ~ f f e rbetween the power generation and load consum~tionand aids in the load levellin~and matching (within a few cycles) of the two, enabling a greater control and flexibili power system [IS]. The benefits o f energy storage systems are offset by the losses of s t o ~ n genergy. The round-trip efficiency of a SMES system is claimed to be 90%. The SMES coil is fed by a current source GTO inverter h m the AC n required, the SMES can supply transient active or reactive p o w e ~to the AC supply to support it. The technology holds promise for improved en 9 d flexibility to meet peak utility system loads. A multi-temi P to act as a power flow control device also [6,7]. A fairly rec

Flexible AC Transmission Systems (FACTS)

71

suggested the use of a SMES system for SSR damping of turbine generator units. A SMES unit has been in com~ercialuse on the BPA system.

I Superconducting Coil

~ i ~ M 8.6 r e SMES operating principles

imilar to a large uninte~ptiblepower s~pply.A VSC connects the DC battery to the AC system. Such applications provide load-levelling benefits and act as a spinning reserve on islanded networks. Modulation o f the 10 MW BESS at Chino has increased the transfer capability from Arizona to California by several hundred megawa~s. Battery storage has been applied at a number of locations including: an 85 ~ ~ / minute 3 0 system in ~ e ~in~1986, i n a 10 MW/4 hour station commissioned in Chino, S . California, in 1988 [17], and a 20 MWI4 hour station commissio~edin Puerto Rico,

8.4.3

Series C o ~ t r ~ l l e ~ ~

g AC lines for increasing line loa~bility has been known for a long time. Adding a isto tor-controlled series capacitor ( however, is a more recent pheno and provides greater flexibility in ission line impedance continuously ~ a n s ~ i s s i oAn .TCSC can vary the below and up to the line’s natural to force power flow along a ‘contract path’. The advantages of the TCSC are: ability to mitigate sub-sync~onousresonance (SSR), ability to balance three-p~~se power flow, ability to control power flow flexibly, ability to reduce short-circuit currents by rapidly controlling the capacitive to ~nductive impedance, and ability to damp power system oscillations. The controlled series compensat~oninstallation will likely have two key componen~ (Figure 8.7). One element will be the mechanically switched portion, and the second will be the thyr~stor-con~rolled portion. The relative sizes of the fixed and c o n ~ o l ~ emode d portions will vary with appl~cation.The TGSG portion i s made up of a number of small series-connected modules. Each module is either inserted (with the thyristors blocked) or b ~ ~ (with s ethe ~ y r i § t o ~fully s conducting). In this manner, a stepwise ~ o n ~ iso l

Power System ~ e ~ t ~ ~and~ Deregulation r i n g

27

ach~evedwith minimal losses and harmonics. There also exists the possibi~ityof op~ra~ing in a vernier mode where partial conduction of the lhyristor path during each-kalf-cycle is used to circulate inductive current through the capacitor and boost its effective ohmic value. One advantage of such small-signal ~ o d u ~ a ~isi othen control of S Transmission line 5% sections

Breaker switched

.7Thyristor-controlled series capacitor

A new control scheme with a TCSC [IS] indicates that a method of modula~ingthe firing angle can be used to boost the series capacitor voltage and virtually eliminate the possibil~~y of SSR oscillations. A phase-locked loop (PLL) is used for synchron~s~ng the thyristor firing with the line current rather than capacitor voltage for a more stable operation.

~ ~ n t (I r ~ ~ ~ e ~ IPC [19,20] does conlain any PE equipment, it is included here as a FACTS device that can aid in the ma~agemcfltof power flow between two synchronous t ~ gs u s c e ~ ~ n c e s , s y s ~ e ~The s . basic IPG consists of a series-connected device c o ~ ~ r i s itwo one inductive and the other capacitive, subjected to properly ~ h a s e - s h ~ ~voltages. ed Thus, whatever the angle 6 at the TPC terminals, some of the cQmponentsare always subjected to voltage. By adju§ting the value of these compo~en~s, it is always ~ o s s i bto l ~force t in each of the networks even if the a n ~ I eat the t e ~ i n a ~ is szero. When all t set in one o f are energised, the ampIi~deand phase angle of the c u ~ e nare s to wh~chthe IPC is coniiected. This current contro~thus en d reactive power through the device. any types of IPC are possible and each type can have d i f f e ~ ~con~gurations~ nt In one type cai~edthe IPC 120 (Figure 8 . Q the voltage phase shifts are achieve^ with a crossconnection between phases using an inverting transformer to reduce the voltage ~ a g t ~ ~ ~ d applied to the reactive compQnents.One practical appl~cat~on o f such an ~ n s ~ l ~ a has ~ion appeared in ~ e ~ oUSA. n t ~ ~~~~~

group of three-phase reactors and c a p ~ c i t oeac ~ ~ , ~ ~ s ~inl series ~ e d between two AC systems. The IPC is different from other series corn nsation devices in the way the series elements are connected. For exa of the s ~ n end~ system i ~ i ~s connected to phases Thus, whatever the angle B a IPC t e ~ ~ ~ asome I s , of the ~ o ~ p o n e nare t s always adjusting the value of these components, it is always subjected to a certain voltage. po§s~b~e to force a current in ystem even if the angle is zero. ~ e all nc o ~ p o n e n ~ s are. ener~isedthe amplitude and phase angle of the current are set in one of two buses to which the IPG is connected. This current control thus enables the power carrie to be set, as well as the reactive power ~bsorbedor generated at one ofthe buses,

Flexible AG Transmission Systems (FACTS)

73

r

VCS

.8 Three-phase diagram of the IPC 120

The SSSC, a solid-state voltage source inverter [21,43,44], coupled with a transfornm, i s connected in series with a transmission line. An SSSC (Figure 8.9) injects an almost e ,series with a transmission line. This injected sinusoidal voltage, o f variable m a g n ~ ~ din voltage is almost in quadrature with the line current, thereby emulating an inductive or capacitive reactance in series with the transmission line. This emulated variable reactance, inserted by the injected voltage source, influences the electric power flow in the transmission line.

dc bus

.9 Static synchronous series compensator

er c o ~ ~ r o ~cu~ently ler in use is the ~ ~ ampe~ ~ [22] to ~~ o ~ n t eS r S was first o b s ~ ~ at e dthe $quare utte ~roject.SSR ins~bilitiesare at. times an side effect of us~ngmec~anical~y controIled series capacitors to a t r ~ n s m i s s ~ ~ The ~ e n e of~ a~ ~ s ~series i ~ ca~acitors g are to lower the line’s i m p ~ ~~ ~~ cc ~r ~e, oa w~ ~~ r sists of baclc-to-~a~k thyristors connected in series with a ss the series capacitor (Figure 8.10). The operat~onof the damper is based on two principles. ne is to fire the switch 8.33 ms after each zero crossover of the capacitor’s vol e, or half a cycle (or 180 degrees) at 60 13%.But if the voltage wave contains other frequencies9some half-cycles will be longer than 8.33 ms. In this case, the valve firing at 8.33 ms causes some current to flow during the exten~edpart of the half-cycle and damps the oscillations. The second principle is to fire the switch somewhat earlier tlxan 8.33 ms or less than 180 degrees following the voltage zero

Power System Restructuring and Deregulation

274

crossover. Earlier firing causes the impedance of the combined circuit to be more negative than that with the capacitor alone, thus de-tuning the circuit. F u r t h e ~ o r eb~modulation ~ of the firing angle, the impedance can have a powerful damping effect at any unwanted frequency below the main frequency. Similar effects can be achieved with HVDC controls. A l ~ e ~ a t i v eactive l ~ , filters can also be used. Transm

-set time resonance damper

S controller for using a TCSC to damp out SSR-related problems was presented in [1 87. The new method controls the amount of voltage boost of the TCSC that makes it exhibit a virtually inductive impedance in the frequencies from 15 to 45 Hz where SSR problems may exist. Basically, the TCSC firing angle is modulated to provide damping at SSR frequencies.

he team at the Kayenta ASC [24] showed similar results that the TCSC exhibits an inductive impedance at sub-synchronous frequencies~and the danger from SSR problems was alleviated. However, the main SSR danger resulted from the unco~troiled portion of the series capacitance in the transmission tine.

PE switches (either thyristors or GTO thyistors) can be used to interrupt AC currents. The thyristor depends on current interruption at the natural zero crossover point of the fault current, whereas the GTO thyristor may intempt at a specified current setting (which is below its interruption capability). Such static switches have been applied mainly to distribution systems where the switch ratings are lower [25]. The static breaker can have two parts in it, one a static switch and another a current limiter (Figure 8.1 1). When a fault is experienced, the c~rrent-~imiting switch is firstly triggered to take over the fault current, and the main static switch is opened. This forces the fault cment into the current-limiting path owing to the series inductive element. The non-linear arrestor across the static switch is used to contain the overvoltages [26]. arrestor

F i ~ ~ 8.1 r e1 Solid-state breaker and current limiter

Flexible AC Transmission Systems (FACTS)

275

It is possible to consider the switching capability of thyristors to use as c u ~ e n limiters t in the application of TSCS in the future [27]. The increasing interest in FAC in paI~icularseems to indicate that fault-current-limiting functions can be economically added onto TSCS units. F u r t h e ~ o r e ,these additional features lend themselves to be retrofitted to existing facilities.

8.4.4

Combined Series/Shunk Controllers

rh the transmitted real power and, in~e~ende~itiy, the reactive power flows at the sending and receiving ends of the transmission line. The UPFC consists of two GTO-based converters connected together by a DC link having a storage capacitor. This arrangement functions as an ideal AC to AC power converter in which real power can flow in either direction. Each converter can either generate or absorb reactive power at its own AC terminal. Converter 2 of the UPFC (Figure 8.22a) injects an AC voltage Vm of variable magnitude and angle in series with the line voltage thereby allowing the control of the phase angle between the resultant voltage and the line current. This injected voltage can be considered as a synchronous AC voltage source. The line current flows through this voltage source exchanging real and reactive power between it and the AC system. The real power exchanged is inverted into DC power and is stored in the DC link. The reactive power exchanged is generated internally by the converter. Converter 1 supplies or absorbs the real power required by converter 2 through the link. Inverter I can also generate or absorb reactive power as a shunt device from the line. Converter 1 can be operated independently of converter 2.

P

Parallel lransformer

V’

I

V’

I h

V’

.I2 Unified power flow controller (UPFC)

The operation of the UPFC can fulfil the multiple functions of reactive shunt com~ensation,series compensation and phase shifting by ~nject~ng a voltage Yw with appropriate ampiitude and phase angle (Figure 8.12b). Comparisons between the UPFC and TCSC, and between the UPFC and TCFAR, are made in (281. Results from transient network ana~yser(TNA) simu~ationsand computer studies are also shown. An application of this technique is presently underway at WAPA, located at Mead, and is rated for I060 NIVA (series injection) and 475 MVA (shunt compensation) capabiIi~,

Power System ~ ~ s t ~and c Derenulation ~ u ~ n ~

276

concept (301 for the compensat~on power flow mana~ementof multi-line transmission systems. In its general form, the IPFC employs a number of converters with a common DC li each to ~rovideseries ~nsationfor a selected line of the ~ransmiss~on system. cause of the c o ~ m DC o~ link, any inverter within the IPFC is able to transfer real power to my other and thereby faciiita~ereal power transfer among the lines of the transmission system. Since each inve~eris also able to provide reactive compensation, the IPFC is able to cany out an overall real and reactive power compensation of the total ~ a n s m i s s ~system. ~n This ~ a p a b i makes l ~ ~ it possible to equalise both real and reactive power fl lines, ~ a n s f e power r from overloaded to under loaded lines, c o m ~ ~ n s a t e rops and the corresponding reactive line power, and to increase ng system against dynamic disturbances. In its simplest form, the IPFC

line I

.13 I ~ ~ ~power r ~ flow i ~controller e (IPFC)

fS

A s c h e n i ~ t i~~~ a g r aof m a phase shifter 1311 is sho

a ~ c o r n p l by ~ ~adding h ~ ~ or subtracting a variable vol

fkom a o ~ e isn o~tained ~ have voltages p r ~ p ~ ~ i oton a l be included or e x ~ ~ u d e ~ ' and 9 - along with the plus or e range of -13 to +13, thus gi var~ableh i g h ~ ~ control p e ~ ~of the p e ~ e ~ d i ~ uvoltage l a r co~ponen~.

Flexible AC Transmission Systems (FACTS)

77 VB 4 - -

V'

V'

V - Input voltage i - Linecurrent

The principles of a pliase-shifting transformer (Figure 8.14a) with a thyristor tapchanger are discussed in [31]. Similar to a conventional phase~shifterwith a mechai~~cal switch, a c o n ~ ~ n u o ~ svariable, ly quadra~revoltage is injected in series with the transmiss~online v o ~ ~ (F'I ge 1. It uses three ~ i ~ f e r etnrta n s f o ~ e rwind~ng§(in 3:9), with switch arrangements that can by-pass a winding or reverse its oduce a total of 27 steps using only 12 thyristors (of 3 different volta , There is no thyrisgor"con~o~1ed phase shifter in service ifter does not have the ability either to gener~teor ~bsosbreactive wer it absorbs or supplies must be suppl~edor absorbed c a n s ~ o ~must ~ e rbe locatea close to a ~eneration to reactive power transfer. ~innesotaBower has deve~opeda novel and P single core/single tank &bangeconom~cversion o bang' %ype TCPAR which ical and thyristor switches. An advance^ phase sh ng voltage source inverters (VSIs) using 6 shown in Figure 8.12 in an earlier section. The converter 2 is used to inject v o ~ V,~, in~ e series with the line. The phase relation~h~p of this voltage V, to the line vol a r b i ~ aas ~ shown , in the phasor d i a g r a ~Thus . the injected voltage can be used fo ulat~onor both. ~ u r t h e ~ o rthe e , VSI can itself generate or a ~ s o r ball ompensating voltage injection. On the 0th must be provided by the AC sourc available). ~ w i t c h ~converter n~ 1 supp~~es to or abs dc link capacitor the real power involved in the overall compensation. Since CO handles only real ower, and as its AC side is in shunt with the transmis largely i~nmunefrom the effects of surge currents during any line faults. C o n ~ e r ~ e2,r however, has to handle its total injection FA as well as any surge currents during faults. Consequen~iy,the rating of converter I is smaller than converter 2. The phase shi~terof this type is econo~icalto a total angle variation of 120 degrees. Above this value, the rating of the injection converter becomes larger than the power transmitted through the line. In such a case, it might be economical to consider the approach of the HVDG back-to-

Power System Restructuring and Deregulation back c o n f i ~ ~ t i considered on in Section 8.4.6. The advanced phase s h i ~ ehas r the ability to control all three parameters affecting power ~ansmission:phase angle, voltage and edance. For this reason, it has also been called u n i ~ e dpower flow contro~ler~ ~ P F C ) Il28l.

Strictly speaking, HVDC transmission does not fit in with the definition provided for ACTS controllers. However, HVDC systems have been a dominant player for such a long time in the usage of PE controllers for transmission that their role in promoting high PE c o n ~ o ~ l ecannot rs be overlooked. With the latest developments in PE t e c ~ o l H oV ~ ~C sys~emswill play an even greater role embedded in AC systems. Trad~tiona~ly, HV ~ r ~ s m i s s i oisnused only for special situations and applications: ~ong-distancebulk power ~an§missionwhere it was cheaper than the AC a l ~ e ~ a t i v e ~ back~~o-back asynchronous interconnections, and in~~rconnect~ons using a submarine (or underground) cabie,

ace ~ a f i s ~ ~ ~ s i ~ ~ iona transmission, power is electronically controlled, and hence an ~~~C line can be used to its fill thermal c e converters are adequate~yrated. line can help a p ~ a l ~AC e l line to F u ~ h ~ ~ oowing r e , to its high-speed control, maintain stability (as long as the E'NDC con not sustain ~ ~ u ~ failures). a ~ o C ~ n s m i s s i o nis used only for ~ o w e v e r owing , to its expensive impleme special situations and appiications. An alternative a ~ ~ g e m with e n ~a con~olledseries capacitor in an owing transmission line may provide similar advantages at a lower cost. ~ o w e v ein ~ ,i n t e ~ a AC-DC ~d systems, it is now possible to have a DC link in parallel an ac link, Ln such systems, and there are a n ~ ~ b ofe rsuch instances (i.e. Intertie, C h ~ d r ~ p u r - P ~ dtie, ~ hetc.), e h e DC link can be used to increase the power ~ a n s ~ i over ~ e the d AC system and provide additional d ~ p i n when g requ~redfor stability ith the availability of GTO/IGBT converters, it is feasible to conside inverters feeding into very weak and even dead AC systems [34], which have no s ~ c h r o n o machines ~s at all. Some of the problems previous~yassoc~atedw terminal HVDC systems using conventional thyrigtors may now also be addressed with parallel taps using f o r c e ~ c o ~ u t a t econverters. d This means th consider multi~term~al C systems more sympathetica~~y. G systems can materialise, however, one additional device ects for this are excellent. opmen~will be the HVDC breaker; the The practical d ~ f ~ c ouf l impleme~tin~ ~ ~ T ~ - ~ a conve~ers sed for high a p p l ~ c ~ ~ has ~ o nbeen s the problem of operating GTQ devices in series. Some tec een suggested to build up the high voltage required for DC t r ~ s m ~ s s i obyn using ~ ~ l t ~ - ~ n v e rint eseries rs [353, or the use o f mul~~-level converters; in either case, capacitors are used to equalise the voltages across the ~ u ~ t i - c o n ~ e ~The e r s economic . vi~bi$i~y of such t~chniquesfor h i g h " v o ~ ~ app~ications ge is far from clear at present.

n

Flexible AC Transmission Systems (FACTS)

79

ac~-~o~ ~ ~ v ~ ~ e ~ s Up rill now both converters have been line-cominu~tedand therefore havs had control only over the direction of active power flow. With the use of self-commutated GTO converters (Figure 8.15), reactive and active power flow can now be controlted in any one of four ~uadrants,since there is no restriction &om the commutation voltage of the valves. Additionally, use of PWM techniques will assist in the minimisation of harmonics generated by the converters and lowering the overall cost of the terminal equipment. We can expect further applications o f BB ties at Iower cost and improved performance.

I

AC systm1 I

J

Active and reactive power can flow in either direction

AC system 2

.I5 Force-commutafed BB link

In this respect, two recent deve~opi~ents that will have significant reper~uss~ons for future ties are: capacitor commu~atedconverters (CCC) [36], and controlled series capacitor converters (CSCC) 1373. Both these t e c ~ i q u e srely on utilising capacitors in series with the converter wit effect that the reactive power demande~by the converter is effectively compensat the series capacitors. This i s a ~ u n d ~ e n tdeparture al from the previous HVDC converter practice o f empIoying shunt capacitors for reactive power compensation. The beneficial impacts o f the series capacitor are as follows:

The capacitor voltage assists in the commutation voltage for the converter which allows operation with a very weak AC system. Since the reactive power flow through the converter t r a n s f o ~ e ris reduc~d,the d~mension$of the converter t r a n s f o ~ ecan r be reduced. The valve short circuit current is reduced to about 50% when ~omparedwith a c o n v ~ n ~ o nconverter. al Since the AC filter is reduced in size, the load rejection overvolfages are much smaller. Coupled with these trends, ~ a n u f a c ~ r eare r s now offering more efficient, cont~nuously tuned AC filters, active AC and DC filters, compact and modular outdoor valves and fully digital controls. These new concepts are going to reduce the cost of converters and improve reliability. A new g e n ~ r a t ~of~DC n cables is available based on polymeric i ~ s u ~ a tmater~a~ i n ~ instead of the classic paper-oil insulation. The mechanical strength, flexibility and low weight of the cables make them suitable for severe installation conditions. The cables use copper conductors for submarine usage and aluminium conductors for land usage. Land cables

be inst~lledunder~oundby plough~ngtec iques or go overhea elopment of IGBT valves c the use of newly designed DC c to new app~ications,Usin es with switching f~equenc~es e new IGBT-based, VSCs are ~e1f"~omrnutated and can c o n ~ o lactive and er flow. This reduces the size of co~ponentsrequired a ~ p r ~ c i ers are c o n s ~ c ~ eind a modular concept and are e n c l o s ~i ~ ng range can vary from 7- 60 over distances of 0- 1 lications scenarios are envis ith this concept:

ulk power ~ ~ i s ~ i s s ~ o n . eacctive li er controller, coupled wiek an active dilteri l-scale genera~ionfrom w i ~ d , ding of new r i ~ ~ ~ ~ omay f-w not~be y ava~labIe. q u a i i ~control by iso~a~ing dis~rbingIoads such as smelters. of appl~cat~ons have a ~ r e a ~been y ~ ~ p with Q this ~ ~c od~ c e p(T~ future pr~spectsare excellent.

2 Gotland

50

i80

65

active and reacti

Async~~~ou§ ~terco~~6~ion

Flexible AC Transmission Systems (FACTS)

8.4.7

Other Controllers

These can comprise (a) T~zyr~stor -co~~ro

ected in series with a part o f a In this application a arrestor to lower the voltage limiting level dyn~mically. (h) ~ ~ i y r i s t ~ ~ - ~ ~ ~ t r o ~ i e

This could be a regular t r a n s f o ~ e rwith a thy~s~or-controlled t a p - c h a ~ ~ or e r with a thyristor-controlled AC-AG voltage converter of variable AC voltage, in series with the line. Such a reiatively low cost contro~lercan be used for controlling the flow o f reactive power between two AC systems.

s 8.5.1

svc

Northern States Power Co. ( ~ S o ~f ~ )i ~ n e s oUSA, t a ~ has installed an SVC in i power tr~smissionnetwork, a part of the ~anitoba-~innesota Trans~issi Project, the purpose of which is to increase the power i ~ ~ e r c h a n Wi~nipegand the Twin Cities on existing transmis~~on lines. This solu~~on instead of build in^ a new line as it was found to be supeRor with respect to in utilisation as well as minimised e ~ v ~ ~ o n m eimpact. n t a ~ The main eneration and ~ a n s ~ i s s i osystem’s n d y n a m ~response ~ to ~ e ~ o r k also provides improv~mentduring steady-s~ateconditio adequate reactive power support. Wi the SVC in operation, the po capabiIi~o f the § y s t e ~has increase y some 200 MW. W i ~ o th~ t ~ansrn~ssion ~ a ~ a ofc the i ~ ~S~ n e ~ o r kwould be §evere~y1 excessive voltage ~ u c ~ a t ~ following ons certain fault situations system, or to severe overvoi~a~es at loss o f feeding power from ~a~~ito~a. The system has a d y n ~ ~range i c of 450 MVAr inductive to 1000 500 kV, r n a ~ i ~it gone of the largest of its kind in the world. It consi chan~callyswitched capacitor banks required to control the ovewoltage e n o ~ h end e ~of the 500 kV line, The SVC consists of f x o t h ~ r i ~ t o r s w ~ t ~ h ereactors d ( T ~ ~and s ) three ~yris~or-switch~d ca~acitors(TSCs). ratings are utilised only during severe d i s ~ r b a n cin~the ~ 500 kV networ the SVC has been d~signedto withs~ndovervoltages up to 150 % of rated v o ~ ~ a gfor e short periods (< 200 ms).

Power System ~

~

s and Deregulation ~ ~ ~

essee Valley Authority teamed up with ~ e s t i n ~ ~ otou sinstall e a 100 at TVA’s Sullivan 500 kV s u b s ~ t ~ o~n l ~ ,inl ~~o ]~ § City. o n This trated in 1995. The selection of this site was made to:

Test the full range of reactive power of the STATCON. Aid in damping the oscil~ationsin the TVA system fed in from the n e i g h b o ~ n gAEP bus voltage during the daily load bu~ld-upso that the 500/161 kV m e r bank can be used less often. bus at Sullivan during off-peak periods. are c o ~ s i d e r ~STATCON n~ a p p ~ ~ c aon ~ ithe o ~C Q ~ O ~ W ~ d Electric Co. Some cost evaluations have been reported at 3 ~onference[ 121.

8.5.3

TCS

i ~ a n u f a c ~ r e rare s resently being tested in North In 6991, AEP of Columbus, Ohio, with the ~ ~ u f a c ~ ~ e ~ of a single-~hase series capacitor b ~ atk ~ r o t o switch ~ e er sub§~tionin W. Virginia. Fo~~owing s ~ c c e s tests, s ~ ~ a 788 A, 42 ohm series three-phase capacitor bank was installed. Each p consists of two p l a t f o ~ s one , with both a 10% (7 the other with the remain in^ 30% (21 ohm) s e ~ ~ e Power A u t h o r i ~( e first t~ee-phaseTCSC 230 kV, 33 in Arizona [40]. For the requirement of inc ~ a n s i ~ i s s line ~ o nb e ~ e e nShiprock Subs~a~ion and Siemens~okiajointly agreed t yenta substation. In addition to the benefit of adjastable i ~ p e d a n c ethe ~ ed reactor can provide high-speed pro~ectionof the 15 ohm capacitor section. managed the install ith n ~ a n u f a ~ ~GB ~ r esuccessfully r ptember 1993 on a 500 kV line at Slatt subs~tionof the TCSC consists of six series capacitor modules. Each modu~ehas ohms at 60 Hz, in parallel with a ~~yr~sto~*control~ed inductor of 0. of the ~ o d ~isiachieved e by firing angle control.

~

Flexible AC Transmission Systems (FACTS)

Z$~

UPFC application was commissioned in mid 1998 at the Inez station of Kentucky for voltage support and power flow control. regulates the substation 138 kV bus voltage by controlling six capacitor banks Ars to reduce daily and seasonal voltage fluc~ationsto within acceptab~elimits. The controllable rea~tivepower range of the shunt converter is from -160 to t-160 ~ V A r s to Compensate for dynamic system disturbances. The PFC is maintained at a level of 300 N W on the line between ig Sandy and Inez to minimise system losses. Under severe contingency conditions, the UPFC controlle~line i s capable of ~ a n s f e 950 ~ n MVA. ~ In order to increase the system reliability and provide flexibility changes, the UPFC installa~ion allows the operation of the shunt ~ d e ~ e n d e nS tT A T ~ ~and M the series converter as an independent ible to couple both converters together either in shunt or series over a double control

Each GTO converter is rated at I- 160MVA. The converter output is a t~ee-phas set of nearly sinu$oidal (48-pulse) quality. Each converter feeds an in transformer that is coupled to the transmission line via a conventional thr te is 50 % of the main Iran transformer. The rating of the i n t e ~ e ~ atransformer The converters are c o n s ~ c t e dfrom three-level poles, each c o ~ p o s e dof four valves. This a~angementassists in waveform construction to facilitate harmonic elimination. Each converter employs 48 valves in 12 three-level poles with a nominal dc voltage o f + 12kV and -1 2 kV with respect to the mid point. The mid point voltage is maintained by ~ e a ofn ~ a split capacitor and diode arrangement.

lation of the power i n d u s ~[42], FACTS controllers will be r $ ~ u ~ r eby d power systems to ana age power flow to utilise transmission lines nearer to their t limits. The ability to transmit at higher transfer limits will necessitate to balance reliability and economy of operation of the power sys adoption of FACTS controllers, the following concerns o f the power industry ~ ~ toebed addressed: Transient ove~oltages. System restoration. Generator torsion behaviour. Power quality. Economic ~onsidera~ions and cost benefits.

ower System ~ e s t ~ c and ~ n g

e ~ o n c e ~ study s, tools are require^ to test. the FACTS con~o~lers as c o n c e p ~prototypes ~~ or before the er c o ~ ~ e service), r ~ ~ The a ~ ) and the red-time power on ~ l e c ~ o m a ~ etransient tic pro ly available. Noweve these c o n ~ o l l e r sare still Owing to the capital costs involved, FACTS d e s ~ ~ ewill r s seek to add featur FACT^ c o n ~ o ~more l e ~viable, such as the feature of fault c ~ ~ l iem ~n t i~nwith ~ FOP the a p ~ ~ ~ c a t of i o nS~T A T ~ O N ~ value , may be % ~ d if~ fde a ~ r e ssuch ondition~ng(I.e. h a ~ i ~ n cancellation) ic can also be provided along with ower.

The author pays tribute to the many pioneers whose vision o f the FA led to the rapid evolution of the power industry. Although it i o f them i n d i v i d ~ a l ~ y ~c o n ~ ~ b ~ t iof o nDrs s N, ~ i n ~ o ~ ~ i The author also than s wife Vinay for her considerabIe this ~ a ~ ~ s c r i ~ ~ .

i, ‘FACTS - flexible ac ~ansmissionsystems’, IEE ~ n f e r n a ~ i ~Conference nal on wer Transmission, 1991, pp.1-7. mgorani and L. Gyugyi, ~ n ~ e ~ s t aFACTS n ~ i n- ~Concepts and Techno AG T r a n s m ~ s Systems, ~o~ IEEE Press, 2000 L.Gyugyi, ‘Solid-state control of electric power in AC ~ m i s s i systems’, o ~ Interiiational ~ ~ m p ~ons Electric i u ~ Energy Converters in Power Systems, Invited Paper No. T-IP. 4, Capsi, Italy, 1989, PACTS Overview, XEEE PES Working Group Report and CIGRE ~ n f ~ r n a t iConference on~ on Large High voltage Electric Systems, Chairmen: E.Lassen and T. Weaver, April 1995 T V.K. Sood, Position Paper on FACTS T e c l ~ o l o ~Canadian y, Electrical A§§ociation~C o n ~ c t CEA ST-460, March 1995. A. Erinmez, Ed, ‘Static Var Compensators’, Working Croup 38-01, Task Force N o 2 on SVC, undamentals of thyristor controlled static var compensators in electric power system app~i~ations’, EEE Special pub~ication~ 7 ~ 0 1 $ 7 -p ~ e- ~~n in ~~ 1987. e~ d, A. ~ a ~ a ‘ A d ~, l y s of i ~power system s ~ a ~e ~i ~l n~c ~e m by e n static ~ var compen§a~o~s’~ IEEE Transactions on Power Systems, Vol.PwRs-1, No.4, November 1986. . Machur, S u p p l e ~ ~ton ta b ~ b l i o ~ a pforh ~static VAR co~pensa~ors (SVC) and r e ~ a ~ flexible ed ac transmission system (FACTS~devices [ 1988-19941. C.Schauder et al., ‘ ~ e ~ e l o p m of ~a~100 t MVAr static condenser for voltage control of ~ a n s ~ ~ s s~~i toen~ s IEEE ’, ~~ansactjons on Power Delivery, Vol. 10, No.3, July 1995, pp. 1486- 1496. . Mehta, et al., ‘Static condenser for flexible ac transmission aystems’3 EPRi FACTS ~ ~ n ~ e 2:~TnR 101784, c e ~ e e t i n gin May 1992, Procee

Flexible AC Transmission Systems (FACTS)

5

1121 A. Ekstrom et al., ‘Studies of the performance of an advanced static Var compensator, STATCON, as compared with the conventional SVC - EPFU Project ~ - 3 0 2 3 EPRI ~ ’ ~ FACTS3 Confere~ce, Baltimore, ~ a ~ l a nOctober d. 1994. ittlestadt, ‘Four methods of power system damping‘, B E E T r a n s ~ t on ~ o Power ~ Apparatus and Systems, Vol.PAS-89,NOS, May 1968. [141 P. woo^, ‘Study of impro~edIoad-tap-changing for ~ r ~ ~ f o ~ e r s damping improvement [l4] C. Wu and U. Lee, ‘Applica~ionof simultaneous active and reactive power modula~ionof super~condiictingmagnetic energy storage unit to damp turbine-gene~torsu~”sync~onous osci~lat~ons’, IEEE Transactions on Emrgy Conversion, Vo1.8, No. 1, March 1993, ava and G. Dishaw, ‘A~plicationof an energy source power system stabilizer on the battery energy storage system at Chino substation’, IEEE Trunsact~o~s on Power Vo1.13, No.1, February 1998 “A [ 181 L. Angquist, 6. Ingesbrorn, and H. Othman, ‘Synchronous voltage reversal new control method for thyristor controlled series capacitors’, E P FA ~ ore, aryland. October 1994. nl by [ 191 M. Gavrilovic, 6. Robcrge, P. Pelletier, J-C. Soumagne, ‘Reactive and acti means of variable reactances’, 11th Pan-Amerjean Congress (COP1 treal, Nove~ber1987, P, Pelletier, F. Beauregard and 6. MO&, ‘hterphase power controller r m ~ a g i n power g flow w~thinac networks’, IEEE Transuc~ionson Power Vo1.9,N0.2, Apfil 1994, ~ ~ . 8 3 3 - ~ 4 1 . [a I] K.R. Sen, ‘ S T A T C -~Static ~ synchronous compensator: Theory, modeling aid applica~j~ns’, IEEE PES Winter Meeting, 1999, pp.1177-1183. E221 N.G. ~ ~ n g o r‘A~ new i , scheme of sub-synchronous ~ s o n a n c edamping of ~orsiona~ oscillations and transient torque - Parts I and U’, IEEE Transmtions on PO $yst@ms,Vo~.PAS-~ 00, N0.4, April 1981, and IEEE PE$ Summer ~ e e t ~ n g

/23] J.W. Ballance and S. Goldberg, ‘~ubsynchro~ousresonance in series c ~ m p ~ i s ~ ~ e d transmission lines’, IEEE Trunsu~tionson Power ~ p p a r a and t ~ Systems, Vol. rolled series compensstion to avoid

[25] T. Ueda et al., ‘Solid state current limiter for power d i s ~ b u t ~ system’, on l E E ~ransactions ~ an Power Delivery, Vol.$, No.4, 1993, pp.1994-1801 . Saarkwzi, E.J. Stacey, J.J. Bank and N. t261 dis~ribu~ion current limiter and circuit breaker: Application requiremen~s lE## Transactionson Power &divery, ”IIo1.8, No.3, July 1993, pp.1155-1 arady, ‘Co~ceptof a ~ m b i ~ eshort d circuit limiter and series c o ~ p e ~ s a t ofEEE ~’, ~ 7 3 ~ r a n s a ~ ~ i on a nPawer s Delivery, Vol.6, No.3, July 1991, pp.103~-1037. [28] L. Gyugyi et al., ‘The unified power flow control controller for i n ~ e p e n d eP~and ~ control in. transm~ssionsystems’, FACTS3, Baltimore, M ~ l a n d~, c t o b e 1994. r

Power System Restructufing and Deregulation

286

[30]

[32]

[33]

1361

[37]

[38] E391 [40]

[41]

.K. Sen and E. Stacey, ‘UPFC - Unified power flow controller: Theory, odel ling and applications’, IEEE Transactions on Power Delivery, Vol. 13, No.4, October 1998, pp.14531460. L. Gyugyi, K. Sen and C. Schauder, ‘Thc interline power flow controller concept: A new approach to power flow management in transmission systems’, IEEE Transactions on Power Delivery, Vol.14, No.3, July 1999, pp.1115-1123 asati, ‘A thyristor controlled static phase shifter for ac power transmission’, E E I i Transactions on Power Apparatus and Systems, Vol.PAS-~OO, No.5, May 1981, pp.2~50-265~. R. Baker, 6. Guth, W. Egli and P. Elgin, ‘Control algorithm far a static phase shifting transformer to enhance transient and dynamic stability o f large power systems’, IEEE Transactions on Power Apparatus and System, Vol.PAS-101, No.9, S e p ~ e ~ b1982. er J. Kappenman et al., ‘Thyristor controlled phase angle regulator applications and concepts for the Minneso~-Ontario Interconnections’, EPH FACTS3 Conference, Baltimore, M a r y ~ a n ~ Oct 1994. ood, Position paper for Canadian Electrical Association on Artificially ~ o ~ u t a ~ e d Inverters, March 1989, Contract No. ST-174B. ng, J. Kuang, X. Wang and B. Ooi, force-commutated NNDC and SVC based on phase-shifted multi-converter modules’, IEEE Transactions on Poww Delivery,Vo1.8, N0.2, April 1993, pp.712-718. T. Jonsson and P. Bjorklund, ‘Capacitor Commutated Converters for HVDC’, Paper SPT PE 02-03-0366, IEEE/KTH Stockholm Power Tech Conference,Stockholm, Sweden, June 1995. K. Sadek, M. Pereira, D. Brandt. A. Gole and A. Daneshpooy, ‘Capacitor c o ~ u ~ a t econd verter circuit c o n ~ ~ r a t i o nfor s DC transmission’, IEEE Transactions on Power DeZivery, Vo1.13, No.4, October 1998, pp.1259-1264. J. Vithaya~hil,P. Bjorklund and W. Mittlestadt, ‘DC systems with t ~ n s f o ~ e r l econverters’, ss IEEE Transactions on Power Delivery, Vol.10, No.3, July 1995, p~.1499-15~4. A. Keri, A. Mehrbahn and P. Halvarsson, ‘AEP expenence with the 788 Mvas series capacn d . 1994. itors and the controlled thyristor switch’, EPRI FACTS3, ~ a I t i ~ o r~e ,a ~ ~ aOctober N.Christl, ct al., ‘Advanced series compensation with variable impedance’, EPRl Conferen~e I on FACTS, Cincinnati, Ohio, November 1990. Proc. March 1992, EPN TR-100504, Project 3022, J.Urbank et al., ‘Thyristor controlled series compensation prototype installation at the Slatt 500 substation’, IEEE Transactions on Power Deliwy, Vo1.8, No.3, July 1993, pp. 1460-1469. enderson, ‘Operating issues for FACTS devices An operations p l ~ i n pg e ~ ~ e c ~ i v e ’ , EPN FACTS3, Baltimore, Maryland. October 1994. Sen, ‘SSSC Static Synchronous Series Compensator: Theory, modeling and applications’,IEEE Transactions on Power Delivery, Vol. 13, No. 1, January 1998. L. Gyugyi, 6. Schauder and K. Sen, ‘Static synchronous series compensator: A solid state approach to the series compensation of transmission lines’, IEEE Transactions on Power Delivery, V01.12, No.1, July 1997,pp.406-417. ~

~

[44]

Kevin Morton London Electricity Group UK

Cliff Walton London Electricity Group UK

Asset management has been one of the most debated topics over the past decade, yet ofien those words are used to label some very different processes. Asset management can range from the ma~ntenanc~ and renewal regime associated with a specific indiv~dualor group of assets to the management of a multi-billion-pound international portfolio of networks of assets spanning a range of industries. This introduction explores the drivers of the development of asset m ~ a g ~ m efrom n ~ a UK electricity distribution ~erspective.The drivers for change have most often arisen from regulatory initiatives or from the ~nancial position of new owners, with asset management evolving to meet each new challenge. U n d e r s ~ d i n gthe drivers gives an insight as to why asset manage men^ means ~ ~ f f e r e n t things to different players depending on where they are in the resmchtring of their business.

In the years i~mediatelybefore privatisation, the electricity indushy 's finances and investments were very much Treasury driven to meet the public sector borrowing requirements. Compet~ngdemands for government investment meant that most e l ~ c ~ i c i ~ companies were required to curtail capital investment and were given annual targets to return cash to the Treasury. At this stage of developm~n~ asset management was normally considered synon with time-based planned maintenance. However, the constraint on the capital expendi~re (Capex) investment meant that as little in the way of reinforcement or renewal was possible ~ were and this brought about a focus of improving asset utilisation. U n s a t i s f a ~ t oassets

2

lation

removed and wherev~rpossib~enot replaced, whilst ~ d e ~ ~ i lp~ant i s ewas ~ recov~re reloca~edto meet load ~ o ~ h . onal areas with local olicies a n ~ o inte r retation of policy rn su~tingin ~ i d e ~ y Is, unit costs and performance of networks. en business o~eratingunits was used to drive lowe utilisation but the availa~ilityof c rmance proved to be a limitation a b ~ the u ~ac~uracyof the statisticsbetween rivals.

~ r i ~ a ~ i s ~~t i o~n a gho ~v e~~s~ e ntot §realise the c o ~ s i d ~ cap ab~~ at the same time free industries from the cons ~e¶uiremen~s. Prices in the UR were initially set vi rec~ivinga ~ o § ~ t i vx,e ~ h e r ~ be yn a b l ~the~ perceiv~d nts to be hnded by the new investors. ut the new sha~eholdersbrought about a strong ~ r o d~r i ~ t e rfor staff num fall, which in turn b ~ o about u ~ ~

Asset ~ a m a ~ ~ r l s eprovider ~ i c e business models beg^ to be a d o ~ ~ e but in a variety o f ~ o ~Often s . initi~~ly with service level a g r e e ~ e n al provider conb-actors, some relatively small asset ~ ~ a g group e ~and~ i nn~ e~~semi a ~ i moved ~ ~ s to adopt ~ o ~ c ao n1~ a c b~ se ~ e the e ~p ies and ou~ourcenon~core b e n e ~ t were s seen to be: a c ~ ~ v ito~ei x~ ts e ~ c~mpanies. a~ g ~ p e r a ~ i n~ gx p e n d i ~(r e rent ~ v e sdeci ~ ~ t ecisions from the doing, ed a d~~ferent skill set f ~ o mthe ~xecut~on. changes in pract~cesan g to be contracted out b the ~ ~ o $ t fied gets done (or

These chan~esmeant that comp

to acquire additional s ision making enabled a i d expertise enabling Iarg

t e c ~ i c astaff. l The asset manager/servi~eprovider model has met with mixed su drivers o f the service provider not neces become ~onfron~ationa~ with those of the asset m 0th sides need experts, one to specify and o

d successive year-on-ye o f unspeci~edpena~tiesan

e achievement of (c~mp~y- set^ per World class studies, bench~arkingand b~sinessprocess re-en~ineenng. Considera~ionof a whole life approach towards vestment. O r g ~ ~ i s a t i omoving ~s towards a three-layer model as they begin to s e ~ a ~ ~ a ~ e r s h i pfrom o~erationa~ m~ag~men~: a Strategy b Asset m ~ a g e m e n t c Service provide^ A more accountable set of relat~onshi~s specifying what needs the doing to the a c c o u n ~ unit. ~~e ~ ~ a t e g asset i c ~ a n a g e m ~ approacl~ nt to u n d e r s ~ d i n gwhere value is c r e ~ ~ e destroyed, s ~ ~ r n ~ins the e d estion - ‘where best to invest the next poun~?’ §cenario ~ a ~ y to s ~ s i n v e s ~ e n tstrategies that are most i r ~ ~ u ~and a tc ox ~~ ~ ~ a l

The second half of the 1390s

Power System ~

e

s

~ andc

~

~

n

Data mining, fault causation analysis and targeting of worst-s customers and most ve to operate n e ~ o r k s . dition monitoring to inforni selective refurbishment or renewal. abi~i~-cen~ design, e d engineer~ngand mainten~ce. ing dormant and problem assets and imprQvingasset y capital project management for smaller and smatle ~ a l u e - b a ~ procL~remen~. ed ovation in technology and processes.

for past i n v e s ~ e n t an some 9 m o n t ~ sahead end of the five year review period. At the same time indications of future income caps and ~ e r f o ~ a ntargets c e were published with 50% of the savings from mergers clawed back. The im~ediatereaction by PESs to the regulator’s initial tho~ghtsdepen~edupon the robustness of their asset management scenario planning and their long-term strategic intent. Some cQntinued much as before but overall the publication of the initial r~viewresults created a drama~~c fall in capital i n v e s ~ e n orders t and in the asset replacement con~acts h limited rewards for excellent p e r f o ~ a n c eand p ~ d ~ capital nt i n v e ~ ~ e nthe t, switched fixed resources onto those targets they saw as ~ a v ~ na ggood ieving without additional investment, whilst ~ o ~ - p e ~on l ~those ~ng that require^ investment and additional resources.

prices cut by 20-35% per-unit from r 3% per-unit price reductions for 4 ency savings of 19%-29%.

PESs from April 2001

and aggregation costs ~ a n s f e ~ to e dsupply. The role of dis~ibutionredefined and almost all customer service costs transferred to al~owedrate of return (good asset m ~ a ~ e mcan e ~deliver t more). A metering ~ ~ ~ p efrom t ~April t ~2000. o ~ Agreeme~ton a business separation comp~ianceplan. P e r ~ o ~ a ntargets c e set by the regu~ator. I n f ~ ~ a t i and o n incentives project to come. ce savings whilst ~mprovingcustomer It is a major cha$~engeto deliver the DPC suggestion from the regulator is that service: with an uncertain incentive mechanis o r ~to mimic a competitive companies will be placed within an incentive ~ ~ e wintended

~

Asset ~ a n a g e ~ ~ n t

91

inarket with Gompanies that do least well in meeting their agreed targets financially r e w ~ d i n gthose com~aniesthat do best by an exchange of penalty payments. The uncertainty posed by Ofgem’s Information and Incentives Project in terms of what will be incentivised, how p e r f o ~ a n c ewill be defined and measured has for many companies effectively extended the moratorium in investment. Companies need to consider how the required scale economies can be effected whilst at the same time d e ~ i v e ~improving i~g p ~ r f o ~ ~Some c e . companies may choose to ~ e f e ~ major new ~ n v e s ~ ~ commitments ent and perhaps org~isationalchanges until there is greater clarity about the rules of the next round of the regulatory game, but this brin own risks of failure to deliver required improve~entssufficiently quickly. The u n c e ~ ~ n ~ high~i~hts the need for a robust frameworks for modelling and valuing the impacts of the various organisational and investment opportunities against a range of scenarios. The scope of asset management has developed with each previous sta r e s t r ~ c ~ r i nofgthe d~stributio~ business and is therefore set to do so again. For companies already recognised by the regulator as being frontier efEcient or as leaders in effective asset management, but still being presented with a very si in r~gula~ed income, a her radical change is essential to achieve the r change in results and still remain at the frontier. ~ o ~ b i n i the n g manage~entof the two power distribution networks v e n ~ r ecompany (2~seven)is LE’S and TXU’s innovative response ~ e r f o ~ challenge. ~ ~ c e Creating an outsourcing a ~ a n g e ~ with e n ~ the tran vehicles and tools, etc., allows the shaxing of expensive ~esou~ces, S U G ~as offices, IT, control, s~ategicana~ys~s and research, applying best practice optimum solutions and delivering a range of services at best value for allowing each company to retain is ownership, distribution licence and to unique com~etitivearke et position should this be appropriate. Such an approach creates the driver for the next evolutionary phase of asset m ~ a g e m e n tand requires the separa~ingout and future comp~titiveassignme~tof the responsibilit~esb e ~ e e asset n owner, asset governor, asset manager and operators.

The owner of major sets of utility assets, whether it be a gov~rnment,~ u ~ t ~ n a t i Q n a ~ co~oration,publicly quoted c o m p or ~ ~m ~ i c i p a lcooperative, will n o ~ a l l yhave a relatively small set of strategic objectives it is seeking to achieve by its ownership, e.g. ansion etc. It will not normally wish to concern itself with the detailed ~ ~ ~ cr ~i a~ ~uor, ~~ e~ c t~ iomc a~an~a g ~ e nof t the assets but mere~yto s a ~ i s ~ at they are in the hands of an effective governor who can reliably deliver its iC objectives with Erontier ef~ciencyand effectiveness. ASS

nc

~ e p ~ a t i nout g the respons~bilities of governance from those of o ~ e r s h management and operations to an organ~$ationdedicated to the creation and release of value t ~ o u g hthe effective m ~ a g e m e nand t exploitation of the assets.

ower System ~ e § ~ ~ c ~and r ~i Ie ir ~~ ~ l a t i o ~

92

The asset ~ o v e r n ~ concept ce provides for even a n o n ~ t e c ~ i corgan~sat~on al to from the ownership of a world class set of distribution assets and services with l~mi~ed d with a minimal stafX s new to the e ~ e c ~ cindustry ~ t y but s ~ m ioppo l~ ities exist in other ~ s the own~rssee the need to i n d u s ~ e ssuch as rail md a i ~ o where a1 ef~ciencyand returns s i g n ~ ~ c a nbut ~ yhave other major an ial o p p o ~ i t i e to s commit their ~ a ~ a g e m etime n t to. ~ ~ ~ ~compliance, l a t supply o ~ business satisfaction, income ~ a ~ i ~ i sand a t value ~ o ~ e~erationre~uirea different skill set from the ~ a n a g e ~ eof n t~ ~ d i v ~~ sds u~or~t ssets of

se, a v a i l ~ b i l i ~capacity , and income ge~erationfrom g and actively managing the portfolio of risks. developing network assets to match new ~ a r k e t sfor

sition with respect to frontier e ~ c i e n rc e~ ~ ~ a t o ~

es arising from the removal of geo ~ompeti~ive cons~aints. allows economies of d maxim~s~ng the returns from

sibly unique arke et p~sition

ent and operations to be ~ o n ~ a cout t e to ~ the mo r ean r ~ c i ~role a l of the asset governor is ~ e ~ e f oas e ~ ~ e ~ tand i v eefficient ing the g o v e ~ owill r require ntier u ~ d e r s t ~ ~of:

vol~mesof servic lity of work done and the value add^^ s and s ~ n d a r neces ~s to evel lop and renew con~act

also discharge a ran of the owner that cann

Power System Reshucturina and D e r ~ ~ ~ a t i o n

2

As has been seen, asset manage men^ is given a wide variety of int industry and even within the electricity supply industry. Even with i n t e ~ r ~ t ~may ~ i ochange n with time particularly as the com~anylearn ~ypically,asset manage men^ has been seen as the core of the d being p~marilyresponsible for the strategy of the network and both are derived through teamwork and cooperation throughout main areas of focus are asset and network p e r f o ~ a n c epolicy ~ and s t a n ~ ~ di sn ,v e s ~ e n t , and operating costs. The focus on the latter is t ~ o u g hwork reduction and a v o i d ~ c ewith the operational groups focusing on productivity issues.

-

Data is the essential ~ ~ ~ e dtoi eeffective n ~ asset ma~agement.The asset man process adds value by converting this data into decisions, which reduce the o v ~ ~ a l ~ ycle cost of the network. n service ~ifecyc~e costs can be ~ r o k edown ~ into ~ r e ~e ~ s t i n areas: ct ~nstallation, ~ one o f the major factors in operations & maintenance and d e c o ~ i s s i o n i n gHowever, of the asset. ~ n s ~ l and ~ a ~ ~ d e ~ ~ i n i nthe g overall lifecycle cost is the actual at the concep~alstage of a ave adi it ion ally been evalu d e c o ~ ~ i s s i o n i ncost g intenmce costs being considered over a fixed project, with operating gers today are faced with d ~ i s i o on ~ s whe p a ~ i c ~ asset. ~ a r Asset

invest, but also have a responsibility for the much wider issue of the exisli For all these existing assets, decisions must be taken which reduce the CO asset in service and extending the period for which the ass& provides sati this i s the essence of asset m The i m ~ o r ~ aquestion nt Asset Managers must ask themselves is:

If the answer is yes then the task is s ~ a i g h t f o ~ a-r we d simply record the age of our assets and replace them at the correct time to prevent them becoming a safety hazard to staff, d i s ~ ~ ~service t i n ~to our customers, or becoming expensive to m a ~ n ~ a ~ , Life just isn’t that simple. Assets ‘age’ at different rates depend in^ on the nature o f the duty imposed on them, the environment which they inhabit, the way they were well as a whole host of other c o n ~ b u t factors. o~ Even if we are able to ‘no ageing process it is still necessary to be able to predict the life span for each we are to avoid replacing plant too early, or allowing our service level to deteri Tools exist to enable us to d e t e ~ i n ethe condition of some of our condition of the asset is far more i m p o ~ nthan t its age. The present conditi of how well the asset has aged over time.

Co~~dition ~ o n i ~has r ~come g c o ~ o n p ~ a in c ea number of asset-inte It has the potential for e ~ m p l i c a t einformation ~ systems to capture pre about ~ a ~ i c u laspects ar of an asset’s p e r f o ~ a n e eand present them in a us to fac~l~tate decision m ~ i n gon maintenance regimes and replacement vides the opportunity for an operator to inspect visu on, a piece of equip men^ and report whether a n y ~ h i nhas ~ change since the last visit. g re~o~in~ Taken to its other e x ~ r e ~ite ,could mean a fully auto~ated~ o n i t ~ r i nand system complete with e ~ ~ y - w ~ ai lnmgs for an indication of wear or %heneed for ma~ntenance. The degree of c ity is a major factor in whether the cost o f c can be ~ e c o v e r eb~ ed m a ~ ~ t e n a costs, n c ~ higher utilisation, ex manage men^ of risk here i s little point in just mo~itor~ng the t r a n s f o ~ e ron a re r basis, via expensi~eanalytical e ~ u i p oil sample will provide a more reliable i i ~ e x p e ~ ~ analysis s~ve wear or ~otentialfailure. If neither of these t e c ~ i ~ u eenable s accurate pred ss fai~ure,or reduc~ionin main~enance,then we must question their u s e ~ ~ n e in management process. tage of condition~~ased monitoring is that it allows the as ree o f con~dencein how the assets are p e r f o ~ i n gand le Eime-based preventative maintenance. In short, it provides the the cost of maintenance and extend the life of the asset. If re

2

Power System ~ e s ~ u c t and ~ng

it or in^ is carried out and records collated, a ‘foo I’ for each item of established and trends monitored. This can be useful for predicting potential f a ~ l ~ ~ s correct~vem a i ~ ~ n ~ c e Lace, which is normal~yless expensive than g a catas~ophicf a i l ~ e . ines for ~ n a c c ~ ~ tperformance ~ble are unavai~able,collati pu~ationenables ‘Qut~iers’to be identi~edand e x a ~ i n e d ~ e g r e eof comp~exityof the rnonito remain %hesame - decide on the criteria for perfomm ind~catorof po&entialv ~ i a t i o nfrom this s ~ and ~ ~ d point in con~~nuousl or of wear is the time it take ring equipmen%and techni~uesare ~ ~ e n availab~e t ~ y to the following sections detail a selection of some of those e

many aspects of a transformer which can be monitQ~ed.One simple but e f f ~ t ~ v e form this is ach~eved mon~tori s the level of moisture in the insul vis~ialiyinspect~ngthe colour of the silica r e f ~ g e ~ b~ ~e ed ~ ~ ~which e r s actively , redu s if the ~ o i s t u r level e should rapidly increase. eing’ process can effectively be slowed ~o~ r. London Electricity has effectively employed level of m o ~ s t ~within re the tram A more d e ~ i ~ ~e di c ~ofr the e condition out dissolved gas analysis (DGA) on a sampl sent and what activity is likely to h ite equip~entto provide a coarse i n d ~ c a ~ iby ~n, carbon ~ n o n o ~ i dmonitor. e This provides a$so~iatedwith overheating, elec set levels. This provides the opp lant arising and actual and moisture content tests can also of the c o n d ~ t i oof~ %hep l ~ We t ~ can

Asset ~ ~ a g e m e n t

7

not simply the num duty of the contacts

continuously loaded profile in Outer ~ o ~ ~ o n . trials are now in progress to s i ~ u ~ amany t e yews' w o ~ ho€ onths. ~ n s p ~ of ~ othe n oil and contacts at various i n t ~ ~ ~ enable us to c o ~ our i initial ~ ~ asse~ionand d e ~ ~ within these trials s e t e ~ ~ n i ma~ntenance ng intervals for diffe~entlyloaded ~ a n § f o ~ ~ r ~ "

9.13.3 Apart from the routine oil condition tests mentioned ~ r e v i o u s l ~ circuit , breaker t~mers e arnoun~of wear on the o ~ e r a ~ imec~anism n to be ~ o n ~ t o r e~d .~ n d ~o n~ e c ~ i se of a simple and inex~ensivee ctronic timer when carrying out op

ry ins~ection(~ressureVessel

9.13. city makes e x t ~ s i v euse of infrared detectors and t h e ~ o v i s ~ oc n ts caused by loose c o ~ e c t i o n sor worn c lings on exposed b ~ s ~ aor rs

Ith check nion~torsare used to check for exce§sive vi~rationmd h ~ n c e mon~tor~ e c h a n ~ c aear l on th ~ a n s f o ~cooling er syste~ s. The i ~ i ~ ~is~not t oar vibration truest sense since it is basi ic evice, w ~ i assi~iilates ~ h noise with wear. in~er~ereiice detectors are c ta monitor d ~ s c h a activi r~~ ~ o ~ a transient b l ~ earth ~ o l (TEV) ~ ~ e of the d i s c ~ a ra~c t~i v i ~ ,or c o sirnil~ ~ ~ ~ ion. ~ o ~ ~ n ~u mooun i~t o r iis~ ~a1so availabl on it or 'leaky' ~ l a ~This t . can

compile a f o o ~ r i n tfor a particular substation. Where a high level of le monito~ngequipment with pre-set a1 vels and potential failure.

Understand~~g Long-term Asset Costs If we are to ~nders~and the long-term costs of employing assets, then we must have a good ~ d ~ r s ~ dofi nhow g they perform in service and what t e c ~ i ~ u can e s be e ~ p ~ o y etod xtend asset life or reduce the level of main~enancerequired, to them in service. As ind~cated in the previous section, purely ti~e-basedmainten ent being ma~ntainedtoo early or too late. In both case u n n e c e s ~ ea ~x ~ ~ n d i ~ r e , therefore need to develop a data model of the asset, which can a c c ~ ~ t ereflect l y Its n, m a i n ~ e n ~requirements ce and life span. In many cases this can be wealth of historical data, coupled with on-line indication of performance. ~ n f o ~ a t e l y this is not a~waysthe case and we are left with the problem of developing a model based on tions and very little feedback from the asset itself. toring of assets is not of the pop~ation,accessi sampling techniques to e n based on tests performed on ~ I e c t ~iscthe i ~~ d e ~ g r co ~ d

~ ~ d e ~ g Cables r ~ ~ n d V and 17500 km of LV under The s ~ ~ o n network d a ~ consists of 8500 cable. The e n v i r o ~ e nin t which it exists makes it difficult to mon~tor,~ l n e r a b to ~ e~hirdularly with the high level of excavation a c ~ vwi i~t ~ i nLon~on,and network account for nearly two-th~dsof the inte s ~ a ~ e g i c ai ~ m~py o ~ to t the com~any. d cables, investment must be targete r e a c ~ n the g end of their useful life. It is essential, therefore, that we are able to derive a mea§ure ~ n d i v ~circuits ~ ~ a l and even localised sections on a circuit has ~adi~ionally been accepted e f ~ ~ c t ~The v ~ assumption ~y. has been that the per failures against asset life, or at least the middle this approach is owing when the particular asset has re f a i ~ ~ r ewithout s, the volume of these failures seriously need to consider the generic model of the ba focus on the bottom ~ o ~ ofothen curve. Figure 9.2 d e m Q ~ s ~ taeseries s of curves with v a ~ i n grates of fai~ure their life span. The ideal situation would enable us to ntifj small increases dicted by the latter as the b e g i ~ ~ nofga steep increase in the failure

~ ~ d i c a tthat e s the slope on many of the small va~ationsis s

10

;13

3

40

50

~

~

txl How long?

ath tub curve p r e d ~ ~failure t ~ v ~mode

entify some other means of in^ §tanda~dsof network

9.13.7

edicting ~ a ~ lif~ we r e are to a v o i ~

bles lem is to analyse the fau~ts cable or joint being an to lead to similar failur s provides the crucial key d where they are most likely to occur. o ~ r ~ modes m a of~ failure:

i

3

t ~ g e t e dcon~itionmonitoring t e c ~ i q u e scan i tify i n ~ i ~ i d ucirc~its al with of failure, Some of the condition m o n i t o ~ techniques ~g includ~: Tan 8 and delta tan 6 Zero sequence impedance P a ~ i adl i s c ~ ~ mapping ge time domain reflectrometry. E: Cable failure from overloading itself is rare but most f a i i ~ e cs to thermal runaway in the insulation ~ o i in d~ the insu~a~on. Manufac~~n c o m ~ a r a ~ i ~rare e l y in the UK. Condit

Partial discharge mapping ~ ~ ~ l e cLoss t i cangle (Tan 6) ielectric Loss angle with voltage ( al imaging of t e ~ ~ a t i o n s onic aging and ~ i s c h a r detection ~e uted ~ e m ~ e r asensing ~ r e using fiber optics series of tests have pressure test as an indi re ~ x ~ ~ n stests i v eh e ~ e a s u r e ~ ~The n t . Q ~ j e c ~ of v e these 0th p e ~ f o ~ a n of c e the re~ainderof the circuit, contin~i~y.

9.13.

.3 Partial discharge map of 1i kV circuit

es have been lo eloped at the circuits in ~ o ~ ~ i s s ~ o n ~

Power System Restructuring and ~ere~ulation

302

.4 Zero sequence impedance values for 1 1 kV circuits

S

ers’ expectations for the reliability of e ~ e c ~supply c i ~ have signi~cantlyincreased in the last 30 years and this trend is likely to continue. Reflecti~gthese expec~tions,the regulator monitors closely the performance of the electricity distribution companies and strongly encourages them to reduce the number and the duration o f service i n t e ~ p ~ i o ~ s . Some of these i n ~ ~ ~ pare t the ~ ounavoi~ab~e ~ s consequence of essential m a i ~ ~ e ~ a n c e or repair work, A few results from operating errors while a significant number are caused by acciden~lor intentional damage to the equipment. However, a large majority of these outages is caused by ~quipinentfailures. The rate of occurrence of intemptions caused by premature ageing or deterioration could be reduced if ail the installed equipment were replaced by new equipment.

Asset ~ ~ a g e m ~ n t -

303

~ o n s i d e ~ nthe g enormous investment that such a replacement would represent and the in the demand for electricity, this re~rbishmentmust be d of time, To optiinise this replacement pro~amme,it is in new equipment are likely to have the largest effect on the reliability of service, i.e. to know which equipment is most likely to fail soon and ought to be replaced first, If failures occurred on a purely random basis, rep~acingany piece of equipment would have the same effect on system reliability. On the other hand, if it was possible to show that a single factor (e. insulation used for cables) has a much stronger negative influence on the any other factor, the replacement policy would be simple: all cables ~nsulationshould be replaced first. A review of the existing literatu suggest§ that the actual situation is considerably more complex than either of these extremes, For example, while it is clear that cable failures do not occur on a purely r ~ ~ a number of factors seem to contribute to their probability of failure. These factors include age of the cable, the method of installa~on,the type and e cable is buried, the instantaneous and historical loading of the circuit and the previous o c c ~ e n c of e faults in a particular cable section. Faults are comparative~yrare given the asset base and have multip~ecauses. As a resu~t, chance is the scourge of fault research. The same unsafe behaviour may in one shed yet in another result in a catastrophic fault. All sorts of external ce the outco~e:weather, co-workers, Ioading, mechanical failure, prediction of large amounts of variance in fault likelihood extremely difficult. Future research, having demonstrated a relationship between an unsafe b e h a v i o ~and faults, should then focus on the inves~~gation of factors that predict that unsafe behav~our. This change of focus has ready happened to some extent in relation to driving acci It is well established that driving above the posted speed limit is predictive of road traffic acc~dentsin the long run. However, any attempt to demons~atea direct link between as measured in a single &udy and the occurrence of accidents within that study is to meet with success. Most speed in^ goes unpunished by negative consequences. wever, that does not mean that speeding is not ~ p o ~ inn accident t causation, Therefore, much research is now dedicated to determining the characteristics that are associated with this dangerous driving behaviour. This approach could also be a~optedin fault causation analysis. Cracking down on relatively small numbers of repetitive faults may have ted effectiveness in changing overall performance (though it i s vital in terms ing specific ~epet~tive failure targets). What i s required are co~termeasuresdirec~edat the whole population. Weather-related faults would appear to be such a group where the fault ot be located at the extremes of the normal distribution^ The problem of faults may require an approach which focuses on fault causa~ionmore broadly conceived, rather than maintaining a rather narrow interest in individual differences in fault liability. It is recommended that future research also consider this perspective So far, researchers into fault liability have focused almost exclusively on those factors that predict inc~usionin the fault group, which in most populations is much smal~erthan

o

the no"fau1t ~ r o and u ~ subject to a high chance factor. Perha~sfuture research will also evote a~entionto those networks which manage over a long perio and e the factors that promote fault ~ v o ~ ~ aItn would ~e, ink a i ~ at ~ encoura~ing d such factors. A n ~ t ~ §ensi~le er shift the focus af ~ n t e ~ e ~towards ~ o n s the ~sitivebenefits of av ative ~ f f e of'faults. c~

a m e s had been ~opularfor a n

d as the unit cost of' a

30

Asset ~ a ~ a g e m e n t

~ r ~ d u c t iLevel vi~

~~~~~e9.5 ~ e n c ~ m a r kperformance in~ matrix for subsration m a i n t ~ n a n c ~

9.14.2

Asset Lijecycle

~ a n a g ~ of ~ ea nlarge ~ portfolio of assets also necessitates tbe ~ a f l a g ~ of ~ ~risk. e~r ~ ~ i s ~ o ~ cthe a lgrowth ~ y , in usage of electricity has not been linear and we should not be Pised to find that our asset base has not been c o n s ~ c t e dat a c o n t ~ ~ u o rate. us 9.6 details the a ~ ~ r o x i age ~ a profile ~e of London Electricity’s major assets, i n d i ~ ~ peaks of i n v ~ s ~ ~~ f ~l t rthei 1960s. ~ g

Power System R e s ~ c ~ and ~ nDeregulation g

30

Age-related replacement of assets will clearly lead to similar peaks in invest men^ in the future. Asset management techniques, such as condition based monitoring ( C ~ ~can ) ,be used to extend the life of individual assets - assuming that they are in good condition. C can similarly warn of the need for early repIacement without the need for failure to occur. Another useful technique which is available to companies with dynamic networks is to use other work as a driver for replacement. This is best illustrated by the following example. A ‘typical’ substat~onconstructed in the peak i ~ v e s ~ eperiod n t of the 1960s would be a 4x 15 MVA transformer site with 16 1 1 kQ feeders. Its modem-day equivalent would be a 3x60 MVA double secondary transformer site with 36 feeders. Reinforcement of one substation in an area can normally enabfe a hrthex two similar substations to be removed, thus avoiding the need for replacement. Extensive use of this technique normally requires an element of load growth. Even if we do opt for an age-related replacement programm~,we need to plan for R more gradual replacement programme. The easy option is to replace assets before they reach the end of their useful life. Our task as asset managers is to manage the risks associated with pushing assets closer towards the end of their usef~llife by i n ~ o d ~ c i n g alternative options, or devising ways of closely monitoring their performance. The actual life in service of assets may frequently be observed to be lower than the accoun~inglife of plant, as used for depreciation by compan~es,or the much higher assigned service life. This difference has normally been driven by reasons other than replacement needs such as: upgrade for load growth, faults, change of b ~ i l d ~ noccupancy, g diversions, etc Figures 9.7.and 9.8 show examples of actual life in service where this has been less than the assigned service life. The data represents all secondary transformers and secondary switchgear removed from the London network since 1991.

-

Actual life in service 5 e c ~ n d ~ransfor~er§ a~

-

36 years) (average age at ~ecQrnis~iQ~ing

I

F

~ 9.7 Actual ~ ~ life in ~ service e - secondary transfomler

07

Actual life in service. secondary switchgear (average age a t decomissioning . 34 years) 10.0%

8.0%

6.0% 40% 2.~94

0.0%

43 48 37 34 31 28 25 22 19 16 13 10 7 ure 9.8 Actual life in service - secondary switchgear

Condition monitoring and assessment provides a very useful guide to the s that need a ~ e n t i owithin ~ the next review period but are less usefui at present i n v e s ~ e npt ~ a ~ i n g . Several models have been used by London Electricity to assess possible asset rep~acementrequiremeiits in the long term (e,g, beyond the next rice review period). These use a number of teclini¶ues for projecting the current profile of assets using d~fferent replacement regimes. The most e l e ~ e n t areplace~ent ~ p r o ~ l emodel is one that gets r e p ~ a c e m eo~ ~ assets in the year they reach the end of the assigned service life. his will have the effect of recreating the same age profile curve as the present population The models used by London E l e c ~ c apply i ~ a spread of replacement ages c e n ~ ~ e d around the ~ s i ~ n service e d life. This is ~ n t e ~ d etod rep~esenta more r e a ~ ~ sview t i ~ of the range of ages at which assets will be replaced, caused by the impact of the widely v a ~ i n ~ drivers for rep~acementsuch as safety, obsolescence, eny~ronmentan The shape of the replac~ment profile can be selected to represent how wide the variation from the average service life is likely to be. The most simplistic ap take a flat profile, which replaces an equal proportion o f the asset popula$~onover a given period of time. Figure 9.9 shows 7.5% of the population replaced each year over eriod. The effect o f this is to create a new profile of assets which is smoothe has a wider spread by I5 years.

T

T

Ex

31

Power System ~

~

s

~and~~ e ~ c e ~~ l a it i onn ~

~aintenanceis only one aspect o f effective asset ~anagement.The ability to exten useful life o f an asset can be based on the amount and quality of the r n a ~ t e n ~ c e out, but it can e q u a ~ ~bey affected by the ability to b a ~ ~ bc een e ~ ~a sg ~ n costs s ~ in the overall replace~entand investment strategy. ~ o l e s a l ereplace men^ of assets is m expensive business and we need to ensure that our investment is always t ~ g e t e dat those areas which provide the most benefit. London Electricity has been deriving a methodology for devel ers in the c o r n p ~ yand p invest in order to m a x i ~ i s s assigning values to n o n - ~ o n e benefits t~~ as well as e s t i ~ t i n g~ o t e n t i acost ~ involves c o n s ~ c t i n ga model o f the project or ~nfluen~e diagram to ensure that all internal and external influence in eyaIuatin~the benefits of a particular project and the way in which it is i~plemented. .12 indica~esthe ~ u ~ u ~ a t ci ov se ~ ~ n e fana~ysis it of a Prom this we can judge which projects provide the nt and allow us to prioritise within budget or caqh flow GO

"12Cumulative c o s ~ e n eanalysis ~t of project portfolio

The steep slope at the beginning o f the curve indicates that the projects at this end efit to cost ratio, whilst those at the other end appear need to recognise the impoi~anc~ o f a less b ~ e ~ c ~ a l her, more beneficial, project upon it. This is illus in the c m e .

Asset ~ a n a ~ e ~ e n t

11

Each of these individual projects can similarly be evaluated against a variety of such as: do more or ~ u ~ ~ kdo e rless ’ or slower, do nothing, etc. This enables ev a~tenancedecis~onsto be calculated as well as the rep~ace~en~re of optimal rep1 eh~ Deferring rep~aceme~i~, refurbis ent QT maintenance always has a risk a ~ s o c ~ awtit it. The use of a f Q ~ amethodotogy, 1 which evaluates costs against bene~ts, useful risk management tool for all the staff associated with and affected by the ~ecisiQns taken. The variabili~yof eleme~itswithin each project are also assessed for c r i t ~ ~ toa ~ i ~ eliminate statistical unce~aintyassociated with those elements which do not sign~~cantly impact on the overall project. This allows us to concentrate on those elements where we need to be more accurate in a$sessingprobabili~~es or variabilities.

London Electricity has been developing a technology strategy to ensure that all pot~n~ial for the network are c o m p l e ~ e n tto~ each 0th ogy strategy has been to ensure that state-of-tl~e-~ evaluated and potential operating cost savings are i d e ~ t ~ ~ e d . The various s t r ~ d of s the t e c ~ o l o g ystrategy all need to build tow objective o f p r o v i d ~ ~the g degree of network p e r f o ~ a n c erequi can be evaluated on its own merits but, in general, those projects eluded in the i n v e s ~ e npo~folio. t of the remote t e ~ i n aunit l (RTU) ~ u ~ e n t ~ y he 1 1 kV network. ~ d d i t i o n afeatures ~ have been built into these units to facilitate the transfer of data from the LV s y s t e ~when n . kin s u ~ ~ bdev~ces ie have been ~ a n u f a c ~ r to e dobtain the required ~ n f o ~ a t i o This specification would nob be possible without such a cohesive strategy. Much o f the monitoring experirnen~l’but it is already possible to install power outage d i s ~ ~ a n csensors e ) in the premises of a customer who has s ~ f f e r e ~ will contact the control centre in the event of a s failure via a telephone line. Fault passage ind~catorsinstalled at on the LV n e ~ o r kp r ~ v ~ dmore e localised i n f o ~ a t i o nabout the positio which will eventually be relayed back to the office via the RTU. These RTUs also have the ability to ~rovideon-line loading and status information for the subs~tiQn,which can provide inval~bl~ e n f o ~ a t to ~ othe n network planners and analysts. Other work has c o n ~ e ~ t r aon t e ~ensuring that many of the i n ~ e p e n d e ~ developed t~y i n f o ~ a t i o nsystems, for con~ol,n e ~ o r kdesign and analysis, etc., are able to share i n ~ o ~ a t i via o n a ‘data hub’. A ~ ~ o ~major h e r task has been the development o f a more proactive version of the partial harge mapping ~ e c ~ i q mentioned ue previously. Continuous d i s c h ~ g emoni~orin and EHV f e e d e ~is economic and this, coupled with the ability to switch the network remotely, could facilitate the isolation of potentially faulty sections without

'

re' injection of gas an silicone fluid into the circul cables to remove mo re and fill voids is esrablis core les can be re~rbishedfor ~ i c a l ~less y than a w ~ e r ed ~ e c t - l a ~ d ently most HV cables in the UM have generally be r e p l a c e ~ ~ cost. ~nt paper insulate^ with lead or c o ~ g a t e da l ~ n i u msheaths which are u t e c h ~ i q ~At e . prese~there are no v~abler e ~ r b i s ~t e c~~ti q u e sfor cables but p r e ~ ~ research ~ ~ n work a ~ is under way to establish the i s h ~ e not f oil and gas press~rec f these circuits where t h e ~ a expansi l

s and the ~ o v e m e nof~ uce pressures and real yed. The presence of o s s ~ b i iof i ~limite is extremely rare conditions.

.16 assessment of the rel ~ ~ ewill n tinclu c relati~nscQnse~uencesto the who1 iate risk control n~easures. For power s y s t ~ ~risk s may be consi~e variables:

e its o w mix ~ of the co~ponentsof ris nt.

f~~1tS risks o

. ~ ~ e c t i asset ve m ts with higher that

1.1

witchgear reduces the risks o f failure. with timely r ~ ~ e ~action. ial

~ for this§

31

Power System Restructuring and

A u ~ o ~ a t esecurement d and/or remote restorat~on of supplie§ with r e a l ~ t i ~ e telemetry, which can significantly reduce the p o s s i b i l ~of ~ overloading and s e c o n d a ~failure which if sustained may cause far more exte~sived a ~ a than g ~ the erhaps simple, failure.

mising the number of customers affected through active risk m ~ a g e m e n desi t systems and the use of appropriate protection zones, owing the s t a ~of s the network and keeping cust~mer§advised. viding restorat~onin seconds or minutes, not hours. g ~ e q ~supplies t e of spares and skilled resources. hing c o n ~ ~ n g e plans n ~ y and r e ~ l a r l yexerci~~ng them, Laying off some of the financial risks, contract exclusions and i n s ~ ~ c e . ~ e , as the c o ~ n c i d e nloss ~ ~o f Some events that may ~nitiallybe cons~deredi ~ p r o ~ a bsuch ~ u ~ t i pindependen~ le circuits or of substations, may neve~he~es§ be w respect to the physical, political and economic environment. Some m their flight paths, failure of flood defence^, e a ~ h ~ u a k ete~orism, s, a1 disputes, computer viruses, etc., may when combined with s~ppliesto central business districts, CO the n~edia9 security and transport services. Asset ~ a ~ a g e m e strategies nt for these si~ationsmight from outside the ~ n v i r o ~ e zones, n ~ l fall-back or manned generation or just a ~arge~ed set of cont~ngen~y them.

r ~ncid~nts are f o ~ n a t e l comparative y rare will account for around 10% of customer incidents often seem to arise from a unique set of circ o te types of events using large pop~~ations ~nderlyingp a ~ e ~ strends. Such analysi proportion o f sue for e ~ ~ ~ pthat l e perhaps , a h s and that even these are most often as ~ n § ~ l l eord ~ a ~ n ~ ~orn that e d a9 cerlai u~tomers~ e i n ga ~ e c t e dfor ons can result in 1 # ~ ~ g for e si ~ s ~ e c t ~an ons times w h i ~repairs ~t are e ~ e ~ ~ e d , lar e ~ u i p m e ncan ~ r ~ s a § i ~ n i ~ c aloss n t of r e s o ~ c e romised for prolonged p e ~ o with Erip testing is a key p e r f o ~ a n c e~ d ~ c a tthat o r the ~reakerwill do so. With the increase in remote ~ ~ nfaeili ~ o l system^ from con^^^ ~ e n ~and es ied out on both p ~ dmsec on^^ ~ ~

will inc~~asingly be performed an reported automatically, rdeasing maintenance staff to tackle other activities. Correct ~nstallationand c o ~ ~ i s s i o n i nofg lant and equip men^ is critical to both life cycle costs and system reliability: ro~ec~iQn o~eratio~o~e~ after t i vmodi~cations e and circuit out Primary system and bwbar ~ o d ~ ~ c a t ipost-commissioning ons ins with t h e ~ and a ~discharg~§ ~ e y s . Exercise MSS circuit breakers remote1~re larly, e.g. twice a year. 1 inspection o f outdoor i n § ~ l l a t i ~and n s precautions against flying debris. mise repair time on first circuit outage.

9.163

Type ~ ~ i ~ u ~ ~ s

The economics of purchasing often meam that large ~uantitiesof the Same ty sw~~chgear, ~ a n s f o or ~ ea ~n c i ~ l aequipment ~ are p u r c h ~ eand ~ ~ n § t a ~ ~ine d p r o x i m ~on~ networks hat are being built, e x ~ e n d eor~ ref~rbishedat the Experience has shown that whilst the widespread catastrophic failure o equipment is rare, problems that could lead to longer term failure identi~edcons~derab~y more often. ~ e a ~and t h safety consid~rationsm live operation o f the plant to be restricted until after it can ection and modi~cation. e failures can present the operator with very e sections of networks could be rendered hop con~~tiQns. This is icularly the case with wholly ~ d e r ~ Q u n d e line overhead work does not e where the oppo~uni u n d e ~ k live Is0 be necessary to effect the necessary re repeated outages CO

e failure can be managed by: ~ e l e c ~ enqgu i ~ m ~with nt roven excellent p e r f o ~ a n c erecord. Actively maintaining a d sity of ~ a k e types s ~ and versions o f e q u i ~ ~ et n t ~ networks so that a type failure of one e does not result in widesprea ne~orks. For new types of equipment and changes to existing desi m ~ u f a c ~ r to e r participate in formal and independent failure can ~ d e n tand i ~ addre$s the probabili~and consequ

Common mode failure can occur where a single incident places a n component§ at risk at the same time. Typical causes o f common mode fai~~ires are: age to overhead lines. e c h a n ~ ~excavatQr al damage to several cables in the same ~ e n c h .

o

~

Power §ystem ~

e from frre in a $ w i t c ~ ~

e

~

t and~

c

~

n

~

tive regimes that are increas liver r ~ ~ u ~levels r e d of se loss arising from IOSS rcial tower block coul be found liable for the

Asset ~ ~ is the key ~ to effective o asset ~~a n a g e m ean thowev~r ~ ~ ~ data to c a ~ ~ rhow e , often how to store it an then how to use it e f f ~ c ~ i ~ e or w i ~ ~ cost. o~t too much data in appropriate^^ large costs in sy$te~ r ~ s ~ o nto s ea st m a ~ ~ t ~ n ~ ~ ~ . There is a real cost to collect~~g and ~ i ~ ~ i n ~ ~

~ ~ m ~ time, t e d in order to r factors can be a simpl

~

Power System ~ e s ~ c and ~ m g

cally dis~ibutionc Q r n ~ ~ ihave e s a n u ~ e r o u asset s r n ~ a ~ e m eIT n t s y ~ t ew~ s~ ~ c h aged effectively can exhibit ~ r o b l e in ~ sthe f o ~ ~ o w areas: ~ng

unavai~abilityof data for strategic analysis and business reporting, and ~ ~ lof~a ~dt os~ a t i othat n cannot exchange information. These ~ ~ Q b ~we e rdue n ~to the lack of a strategic ‘integration architecture’ enabling the easy ~ e v ~ l o ~ mand e n texecution of electronic ~terfacesand the rocesses required to on solid ate ~ n f o ~ a tfor i ~strategic n analysis and The IEC 61968 series ‘System Interfaces for on ~ a n a ~ ~ ~iseintended n t ’ to fac~litate inter-application integration of the various d i s ~ i ~ u t esoftware d a~~lica~on orting the rn~agementof utility efectrical ~ s ~ b u tnie~~ no r k s ~ Figure 9.13 clarifies the scope of IEC 61968-1 graphically In terms of business ~ n ~ t ~ oand n s shows a distribution m a n a g e ~ e nsystem ~ with IEC-6 1 9 6 8 - c ~ ~ p l i a n ~ ~ n t ~ ~ f~a cc he i t e c ~ r e .

Distribution ~ a n a g ~ msystem ~ n t with IEC-61968-com~liantim~erfacea r c h i t ~ c ~ r ~

Asset ~ a n a g e ~ e n t

9.17. I

Asset ~

~

~

a~ ygs t e~ ~ s~

e

~

t

Asset management systems typically hold data, including ownership costs, on all the electrical assets, linking them together via parenuchild relationships. These s y s t e ~ s normally share a comprehensive power system model with other app~icationsso that operational and planni tools and data can be employed as part of asset manag~men~. scale of the data ~ n ~ e ~ r systems a t ~ d are mparatively new in many companies and a chance to prove collection va~idationcan be immense, SO many systems have not yet ment are as plant their full worth in monitoring lifecycle costs etc. Proven uses at th database and ma~n~enance schedulers Iising network analysis and continge~cy tools. ~ a r i a b l ema~tenance~ g g e r cs be set within the database and co~dition r inspection visits, and defects recorded from mainten The p e r f o ~ a n c eand effective life of otherwise identical assets is largely driv duty they are required to perform and the environment in which they operate. An benefit o f inte~atedi n f o ~ a t i o nsystems is the ability to download large secti for ‘off-line’ analysis and data mining to understand and exploit the re~ations~i p e r f o ~ a n c eduty , and environmen~.

The data required by asset managers typically resides in several s te databases each of which has to be desi opulated and mainta~nedby an ve set of busi~ess pr~cesses. The equipment database contains information about the items of plant and circuits which make up the dis~ibut~on network. The volume of the assets and the varie i n f o ~ a t i that o ~ i s a~ailableabout each and every type are very jar are p a ~ c u l a r lcomplex, ~ ofcen requiring multiple spatial repre potential users of the data to access the i n f o ~ a t i o nthey need manufac~rer,speci~cation,age, insta~lationmethod, condition, loading, electrical parameters, etc. The recording of costs again ard as cables are continuously being cut into n atabase will typically desc geo~ra~hical and circuit location of the faults which have t network over at least the last three years an customers. of estimatin~the number of custo~ers widely from company to company, Tigh~er r e e l arrangements will require consistency o f reporting be ing o f end use custom a solution to this issue ase could be a major unde~akin c ~ n n e c t i vdown i ~ to indiv~dualckcuits and phases at LV with~nthe d ~ s ~ i b u tsystems io~ ividual blocks of fiats and offices. The costs of such systems and their ma~~tenance not easily be j u s ~ ~ ~ when e d the consistent applica~onof simple e ~ t ~ ~ a t i o could well s u f ~ c e .

Power System ~

~

~and t

~

c

at the most reliable data comes when it is c o i ~ in~the c field ~ ~ by ~ s of faiIure ~~d who the data co~~ection s and to the c o ~ ~ a ~ y . ce of accurate data collection e n ~ ~nleea r - r e ~ ~ - t i ~ e a collection t e ~ ~ n aand l s radio te data with insta~tq ~ ~of ev a ~~ ~ eor ~§ § e~~ u ie n c~eof~ d a t ~that ime series data that is routinely requi t, s u as~t r a~n s € o ~ eand r fee^^ ~ o a ~ tern ~ g , is no longer cost effective in a t i g ~ t ~regul y ngly s e c o n d a ~sub able to provid~CO

once^^^^

the ~

the ~ n v i ~ o n ~ e f l t ~u~ circuits and

~ in, t hor o ~ ~or ~hov e This will ~ s u a l ~inc y

~

21

~ubte~ai~ uilt e n v ~ r o ~ e n ~

idity, water table, s ~ n ~ s w e l l po~e~tial, resistivity, stability on, co~osion,vibration, tion, thermal s ~ ~ r c edas ,

y no ~ e a exhaustive. ~ s

ill also need to r n ~ the ~ effect a ~ ~ e, Pr the material. patia~i n f Q ~ a ~ icoonn c e ~ i n gc Q n s ~ a ~such ~ t sas: e

areas of o ~ t ~ t ~natural n d i beauty ~ ~

is also es§~n~ial.

y people tend to include e~oneousor ~ncompleterec

can also be subject to CO

Power System Restructuring and ~ ~ r e g u ~ a ~ i o

tier registration information, and yarticularIy information that wou~df a c ~ ~ i ~ a ~ c u s t o ~ esegmentation r ana~ysis, ust tom er c o n s ~ p t ~ data, o n and particularly half-ho~lyload shapes. Advance information regarding the setting of future DUOSprices. ~ ~ f o ~about a t the ~ c~r end ~ ~ - w o ~ hof ~ esuppliers. ss Asset ~ a ~ a gwill e ~need s to ensure that only those having a p p r o p ~ a rights t ~ and a ne ow can have access to, copy or export ~ f o ~ a t i that o n is to be regarde~as ~ o n ~ d e n t i a l ,

As operating margins become smaller and further efficiencks becomes more difficult to ~ e ~~ ~~ ~~ r~q~~ $ aa, lsof i ~~i ndf o ~ a t ~ obecomes n ever more essential for the eff~ctive t it is essentia~ For r e ~ l a t as o ~well as asset m ~ a g e m e npurposes

~ o m p a r a bacross l ~ companies onsistent over time Collectable at reasonable cost. ata c o l ~ e c ~will d be used for asset r e g u ~ a t o ~ c regimes, there will be a need to be able to demon and timeliness of the data. One way of managing q u a ~m i ~a n a ~ e ~ e n t cesses such as those QE the IS0 900 audits that run s ~ c ~ e d sequenc~sof data studies, ~ o ~ ~ results a ~ with n gthose p~viouslyobt for ~nvestiga~ion, can also be used to s ~ ~ i ~ cadv ant

None of the above examples indiv~duallycan provide the solution to the problem o f how we ana age our assets. A combination of all, or at least so^^^ of th s o l ~ t i ~which n , matches the point on the e v o ~ u t ~ o cn au~~ which e L at this moment in time. The only ~~~~~g that is certain is that ' overall model will continue to change as more information bec ome more e s ~ b l ~ s h eord more varied, or if p r ~ s s u r ~ other s~keholderspushes investment decisions in a new is to eva~uatec Q n t i ~ ~ a l the i y b e n e ~ t sof in& inst the cost of installation and operation. We the principle of condition ~ o n i t o ~ and n g data collection, lest we forget to en~~ally high cost associated with both the co~~ection and an ~eaTingin mind that the most effect~veway of i ~ e n t i ~ i when ng of a ~ a ~ u a l~l y~ ~ e n item d e of n ~ ~ u ~ p msuch e n ~as, an isolator, requires m a i ~ t ~ n aisn to c~ ask tbe last p~rsonwho o ~ e r a ~ it. ed

23

Asset Management

sis

t

Large power t r a n s f o ~ e r sare probabIy the most important equipment in an e ~ e c ~ c a l system. Correct diagnosis of their incipient faults is vital for the safety and ~ ~ I ~ aofban~ l i ~ in-service ~ a n s f o ~ ise rsubject to electrical and the c l e c ~ c anetwo ~ n the insulating materials and release gaseous which can bre products. Qverheat~g,partial d i s e h ~ g eand arcing are three primary causes of faultrelated gases. There are many i n t e ~ r e ~ a ~methods ive based on DGA to diagnose the nature of transfomer detcnoration, such as the IEC ratio codes which were developed from ions on gases generated from individual faults. has widely been used in the industry, in some cases, the conv~nt~onal osis incipient faults. This normally happens for those tran e 1ype of fault. Actually, the conventional d i a ~ o s ~me ic based on the ratio es generated from a single fault or from ~ u l t ~ faults p~e one of dominant nature in a transformer. When gases from more than one fault in a t r a ~ § f o ~ are e r collected, the relation between different gases becomes too comp~ieated ~ e d and may not match the pre-de~nedcodes. For instance, the IEC codes are d ~ ~ from certain gas ratios. When the gas ratio increases across the defined limits (boun~aries),the code changes suddenly b e ~ e 0, ~ 1nand 2. In fact, the gas ratio boundary may not be clear (i.e. fuzzy), especia~~y when more than one type of fault exists. there fore^ between different types of faults, the code should not change sharply across the boundaries, A new m e ~ ~ has o d been developed to employ fuzzy boundaries between differen~IEC codes.

9.19.1

The~~CD

, the IEG codes have been used for several decades and eonsiderable ~ x p e ~ e n c ~ a c c u ~ u I a ~ ethroughout d the world to diagnose incipient faults in t r a n s f o ~ e used to determine each ratio and its assigned limits are shown i s are then allocated according to the value obtained for each ratio corresponding fault characterised, .19.2

The Fuzzy IEC Code - Key Gas Method

The fuzzy IEC code-key gas method (FIK) developed is a eomb~nationof diagnosis using IEC codes and key gases. This method produces nine fuzzy comp onents are related to the fault types as d e ~ ~ i in e dT

IEC codes

Fault c l a ~ s i ~ c a t~i oc~c o r t~o nthe~ E C Gas

1 or2 ischarges o f high energy

2

Thermal fault of low t ~ ~ p ~ ~ a ~ ~ r e

1

1~0-330kV power ~ a n s f o ~ ewere rs

with IEC me~hod,the FIK method also h es, 13 ~ a ~ s ~ ocould ~ e not r s method, as shown in Table 9.3 Its may be only at the early ive a s ~ o n g e indica~~on, r s

nspectian of anothei re d ~ due to a e ~

~

charge o f high energy

Thermal fault { 1~0-300"~) Thermal fault ~300-700'~) Actual fauIt will be Thermal fault (>70QoC) checked during the next o v ~ r h ~ ~ ~ .

Power System Restructuring and Deremlation

326

____

F(0)=0.525 F( 1)=0.053

Normal ageing PD of low energy

IEC cannot diagnose but FIK indicates a

F~2)=0.231 I00

121

No match

No match

F(3)=0.045 F(4)=0.050 F(5)=0.000 F(6)=0.047 F(7)=0.000 F( S)=O .050

Discharge of low energy Discharge of high energy Thermal fault (450°C) Thermal fault (150-300°C) Thermal fault (30O-70O0C) Thermal fault (>700"C)

F(0)=0.005 F( 1)=0.052 F(2)=0.052 F(3)=0.000

Normal ageing PD of low energy PD of high energy Discharge of low energy Discharge of high energy Thermal fault (<150°C) Thermal fault (150-300'C) Thermal fault (300-700°C) Thermal fault (~700°C)

F(7)=0.161

.431

F(0)=0.479 F( 1)=0.005 Low values

No diagnosis

F(4)=0.0 13 F(5)=0.000 F(q=o.ooo F(7)=0.000 F(S)=0.005

Actual fault will be checked during the next overhaul. IEC cannot diagnose probably due to the e x ~ s ~ ~ n of c e more than one fault. The fuzzy compo~ent of the early thermal fault indicated by FIK is useful for future trend analysis.

Actual fault was an arc damage to the core. IEC diagnoses Normal ~ e ~ ~ PD of low energy fault but actually PI3 of high energy Discharge of low energy both medium- and hi~-tempera~re faults existed as Thermal fault (cl50'C) Thermal fault ( 150-3OOOC) indicated by FIK. Thermal fault (300-700°C) Two locations o) Thermal fault (>7OO0C) overheating damages were Jound due lo eddy currents and a bad cantact.

F(O)=0.007 F(1)=0.026 F(2)=0.026 F(3)=0.000 Thermal fault (300- F(44)=0.030 700'C) F(s)=o.ooo .003 ,477

P

which could be at an early stage.

-

Normal PD of low energy PD of high energy Discharge of low energy Discharge of high energy Thermal fault (450°C) Thermal fault (I 50-300°C) Thermal fault (300-700'C) Thermal fault (>700"C)

Although the gas level is below the guide value, an early indication of low. energy discharge by FIK should be useful for trend analysis in the future. Actual fault will bc checked during the next overhaul.

~

Asset Mana~ement

9.19.4

7

~~e~~~ ~ u l y sqfi sI ~ d i v i ~d ~ ~a ~ u

~

~

~

In FIK ~agnosis,a fault can be more accurately determiiied by its fuzzy component that indicates the likelihood or dominance of the fault. Deterioration of the fault may eref fore be closely monitored from trend analysis. This technique has been used for a that was tested over a 15-month period. ermal faults of medium- and high (300-700°C and >7OO"C) were diagnose y the FIK method and the fuzzy agaiiist the test time are plotted in Figure . The graph clearly shows the de each thermal fault in this t r ~ s f o ~ eItr ,can be seen that at the begi monitoring period, the medium ~ e m p e ~thermal ~ r e fault F(7) was the main p r o b l e ~o f this ~ a n s f o ~and e r the fi~zzycomponent of the high-tempera~retherm mall, i.e. below 0.05. The high-tempera~rethermal fault F( 14 onwards and then become stable until Day 406 when the ssing, because the ~ ~ efaults ~ remained, a l the fuzzy compo went up again from Day 453. It took a few weeks for the gases to be re1 in the oil to a sufficient level for accurate diagnosis. A small fluctuation of F(8) was c recorded on Day 178, which might be due to the lighter load during the s p e c ~ ~time period. mponent F(0) always It must be noted that if a transformer has no fault, th gives a large value in th anga of 0.6-1. For example, results for a hea~thy t r a n ~ f o ~ are e r (in ppm} - 95, N2 - 73000, 02 - 11000, - 25, C,H, 45 and C2H2 2. The fuzzy component o f no-fault ~ ( 0 ) ~ . $ 4 3 at no fault exists in the ~ansformer.The IEC codes are 0, 0, 0, also i ~ d i c a ~ n g no fault. From OUK experience^ when the value of F(0) is between 0.3 and 0.6, an inci fault may have occurred at earlier stage. When the fault is getting worse, F(0) will decrease to CO, 1. ~

-

0.6

+.=.

0.5

0.4 0.3 0.2 0.1

0 1

114

147

178

406

218

191

413

453

469

471

a1 fault ~ 0 ~ ~ degree 7 0 0C

The trend of tv.70 types of thermal fault in a 330 kV transformer determined by the FI methad

method developed has been succes ers in Au5~alia.It has been proved that, using the

costs of ~ ~ ~ f o With ~ ethe~ aid 5 of . ~ e ~such ~a5 the~ FIK~method, ~ the ~ longer s c ~ ~ life c e could be achieved.

§

,

123

131

[SJ [6]

[7]

ario V.F. Pereira, Michael F. McCoy and Hyde BA. Merrilli, aging risk in the new power usiness’, IEEE ComputerApplications in Power, Vol. 13, No.2, April 2000, pp. 18-24. Ceorge Anders, Robert Entriken and Puica Nib, Risk Assessment and Financial ~ a ~ a g e ~ e n ~ rial, IEEE Catalog Number 99TP137-0,1999. Gorenstin, N.M. Cam~donico,J.P. Costa and M.V.F. Pereira, ‘Power systcm ~ l ~ i i ~ g under uncertainty’, IEEE Transactions on Power Systems, February 1993, pp. 129-136. ofilo De la Torre, James W. Feltes, Tomas Gomez San Roman, Hydc M. M e ~ I ~ , i~atiza~ion, and com~etition:~ansmissionplanning under ~ c e K ~~ ~E ~ ~~ E’ Transactions on Power Systems, May 1999, pp.469-465. J.C. Hull9Options, Futures and Other Derivatives, Prentice Hall, New Jersey, 1998. J. Schwager, A Complete Guide to the Futures: ~ ~ n d a ~ e nAnalysis, tal ~echnicalAnalysis~ Trading, Spreads, and O p t i ~ l lJohn ~ , Wiley & Sons, New York, 19%. Price W~t~rhouse LLP, The CorporateRisk Management Handbook, Risk ~ublications* London,

1996, [S] P. Jorion, Vdue at Risk: The New Benchmarkfor Controlling Market Bisk, Irwin Professional Pub., Chicago, 1997. ouglas, A. A ~ ~ i aV. n , ~iemcyer, . Goldberg, and C. Claxk, ‘ ~ a ~ i g a t ~the n gc ~ ~of ~ t risk‘, IEEE Power Engineering Review, March 1998, pp.6- 10, [I01 D. Duffie and J. Pan, ‘An overview of value at risk’, Journal of De~vatives,~ s ~ i ~ t i o n a ~

en & Co., The JP M ~ r g a n / ~ r Andersen ~ ~ u r Guide to cations, London, 1997. [123 G.L. ~ a s ~ i n e a~ictionary u, of Financia~Risk Management, Swiss Bank Corporation, New York, 1992. [I31 Eilron Capital Trade Resources, an aging Energy Price Risk, Risk ~ b l i c a t i o n sLon ~ 1995. E141 R.L. Nersesim, Computer Simulation in Financial Risk Manugernemt: A Guide for Bwiness P l ~ n ~and ~ r Stra~~gists, s Q u o Books, ~ ~ New York, 1991, [ 151 Q. Su, 6. Mi, L.L. Lai and P. Austin, ‘A fuzzy dissolved gas analysis method for the di of multiple incipient faults in a transformer’, IEEE Transactions on Power Systems, 2000, ~ ~ . 5 9 3 - 5 9 8 ,

Prof. JQS ~ ~ ~ l a g a University of Canterbury New ~ealand

University of Canterb New Zealand

to deregulation, electricity has been generally sold from one supplier to th ownership ~ h a n ~ i nhands g at only one piiysical point. In con~ast,after it is expecte~that the product will be exchanged at several points along th t~ansmi~s~on and distribution systems and there will be power quality (PQ) i 1 location where owner€hipis transferred. s, of course, an ~ b i ~ oterm u s which in its b r o ~ ~ esense s t is quality including reliability of supply, waveform In a d e r e ~ ~ a t environme~t, ed only nationa~ and act on the i n f o ~ a t i o nnecessary to pro~idesystem secu position, the grids can be unreasonably d ~ ~ a n ind ~ n ~ ation plant. In the long term, however, the expec~tio will find s ~ ~c o~~ p ge t i ~t i ofrom ~ r dis~ibutedgeneration, bo micro-hydro, wind and solar) and non-renewable energy ~ i c r Q ~ r b i nand e s fuel cells), the latter in the k i l o ~ rather a~ logy used in these energy sources involves power is now commercia~lyavaila~leCO links and FACTS ~ ~ e x AC ~ b ~ e power devices. At the generation level, an increase in the connection of IPPs ( ~ n d e ~ ~ ~ d e n t ~ Q ~ u c e rsuch s as wind and gas"~e11ed ~ i c r o t ~ r ~ i n ewith s ) p~Qrlyc

Power Quality

1

sy~ic~onisation will make PQ more difEcult to control. The increase in embedded ~enerationwill cause ~ r t h e voltage r ~ a g n i ~ variations de as well as introduce additiona~ voltage m a ~ i ~ steps d e [2]. Wind power is known to lead to an increase i severity. Solar power and the more advanced ways of connecting wind power wi an increase in h a ~ o n i cd i s t o ~ o nAt , the ~ a n s ~ ~ i s slevel, i o n the need for ~ y s t e ~ to transmit power according to contracts between the requested locations is a ~ c e l e r a tthe ~ d ~ ~ a for n d s ~ ~ ~ e s - c o FACTS ~ e c ~ econtrollers. ~ In the c o ~ p ~ n s a ~and i o nunified power flow controllers are expected to be used extensive~yonce they are shown to offer better technical features at reasonable costs. m planning under deregulation will be more difficult owing to u n c e ~ a i n ~ in the gene~tionand load locations, fast solutions will be needed to improve the o~erating conditions and FACTS controllers can offer such solutions with short delivery installa~iontimes. The use of a s ~ c ~ o n o ugrid s intercQ~ections,both national an i n t e ~ a t i o n is ~ ~also likely to increase with dere~lation. The control1 asynchronous ~ ~ e r c o ~ e c t is o rcurrently s limited by the switching restricti silicon-controlle~rectifier, which only permits two-quadrant converter direc~ionalactive power transfers, The a v a ~ l a b i l iof ~ gate turn-off permits four-quadr~tconverter operation and considerable developan on to ~ m ~ r o the v e effic d power h~ndlingcapabili~of these d of two~quadra~tor f o u r - q u ~ ~ a n t that w h e ~ e r in the ;lijynchronous link is be an important player in modern ~ a n s m ~ s s ~systems on ~ l ~ n and ~ nitsgi ~ ~ a c eeds to be carefuIly exa~ined, Power elec~onic whether in the form of as~chronousinterco~ec~ors, FACTS or custom power, have the poten~alto improve various aspects of e ~ ~ ~ control o n ~at cd i s ~ i b ~ t i olevel n may ~ i t i g a t evoltage v ~ a t ~ o n s ~ voltage sags. But the increased use of ower electronic controllers may introdwce new erns like a ~ d ~ t i o nharmonic a~ voltage distortion, especially in the form of higher order n a c o ~ p e t i t ~ environ ve t there will be reluctance to expand distribution sys~em, customer interaction, And, at the loads ~ h e m s e l ~i ~ § ~ costs will create an emphasis on local co~pensati or active coan~onents. Some of these changes tend to de loads of a cons~ant-powertype. more c u ~ ~wh nt ltage drops causing additional vo use of Compensation equipment may even become t of these prob~emsare not ex~~usive to dere~ulation. In fact, there is a c ~ ~ t i n u i n g s, such as adjustable speed drives, office equip~ent, and ~gh-efficiency fluorescent lighting. At the same time, sensitive ~ n f o ~ a t ~ o n ment, such as PCs, continues to be dispersed into power locat~onsthat previously were res~ictedto lights, motors and heaters. There is no reason to b e ~ ~ e v e this trend will reverse. ~ollowingderegulation, the power exchanges should be s~bjectedto close s c ~ ~ i on n ya continuous basis, This requires dynamic evaluation of the ~ e n ~ or s by a combina~onof and current waveforms, either by local ~ ~ a s ~ eexclusively ~ e a s u r e i ~ e nand ~ s sys~ems i ~ u l a ~ using i o ~ h a ~ o n i cstate estima~io~ tec~iques.The

latter should provide more intelligent an economical solutions for the control of the di§to~ionp r Q b l e on ~ a system-wide basis. ~ e r e ~ ~ a t ~ o n clear, for the most part, that the utilir the customer. After ~eregulatio~, however, who i s responsible for the enerator? The e ~ e r g ysup lier? The d i s ~ ~ ~ t o r ? to con~sion,and po~siblyto an i n c r ~ a s ~ in d i ~ ~ t e s .

the quality ofpower has become e cts that help correct PQ problems

y local electric utilities have

une~pec~ed b e n e ~ t sfrom moiiitori ta with ~ndividual~ at those custo

a d i $ ~ b a n c ei s a temporary deviation from the steadyIn the con~exto Its of brief durat~onor by sudden changes in w a v e f o ~caused dis~rbancescon~ideredby the ~ n ~ e ~ a t i o~n la el c ~ o t e c ~ Ci c a ~ age dips (sags), brief i n ~ e ~ p ~ i voltage o ~ s , increases (swells), oscillato~ ~ a n s i ~ nThese ~ s . are illus~atedin Figures 10.1 and 10.2.

Voltage d ~ s ~ r b a n ~ e s

.2 Voltage transients

supply ~ e ~ o rThe k . main cau$e

xtinction of discharge 1 of control devices; speed variation or s ~ o p p i nof~motors; trippin CQntactorS; c o ~ p u t e r system crash; or c o ~ ~ u t a tfail~re i o ~ in line commutated inve~ers.The effect of a v o ~ ~ g e

Power System Restructuring and ~ ~ r e ~ ~ a t i o n

dip on equipment depends on both its magnitude and its duration; in about 40% o f the cases observed to date, they are severe enough to exceed the tolerance standard ado~tedby er manufac~rers.Brief interruptions can be considered as voltage sags with 100% de. The cause may be a blown h s e or breaker opening and the effect an expensive s h u ~ d oFor ~ . a given system design and fault location, a certain number of c u s t o ~ ~wili rs be ~ ~ e cand t ethere ~ i s no way to prevent this process without major system s ~ c ~ r a changes. I-lowever, i n t e ~ p t ~ o due n s to over~oadare somewhat more ~redictab~e. These include overload of the whole system (due to lack of generation) as well as ind~vidua~ lines and cables, Voltage collapse can also be view as an overload situation, but in this case load shedding can alleviate it. In the pre-dere~lationera, load shedding took place accord~ngto utility ides. ~ e r e ~ ~ a t aIlows i o n utilities to offer i n t e ~ u p t i ~ land e non-inte~ptible supply. During Limes of overload or overload risk, utilities may decide to increase the inc~ntivefor customers to be i n ~ e ~ u p ~[8,9]. e d At present, this action only covers a very s r n ~ fracti~n i~ of the ~ n t e ~ p t i o but n s this will obviously change if the congestion in the system increases, Voltage swells are brief increases in r.m,s. voltage that sometimes a c ~ o m p voltage ~y sags. They appear on the unfaulted phases of a three-phase circuit that has developed a single-p~aseshort circuit. They also occur following load rejection, Swells can upset electric controls and electric motor drives, pa~cularlythe adjus~b~e-speed drives, which can trip because of their built-in protective circuitry. Swells may also stress delicate computer components and shorten their life. Voltage disturbances swells are classified as transients arid are caused by s u ~ d e nchanges

cw.

According to their duration, transient overvoltages can be di into sw~tchingsurges (du~ationin the range of ~ ~ ~ i i s e c o n dand s ) , ~mpuIsespikes ion in the range o f sing from power s y s t e ~switchin microseconds), Surges are high-energy pulses ssociated with swit d i s ~ r b ~ c eeither s , directly or as a result o f resonating circ capacitor s w ~ ~ h i n devices. They also occur during step load changes. In parti cause resonant oscillations leading to an o v e ~ o l some ~ ~ ethree to four times the n o ~ i n a l ,causing tripping or even damaging protective devices and equipment. ~ l e c ~ o n i c a l l y based controls for ~ n d u s ~motors a ~ are pa~icularly suscep~ibleto these ~ansients. Impulses result from direct or indirect lightning strikes, arcing, insulation b r e ~ ~ detc. o~,

10.I . 3

Volta~e Sags

In Ehe present stage of deKegulation, no serious cons~derationis @ven to ~ a n s ~ i s s i oand n distribution levels and, therefore, there i s little incenti overall reduction in the f?equency of sags. Although there me indication will increase in the hture, some customers are likely to d e ~ a~reduction d in their nu~~eK. One option is to introduce ‘power quality guarnntees’ whereby the customer receives ~ o ~ p e n s a t i ofor n each event exceeding a certain severity (in ~ a ~ i t u d duration e, or frequency). Such an additional service may be offered by the (monopo~ised)distri c o m p ~ ~ yby, the supplier, or by any other pfayer in the market (e.g. an insurmce c o ~ p ~ yA) ~, t e ~ a t ~ vae regulatory ~y, body may decide to enforce a basic compensation

Power Quality

scheme for all customers as part of the connection fee [I 11. However, some customers may not be satisfied with any compensation scheme, safety being their main consideration. The option in this case is for the utility to offer high-quality power to a small customers. These customers will experience less voltage sags than similar customers elsewhere. This special service will require the installation of m ~ ~ g a t i oequipment, n which may be offered by the dis~butioncompany, by the supplier, or by any other player in the market. Additio~a~ regulations are needed to guarantee a minimum level of ~ o m p a t i b i l ~ ~ between equ~pmentand supply: R e ~ u i r e m e n for ~ equipment immunity must be produced by standard-se~ng organisations. The IEC is obviously the best platform for the development of such a s~andard.In the USA, the IEEE may take the lead. Standards for equipment test~ng,like IEC 6 1000-4-11 [ 121, are also needed to obtain and verify equipment immunity. As a complement to equipment immunity requirements, voltage characteristics for the supply must be made available to the customers. The E u r o p e ~s ~ d a r dEN 50160 should be extended with voltage characteristics for voltage sags and other events. Equivalent documents should be written for other parts of the world as well as local s t ~ d ~for d sindivid~dlcountries [13]. latory bodies should pub~ishstatistics on the PQ performance of uti~~ties. Such a e is already in place in the UK for long i n ~ e ~ p t i o[14]. ns Voltage sag ch~acterisat~on is an important basis for the above s ~ d ~ d regulations. At the time of writing, standardisation on this issue is under develo both in the IEC [4] and in the IEEE [lSJ. However, current activities concen sags experienced by sin~le~phase equipment. A technique has been proposed for the characterisation of voltage sags [16] e x ~ e ~ e n c be d three-phase equipment. It enables the characterisation through one complex vol wi~houtsign~ficantfoss of information. The method is based on the decomposition o voltage phasors into symmetrical components. An additional characteristic is introduc e n ~ b l the e exact recons~ctionof the three complex voltages. The m a ~ e ~ a t ibehind cs the method and additional examples is described in references [2,17-20]. The ITIC (Information Technology Industry Council) curve [21] shown in Figure 10.3 can be used to evaluate the voltage quality of a power system with respect to voltage i n t e ~ p t i o n ssags , or unde~oltagesand swells or overvoltage. This curve was ori deline in the design of the power supply for computer and electronic in the 60 Hz, 120 V distribution voltage system. By noting the changes of power supply voltage on the curve, it is ossible to assess if the supply is reliable for operating electronic equipment, which is generally the most susceptive equipment in the power system. The curve shows the m a g ~ i ~ and d e duration of voltage var~ationson the power system. The region between the WO sides of the curve is the tolerance envelope within which electronic e ~ u i p m e nis~expected to operate reliably. Rather than noting a point on the plot for every measured d i s ~ b ~ cthe e ,plot can be divided into small regions with a certain range of magnitude and duration. The number o f occurrences within each small region can be record~dto provide a reasonable indication of the quality of the system.

Power System ~ e s t ~ c and ~ n g

33

Percentage of nominal voltage (ms.of peak ~ q u i v ~ ~ e n t ~

110

90

0 Ims 3ms

Fi

2Oms

0.5s

1OS

Y State

ETIG curve

elet ~ a n s f (WT) o~ c u ~ ~ wav nt f r ~ q u ~ re n~y

ides a fast way of an the ~~~~~r ~ r a ~ ~

arly in the ~ r e s e n cof~ a

(10.1)

9

Power Quality

A sample mother wavelet

(10.2) he WT of a ~ o n t ~ u siQ ~ s

time e~~~~~of the w a v ~ ~ ~ t

are ~ i $ c r e t i ~but ~ dnot the i

Power System ~

e

~ and ~~ e ~ ~ c ~ l ~a t i o~ n

and the discrete wavelet coefficients are given by

(10.6)

Although the ~ a n s f o ~ a t i is o nover continuous time, the wavelets represen~tionis discrete and the discrete wavelet coefficients represent the c o ~ e l a ~ i obetween n the original signal and wavelets for different combinations ofm and n. The inverse DWT is given by:

= (A + B)/2, and A and B are the f i m e bounds (maximum values of a and b).

10.2.2

W a v Analysis ~ ~ ~ ~

lysis is normally implemented using ~ult~-resolut~on s h- and low-pass equivalent filters, h and g respectively, ana~ys~ng wavelet. The digital signal to be analysed is then decomposed (filtered) into smoothed and d e ~ i i e dversions at successive scales, as shown in ~ i g u r e10.5 where (24) represe~~s a down sampling by half, Scale 1 in Figure 10.5 contains i n f o ~ a t ~ ofrom n the Nyquis~~equeney(half the ains i n f o ~ a t i o n frequency) to o n e - ~ u ~ the e r sampling frequency, scale -quarter to one-eighth the sampliing frequency and so on. at any scale, with the final smoothed is is one of the sirab able s, i.e. scales 8,16,32, if it is . The choice of mother wavelet has a nt effect on the results obtained. The o ~ h o ~ o n aofl iwavelets ~ ensures that the signal can be recQns its ~ ~ s coeffic~ents f Q [23]. ~ Wavelets with s y ~ m e filter ~ ~ c o e ~ c i e n ~genera^^ s l~near phase shift. A large wav~letfamily derived by Daubechies [2 ] covers the field of o ~ h o n o ~ a l wavelets. It includes embers ranging from highiy Daub6 wavelets ape the best choice for short and fa ~ a ~ s id ei s~ ~t r b ~ cDaub8 e ~ , and Daub10 are the mo of a mother wavelet without knowledge of the types simpler solution is the use of one type of mother wavelet in the wh de~ect~on and localisation for all types of d i s ~ ~ ~ c e s .

Power Quality

xi31

scale 2

scale

. I .

M u I ~ i ~ ~ ~ s osignal l u ~ odne ~ o m p o s i ~implementation ~o~ of wavelet analysis

In doing SO, higher scale signal decomposi~ionis needed. At the lowest scale the mother wavelet is most localised in time and oscillates rapidly within a very short p e ~ o dof time. As the wavele oes to ~ ~ g hscales, e r the analysing wavelets becom~less loc~lisedin owing to the dilation nature of the WT analysis. As a result of time and oscillate 1 higher scale signal decomposition¶ fast and short transient d i s ~ b a n c e sare de~ectedat lower scales, whereas slow and long transient d i s ~ b a n c e swill be detec~e scales.

10.2.3 F i g ~ r e10.6a shows a s ~ u e n c of e voltage dis~rbances.To remove the noise prese~tin the waveform, squared wavelet ~ a n s f coef~cients o~ (SWTCs) are used at scales rn = I 2 3 and 4, ~ ~ s p e c t ~ (v se ~h yo in ~F 10,6b, c, d and e; these are analysed U wavelet. Figure 10.6a contains rapid oscillation disturbance (high fre time 30 ms, and is ~ o ~ l o wby e d a siow oscillation dis~rbance(low freque ms. The SWTCs at scales I, 2 and 3 catch these rapid oscillations, while scale 4 cat slow osci~latin~ d i s ~ r b a n which c ~ ~ o c ~ u ~ after e d time 30 ms. Note that the h i ~ h persist at the same t e ~ p o r alocation l over scales 1 , 2 and 4. It must be pointed out that the same technique can be used to det waveform distortion (like no~chesand h a ~ o n ~ cand s ) other momen~aryinter~ptions,sags and surges. ~ o w e v e r rig ¶ must be developed for each stutbance for the WT to be accepted as au~omatic~ ~ a s s i ~ cof a ~P o n

Power System ~ e s t ~ c t u ~and ing 200

0

0

10

20

30

40

50

60

70

80

90

100

60

70

80

90

100

70

80

90

100

e Voltage disturbance signal (0 1996, iEEQ

0

10

20

30

40

The SWTCs at scale I (0 1996,Z

0

10

20

30

40

50

E

50

The S~~~~ at scale 2 (0 1 9 ~ 6IEEQ ,

~

~

60

0

10

20

30

40

50

40

70

80

90

100

Power System Restructuring and ~ e r e ~ ~ a t i o n

42

.3

istor

W a v e f o ~d ~ s t o ~ i oisn generally disc~ssedin terms of h ~ o n i c s which , are s voltages or currents having frequencies that are w frequency at which the supply system is d e s i to~ operate ~ (e.g. 50 the ~e¶uenciesof these voltages and currents are not an integer of they are termed ~t nic and interha~onic~ s t o ~ i ois ngenerally caused by ~ u i p m e with non-linear voltage/c~ent characte~st~cs.In general, distorting equ harmonic currents which in turn cause harmonic voltage drops across the impedances of the network. Harmonic currents of the same frequency &om different sources add ed vec~orially.It is believed that, in general, harmonic levels tend to be i n ~ ~ e n c prima~ly by local and immediately adjacent conditions rather than wider zonal effects. The main de~imenta~ effects of h a ~ o n i c are s [30]: maloperation of control devices, mains signalling systems and protective r ~ ~ a ~ s , losses in capacitors, ~ ~ s f o r m eand r s rotating ~achines, ional noise from motors and other apparatus, telephone interference, and e presence of power factor corr~ctionc a p a c i ~ oand ~ cable capac~tancewhich can cause shunt and series resonances in the network roducing voltage ~ p ~ i ~ c aeven t~on oint from the distorting load.

As well as the above, i n t e r h a ~ o n ~ ccan s perturb ripple eantml sign& and at s h a ~ o levels ~ ~ ccan cause flicker. To keep the harmonic voltage content within the recom~endedlevels, the main solutions in c u ~ ~use n tare:

the use of high pulse recti~cation(e.g. smelters and R passive filters, either tuned to i n d i v i d ~&e¶uencies ~ active filters and conditioners.

I U,3. I

~ a ~ ~ o Sources nic

Lower order odd h ~ o n i c are s the most proli~camong consumer e~ectronic$yste~s. I~owever,the third harmonic (of zero sequence) is usually p r ~ v e ~ t efrom d en~erin high voltage system by the use of appropriate transformer connections. The fifth harmonic (in the UK) has been identified as the harmonic order exhibit in^ the highest peak levels of high v o ~ ~ systems, ge with values between 2.5% and 3.0%at some locations. The fifth also most ~ e ~ ~ e npresents t l y the highest mean harmonic levels, a characteristic which has been found to be consistent both g e o ~ p h i c a l l yand with time.

3

Power Quality

-n

m -7

1.0

r4

2 0.8 X I

8

@)

-g .-a

5.

0.6 0.4

0.2 01

11 13

Frequency (x fundamental frequency)

igure 10.9 12-pulse converter current: (a) waveform, (b) harmonic spectrum

23 25

‘The § t ~ d a cr o~ ~ ~ g u r a t for i o ~i ~ ~ u s a~ i a ~ ations is the ~ 2 ~ ~ u l s e cQnve~er,shown in igure 10.8. The c~aracteristic ation are o~orders12k-t-1 (of positive s e q ~ e n cand ~ ) 1%a ~ ~ l i ~ are d ei ns v ~ r § ~ pl yr o ~ o ~ to ~ othe~ ~a ~a ~ o ~ i c s ~ e c of ~ ~Figure m 10.9b which c o ~ e s p o n dto~the time wavvefo of course, ~ a ~ i 1 r 1s n for ~ ideal system conditions, e d ~ c eAC ~ s y s ~ e ~ a per~ectly flat direct c u ~ ~ (i.e. n t i ~ ~ n si~ to o~t h i n g en the AC system is weak and the o erfectly s y m m e ~ ~ a l ~ a r ~ o nappear. ~cs ile t~~ c~aracteristich ~ o n i c s it is not e~onornica~ to reduce in that way the un

devic~sand are, there

~ o ~ common h e ~ ex amp^^ of u n c ~ a r a c ~

3

els with v ~ ~ o levels u s of c o ~ p l e x are i ~ app~aring 1 n o ~ - l i ~ e cQ~ponents, ar such as AC/DC converte harmonic Norton equivalents. They involve iterative harmonic analy in~e~action b e ~ thee conve~er ~ and the linear system. Further work is 1 ~ ~ e Q u sthe l y effect of multiple ~tercQnnectednon-li The system s ~ a state ~ y is $ub~ta~tially, but not completely, desmib ark. In many eases, it is a s s ~ e dthat there the ~ n d a ~ ~ e nfrequenc~ tal and its ~armQnic$.

he cedwe used to solve the non-li~earequation set. a set o f accurate non-linear e

Power System ~

6

10.3.3

e

§

~ andc Dere ~ ~

n

~

~ ~ r ~ o Flows n i c (301

In its simples~form the frequency domain provides a direct solution of the effect of d in~ividua~ h a ~ o ~ or i c~ o n ~ h a ~ o~equency nic injec~~ons t~oughouta linear system, without explicit consideration of the harmonic interaction between the network and the n o n ~ l i c~oem~p ~ ~ e n ~ ( s ) . The sources of h ~ ~ o ninjection, ic depending on the available info linear c ~ ~ p o n e n tcan s , be current sources or Norton or Thevenin harmonic ~ ~ u ~ v a l eAn t ~ , c o ~ r n ~experience n derived from harmonic field tests i s the asyrnme~ica1n a ~ r eof the readin~s. justifies the need for three ,being the mle rather than tbe exc s. The basic compon~n~ of a t h e ~ ~ s m i s s line, ~ o nwhich can be accurate~yrepresented as earth return, skin model, including mutual effects a other n e ~ opassive r ~ n line m o d ~ are ~ s then combi~edw c o ~ p o n ~ ntot sobtain t~ee-phaseequivalent h a ~ o n i ic The system harmonic voltages are calculated by direct solution oflhe linear equation

is a reduced system a d ~ ~ ~ a rn n c ex of order equal to ( n u ~ b of e~ inject~onbusbar§,

current waveforms often have an ape~iodiccoinponent. The most c o ~ m o n iodicity in the ~ a v ~ isf o ~ and i n t e r ~ ~ o con~ent~ nj~

also pro~ucesvoltage ~ ~ c ~ a t and i o lni ~~h t~ ~ c ~~ eor ~. e c t toi othe~ voltage level and the of series reactances .The conve~tiona~ P account the ~ p e r i o ~co~ponents. ic For example, the total h a ~ o n i cdistortion ( D) is basically a ratio of the en~rgy t ~is possible to d e ~ n ea onics to that in the ~ n d a ~ ec o ~ p o n e n It r the aperiodic case by defining the power f r e q u ~ ~ c(and y there I, component), and then using the ~ e m a i n i nportion ~ of the s nu~eratorof a ~ ~ - 1 i index k e [39]:

Power Quality

where the power ~equencyis denoted as cooand E[.]denotes the calculation of the ener of a time signal. The prime on th D indicates that this is not quite con~entionalTHD ca~cu~atiQn. Of e, TMD degenerates to TIID for the p e ~ o d i ccase. With re~erenceto the flicker disturbance, the measurement and frequency windows in which flicker exists is d e ~ n e d inte~ationalstandards, mainly thro C). Generally, flicker i s limited to ~ l e c ~ r o ~ e cCommission ~ica~ fluctuations in the supply voltage. A proble~aticf ~ ~of ~thisr index e is how the flicker is to be m sured. As an examp~e, power frequency) be should the flicker energy (i.e. sideband energy in the vicinity of measured in root mean square a m p ~ i ~ dore ,zero to peak? m e a n i n ~ to ~ lintegra~ethe sideband energy over a latter appears to have less phys~ological implic mathemat~cal properties. Also, the integration of energy physiologic~~ weigh~~ng factor as specified by the IEC stand tran§form, short-time Fourier transform, and Fourier linear combiner have been sugges~ed as possible solutions to the problem.

with the intent of s u m m a ~ s ~ rms of the active power loss distortion, and the ~n~er~erence on telephone and data communication ei these indices have evolved from expe~encewith power systems, m i n ~ i t i v er ~ a s o n i n ~ and from heuristics. However, with the advent of power electronics and 0th tronic devices, there are prob~ema~jc cases in the general app~icationof indices. For examp~e~ consid e use of the power factor index to minimi sys~emlosses, with a ti load, such as a pulsating load on the s phase induction motor, become a source over part of the cyc stroke occurs fo~lowedby a re~e~erative period). Thus, the power ~ n d ~ c t i omn a c ~ n eload may go ~ e a ~ and n g lagging. In this case, th correct the power factor to minimise loss in the distribution supply intuitive result. The power factor index should be applied with caution in cases o f time v ~ a t ~ ounbalance n, and presence of on-powe~ frequency signals, The main ~ o t i v a ~ i ofor n using indices is the ease in calculation, th tion of of the definition, the simple a p p ~ ~ c a t oi of ~the indices (and the simplified in some cases, indices should not be used at all. Instead, it ~ ~ gbeh t ~ ~ c e sHowever, ). the time waveshape of voltages and currents directly. ~ o m e ng definitions include sojourn time, wavelet spectrum, Liapunov The in dust^ needs to est~blish~ ~ i andf complete o ~ P that d~~ can be compared (over location, over time, etc.) and such as IEC 61000-~-7,which cov 77A Work~ng~ r o 09 u has ~ made en ds U ~ ~ l ~ ~ - o PQ ~ esn~ ~~ ed da r are of s ~ n d a r d scan be used to set a CO

and they should create a ~ i ~ a mc c ue p~level ~ ~ of ~ PQ. ~ The ean s ~ a ~ ~d ~~e da d y contains some we11"~e~ned margins for harmonic disto~ionancl other variation^ le levels for events li WO&,~ ~ w e ~still e rneeds , to be done to set acc and ~ ~ ~ ~ ~ pThe t i voltage o n ~ characteristic . themselves are not e q ~ ~ p ~i e n t ~re ~ r ~ ~ eand m i~s s i b ~umber ~ n ta~smaxi i of quip needs to be decided on. More work i s needed on PQ standards that can be used by equip~entm a n ~ ~ a c ~ r eItr s . is far less expensive to inform m ~ u f a c ~ r e~r s ~ theoreal~levelt of t to improve the level of power quality. Some in sky, have already developed their own s ~ ~ n ~ ility s t ~ will d u~ ~ ~~ i ~ t ~e li yn i ~alli $P ~ issues, i n c l u d ~t ~~o s e

us effort is needed from $ ~ ~ a r d " § e ~ i ~ lish require~entsfor equ

Power System R e s ~ c ~ r i and n g Dere~lation

35

[ 121 ‘Voltage dips, short intemptions and voltage variations immunity tests’, IEC Standard ~ocument61000-4- 11. [ 131 ‘Basnivo fdr elkvalitet’, (Basic level for power quality, in Swedish), Gdtborg Energi Ndt AB,

[ 151

[l8]

[19]

[20]

221)

~ o t ~ e n b u rSweden, g, 1997. port on distribution and transmission system perfo~ance’,pub~ishedannually by Office of Electricity Regulation, Birmingham, UK. IEEE Project Group 1159.2: Power quality event characterization. llen, J. Svensson and L.D. Zhang, ‘Testing of ~d-connectedpower”e~ec~onics European Power Electronics for the effects of short circuits in the Confer~nce, Lausatme, Switzerland, 1999. . Bollen, ‘A method for characterizing unbalanced voltage dips (sags) onents’, IEEE Power Engineering Letters, July 1998. L.D. Zhang and M.H.J. Bollen, ‘Characteristics of voltage dips (sags) in power systems’, I n t e ~ a ~ ~ Conference onu~ on Harmonics and Quality of Power, Athens, Greece, October 1998. M.H.J. Bollen, ‘Characterization of voltage sags experienced by three-phase adjustable-speed drives’, IEEE Transactions on Power Delivery, V01.12, No.4, 1997, pp.1666-1671. L.D. Zhang and M.H.J. Bollen, ‘A method for characterisation of three-phase unbalanced dips EneW (sags) from recorded voltage waveshapes’, International Telecon~municu~ions Conference ~ ~ T E ~ ECopenhagen, C), Denmark, June 1999. ITIC ( ~ ~ ~ o r Technology m a ~ o ~ Industry Council, formerly known as the Co~puter& siness ss Equipment. Manufacturer’s Association), ITIC Curve Application Note, available at

acharjee, ‘Applicat~onof wavelets to d e ~ e ~ i n e motor drive performance during power system switching transients’, Power ~ ~ a l i t y 1, A ~ ~ e r1994. d ~ , An Introduction to Wavelets,Academic Press, 1992,6-18. L231 [24] 1. Baubechics, ‘Orthonormal bases of compactly supported wavelets’, ~ ~ m r n u n ~ c a ~ini oPure ns and A ~ ~ ~ ~ e d ~ u ~ h @ Vo1.41, m a t i1988, c s , pp.909-996. [as] S . Santoso, E.J. Bowers, W.M. Grady and P. H o f ~ ‘Power ~ , quality assessment via wavelet t ~ a ~mlysis’ ~ f o IEEE ~ Transactionson Power elivery, Vol.11, No.2, 1996, ~ p . 9 2 4 - ~ ~ 0 . egnevi~sky,‘Automated disturbance recognition in power systems’, Power ~ n g ~ ~ e e Conference r~ng (AUPEC 98), Hobart, 1998, pp.593-

[221

597.

1271 P.F. Ribeiro and P. Ceiio, ‘Advanced techniques for voltage quality analysis; ~ ~ ~ ~ e s sophistication or indispensable tools’, Paper A-206, Power Quality assess men^, ~ ~ t e r d a i n , 1994. [28] L. Zadeh, ‘Fuzzy sets’, Info~at~on and Control,Vol.8, No.3, 1965, pp.338-354. Introduction to Fuzzy Logicfor ~ r a c t i c App~icati~ns, a~ Sp~nger,1997. a, D. Bradky and P.S. Bodger, Power System ~ u r m o ~ ~ iJohn c s , Wiley & Sons,

ads and J. Arrillaga, ‘HVDC converter ~ a n s f o ~core e r saturation instability: A frequency domain analysis’, IEE Proceedings - Generation, T r a n s ~ i s s ~Distribution, o~* V01.143, Ea0.1, 1996, pp.75-81. . Ainsworth, ‘The p~ase-~ockedoscillator. A new control system for controlled static converters’, IEEE Transactions on Power Apparatus and Systerrts, Vol.PAS-87, 1968, pp.859865.

Power Quality

1

liaga, N.R. Watson, J.F. Eggleston and C.D. Callaghan, ‘Comparison of steady state and dynamic models for the calculation of a.c./d.c. system harmonics’, IEE Proceed~ngs,Vol. I 34C, No.1, 1987, pp.31-37. R. ~ a c a m i n iand J.C. Oliveira, ‘~armonicsin multiple converter systems: a genera~~sed approach’, IEE Proceedings, V01.127, 1980, pp.96-106. 6. Carpinelli, et al., ‘Gener sed converter models for iterative harmonic analysis in power systems’, IEE Proceedings Generation, Transmission and Dis~ributi~n, Vol. 14 1, No.5, 1994, pp.445-45 1. C.D. Callaghan and J. Arrilla~a,‘A double iterative algorithm for the analysis of power and ~ a ~ flows o at ~ ac-dc c converter terminals’, 115%:Proceedings, Vol.136, No.6, 1989,

. Smith et al., ‘A Newton solution for the harmonic phasor analysis of ac-dc c o n ~ e ~ e r s ‘ , IEEE PES S ~ m ~Meeting er ‘95, SM 379-8. C.D. Callaghan and J. Arrillaga, ‘Convergence criteria for iterative harmonic analysis and its application to static converters’, ICHPS IV, Budapest, 1990, pp.38-43. G.T. Heydt, ‘Problematic power quality indices’, Panel Session on ~ ~ ~ t a n d a~~ dIEEE s, ~ Winter Power Meeting, Singapore, 2000. R. Ott (Chairman), IEC 77A Low Frequency Phenomena, Working Group 9, ‘Power qua~ity 1999. measuremen~s’,Draft in progress, ctric Power Distribution for Industrial Plants. IEEE 141:1986, Recommended Br IEEE 1159: 1995, lEEE R e c o ~ e n d e dPractice on Monitoring Electric Power IEC 61000-2-5: 1995, E l e c ~ o m a ~ eCompatibility ~ic (E~C), Part 2: E n v ~ o ~ e n$ection t, 5: Classifications o f ElectromagneticEnvironments. IEC 61000-2-1: 1990, Electroma~eticCompatibili~(EMC), Part 2: E n v i r o ~ e n t Section , 1: D e s c ~ p ~ i oofnthe E n v ~ o ~ e-nElectroma~etic t Environment for Low-~requen~y Con~ucted Disturbances and Signalling in Public Power Supply Systems. IEC 61000-2-2: 1990, E l e c ~ m a ~ e tCompa~ibifity ic (EMC), Part 2: E n v i r o ~ e n t Sect~on , 2: Compa~~bility Levels for ~ow-~requency Conducted Disturbances md S i ~ a l I i n gin Public Power Supply Systems. AC IEEE c62.41: 1991, IEEE R e c o ~ e n d e dPractice on Surge Voltages in Low-Vo~~age Power Circuits. IEG 816: 1984, Guide on Methods of Measurement of Short Duration Transients on Low to Measurements of Voltage Dips and Short ~ n t e ~ p ~ i o n s Occurring in Industrial Installations. Federal ~ n f o ~ a Processing ~ ~ o n Standards Publication 94: Guideiin~on E ~ e c ~ cPower al for ADP ~n~tallation~, National Technical Information Service, 1983. D.L. Brooh, R.C. Dngan, M. Waclawiak and S. Sundaram, ‘Indices for assessing utility system R.M.S. varialion ~ e r f o ~ a n cIEEE ~ ’ , T r a n s a c ~ j oPower ~ Delivery, PE-

IEEE 519: 1992, IEEE Recommended Practices and Requirements for Harmonic Control in Electric Power Systems (ANSI). IEC 61000-4-7, 1991, E ~ e c t r o m a ~ e ~ Compatibility ic (EMC), Part 4: Limits, Section 7: General pi& on harmonics and inter-harmonics measurements and ~ ~ ~ ~ ~ u m efor n power ~a~~on, supply systems and equipment c o ~ e ~ t thereto. ed

estructtaring and ~ e r e ~ l a ~ ~ Q n irectives conce~ingthe Protection of T ~ l e c o ~ m ~ i c aLines t i o ~against

Group, ~ n t e r - h a ~ Q n in i c Power s Systems, January 1997. [571 Ec 868: 1986,F ~ ~ ~ k e-Functional ~ e t ~ T and design spe~i~cations. [58] IEC 868-0: 1992, Flickermeter - Evaluation of flicker severity, 4: Limits, Section 15:

ems and Equip~ent-

Relevant Standards. E631 lEEE 100:1992, IEEE Standard Dictionary ofEIecCrica1 and Electronics Te 1641 ET4 50160 1994, Voltage ~ h ~ a c t e ~ s tofi cEs l e c ~ i csupplied i~ by Public

1651 IBC 61000-3-2: 1994, E for ~ ~ o ncurrent i c em [66] IEC ~ 1 0 0 0 - 3 1994, ~: E armonic current emission 6~000-~-3: 1994, Elect ~ ~ i ~ tofi Voltage o n ~ l u c ~ t i oand n s Flicker in LQw-volta~e~~~~1~ ~ y $ t for e ~~ ~ u i ~ ~ Rated Current 1 16 A. C), Part 3: Limits, SectiQn 5: ~ ~ ~ iof ~Voltage a ~ Fluc~a~ions i o ~ and Flicker in ~Qw~VoltagG $upp~y ystem for ~ q u i p m e ~ i ~ d Current gncata than 16

G

analysis in real-time’,

Iowa State Univer§~ty USA

University of'~ e § t Australia e ~ A~§t~lia

r Loi Lei Lai City ~ n ~ v e ~ § ~i t oy .n d ~ n

UK

3

Power System ~

e

~ and D~ ~ r ~c~ l a t~i o n ~

S o ~ a r agents e have evolved from multi-agent ~ y s t e ~ s three broad areas which fall under distributed artificial being dis~ibutedproblem solving (DPS) and parallel arti ~ e ~ascwith ~ ~ulti-agent , systems, they i ~ e r i many t potential benefits. For example, s o h a r e agents inherit m o d u ~ a r i speed ~ , (due to parallelism) and reliability (due to redundancy). It also i ~ ~ r i t s those due to AI such as operation at the ~ o w ~ e d level, g e easier mainte~ance,r e u s a b i ~ i ~ and p l a ~ f independence. o~~ The concept of an agent can be traced back to the early days of research into DAI in the 1970s. tudy of mu~t~ple collaborat~v~ agents includes intera~tionand c o ~ ~ n i c a t i o n be agents9 d~compositionand dis~butionof ta coordina~ionand cooperation, conflict resolution via negotiation. These resulted in work such as I planning and game ~ht:Qr~es [17]. ‘ s m a ~ e s s ’derives from the fact that the ‘value’ gained from ~nd~vidual stan agents c o o r ~ ~ n a ~their ~ n gactions by working in coo~erutionis greater than that gained from any individual agent. A p p ~ i ~ a ~ domains ion [ 1x1 in which agent solutions are being applied to or investigated include work~owmanagement, network management, a i r - ~ a f ~control, c business process r e - e n g i n e e ~ ~ i n f o ~ a t i o n re~eval/management, electronic commerce, educat~on, perso ~ s i s t ~ n t sas), e-mail, digital l ~ b r ~ i e sc ,o r n r n ~and ~ ~ontrol, ~ m a ~ s~heduIing/dia~ m a n a ~ e ~ e netc. t, s are still ~ ~ m o n s ~ ronly: ~t~rs It is important to note that most agent-based e even greater cliallenges, some c Q n ~ e ~ ithem n g into real usable appiications would reseen. The essential ~ e s s ~ g e of wh~chhave been ant~cipatedbut, currently, many ity9their wide of this section is that agents are here to stay, not least because of thei r a n ~ eof a ~ ~ ~ i c a and b i lthe i ~broad spectrum of companies investing

a component of software andfor hardware which i s capable of accomplish tasks on behalf of its user. here art: several

first^^, agents may be classified by their mobility, i.e. by their ability to move a ~ ~ ~ n some n e ~ o r kThis . yields the classes of static or bile a~ents. ~ e c o n ~they ~ y , may be classed m either ~ e ~ i b e r a or ~ ~reactive. ve derive from the deliberative thinking p~adigm:that is, the agents symbolic, reasoning model and they engage in p l a ~ i n gand neg~tiationin order to achie~e with other agents. Work on reactive agents o r i ~ i n ~ t from e s research carrie oks [19]. These agents on the contrary do not odds of their environment, and they act using a stimulus state of the env~onmentin whi that intelligent behaviour can olic re~resenta~~ons of traditional AI [21].

Information Technolonv A ~ ~ ~ i c a ~ i o n

355

Thirdly, agents may be classified along several ideals and pfimary attributes that agents should exhibit. At T Labs, three main attributes, namely autonomy, le cooperat~oi~, have been ~ d e n t i ~ e~d . ~ ~ refers o to~the oprincip~e ~ ythat agents can operate on their own without the need for human guidance, even though this would sQmetime§be invaluable. Hence agents have individual internal states and goals, and they act in such a manner as to meet their goals on behalf of their user. A key element of their autonomy is activeness, i.e. their ability to ‘take the initiative’ rather than acting s to their environme~~ [22], Cooperation with other agents is paramoun~. to cooperate, agents need to possess a social ability, i.e. the ability to interact agents and possibly humans via some communication language [22]. Having said this it is possible for agents to coordinate their actions without cooperation [23]. Lastly, fo systems to be truly ‘smart’, they would have to learn as they react and/or interact w external environment. Agents are (or should be) disembodied bits of ‘intelligenc these three minimal es, Figure I 1.1 was used to derive four types o f agents, namely c~llaborativeagents, collabo~ativeleaming agents, interface agents and smart ~ ~ e ~ t ~ smart

\

agent^

/

Collabo~at~ve

Agents

1.1 A part view of an agent typology

It must be emphasised that these distinctions are not definitive. For ~ x ~ p l with e, ative agents, there is more emphasis on cooperation and au~onomy ; hence, it is not i ~ p ~ i ethat d collaborative agents never learn. Like ere is more emphasis on autonomy and learning than o Ise which lies Q u ~ s ~the d e ‘~ntersecti~g areas’ is not con most expert syste~nsare largely ‘autonomous’ but,

may s ~ ~ e t i mbe e sc l ~ s i ~ by e dtheir roles ~ r e f e r a b ~ y , odd Wide Web ( ~i n ~ Q ) ~ a t i agents. on Again, info

~ i ~ h ltwo y , or more age^^ ~hiloso~hies are combined in a ~ y ~agent. r i ~There are Qther a ~ b u of~ agents, ~ s which we already m e n t i ~ For ~ ~example, ~. is an agen~versatile (i.e. does it h in a variety of tas Is an agent benevolent or non-help Does an agent lie wingly or is it always ~ u t h f u(&is ~ Can you trust the agent enough to (risk) delegate task in contrast to failing ast tic ally at the b o ~ ~ ~Pee s ?

Power System ~ e s ~ c t u and ~ n Dere g

~esearc~ers are also a ~ b u t ~ en~go ~ i o nattitudes al to agents - do ~~~yget ‘fed up’ to do the same thing time and time again? role does e ~ o t i o nhave in c o n s ~ ~ c t believab~e in~ agents [NI? In essence, agen in a truly ~ u l t i - d i ~ e ~ ~ i ~ n a space. It is quite possible that agents may be in c o ~ p ~ t i t i owith n one another, or per stic towards each other. In agen~-ba involves high-level messages. The use of lower ~ ~ r n ~ ~ costs, c a easy ~ ~reimplementabi o n d ~ o n c u ~ e nLastly, c ~ . and p most i ~ p o ~ tagent-based ~ y , applicat~onsop ally at the ~ o w l e ~ level ge [

.l, collabo~tiveagents emphasis^ a u t o n o ~ yand s) in order to p e r f o tasks ~ for their o ~ eThey ~ may . le rnphasis of their operation. In order to have a CO ey may have to ne~otiatein order to reach a ~ e e ~ e i ion t ssome matters The ~ o t i v a for ~ ~ having ~ n c o ~ ~ a b o r a ~agent ~ v e systems may include one or several of the ~ o l l o w i n ~ ~ to solve probI~msthat are too large for a cen~al~sed single agent to do owin resource li~i~ations; to allow for the ~nterconnectingand interoperation of so~utionsto i~herent~y dis~buted e solu~onswhich draw from

speed (due to p ~ a ~ ~ e ~ ~ s r n ~ h a r e a bof i ~resources); ~~ to re~earchinto other issues, e.g. understanding ~nteractions asise autonomy and l e a r n ~ ~ing order to p e ~ o subtle emphasis and distinction between c o ~ ~ a b o r a ~ i ~ a t ~ v ~~ o ~ ~ a b o r awith ti~g c o ~ ~ a b o r awith t ~ n ~other agents, as is the case with c o ~ ~ a ~ o r~gents. ay not requir~an explicit agent c o ~ m ~ i c a t i ol an n ~ a g eas one re~uiredwhen ith other agents. Essentially, interface agents support and provide assis~nce system. The user’s agent lar app~icationsuch as 1 erface, learns new ‘shortitors the actions taken acts as an assistant, better ways of doing the task. Thus, the user’s the task. As for erates with the user in ac~o~plishing Ily to assist their user better in the ~ ~ l ~ o wfour i n gways [26]:

o b s e ~ ~ and n g imitating the user (i.e. l e ~ n i n g h r e c e i v ~ nposit~ve ~ and ~egativefeedb ctions from the ~ s e r

Information Technology A~~lication

7

other agents for advice (i.e. learni~gfrom peers). e, Their c o o ~ e r ~ t i owith n other agents, if any, is limited ically to asking for a d ~ ~ can ion deals with em, as is the case with c o ~ ~ a b o r a agen~s. ~ v e The ically by m e m o ~ - b a s elearning ~ or other t e c ~ i ~ u esuch s as h are being in~oduced. An interface age^^ is a ~ u a s i - s m ~ where boring and laborious tasks could ith one or or^ computer appbica e tedium of ~ u m a n ps e r f o operatio~s. ~ ~ ~ ~

from a flight reservation to ~

~

aa t e~l e ~io ~ m n u n~ c a t ~ o n s

s neither a necessary nor , ~ u ~ cc ia ~ e d~~ ~t ifor on a ~ ~ ~ t ~ a

wn to other agents.

~ a ~[28]e lists r the ~ a j co~ a~l l e ~ g eThey s . ~ ~ c l u at d eleast the ~ o l ~ o w i ~ ~ : o

move?

~

thow ~ does ~ an ~ ~ : ~move e .From ~ place t to place? How does it

3

Power System Restructuring and ~ e ~ e ~ I a t i o n

Au~hentication:how does the user ensure the agent is who it says it is, and that it is represent~ngwho it claims to be represent in^? does the user know it has navi various networks without being infected by a v Secrecy: how does the user ensure that the agents maintain privacy? How does the user ensure someone else has not read the personal agent and executed it for their own gains? How does the user ensure that the agent is not killed? Security: how does the user protect against viruses? How does the user prevent an i~comingagent from entering an endless loop and c o n s ~ n all g the CPU cycles? Cash: how will the agent pay for services? How does the user ensure that it does not run up an outrageous bill on the user’s behalf? rmance issues: what would be the effect of having hundreds, thou$ands or millions f such agents on a WAN? Inte~operabi~i~/com~unication~rokeri~~g services: how does the user provide e ~ n g / d i r e c t o r y ” ~services e for locating engines andor s p e c i ~services? ~ How the user publish QT subscribe to services, or support broadcasti~gnecessary for some other coordina~~on approaches? The ~ ~ for developing ~ iin f o ~ a t ivo ~ i n t e r~nagents et is ~ simply ~a n e e ~~ d e for ~~ ~ d tools to manage such information explosion. Everyone on the WWW would benefit from ~~e~ agents are going to search the Intern~t,becaus matter how ~ u c h ernet may be organised, it cannot keep pace with the the has (or prom~ses)its own stre~gthsand de~cienc~es, e strengths and minimise the de~cienciesof the most rele rpose. Frequently, one way of doing this is to ad together some of the s~engthsof both the de hybrid agents refer to those whose cons~~tut~on i or more agent p ~ i Z o s o within ~ ~ ~ ~a ssingular agent. These philoso p h ~ ~ o s o ~an h y~,t e r f a c eagent philosophy~collaborative agent s i s h y b agents ~ ~ or ~ c ~ ~ t e c ~ The key ~ y p ~ ~ ~fore having applica~~ons, the benefits accrued from having the combination of ~ h ~ l o s o within ~ ~ ~ ae s roved right; the ideal benefits hies, In such a case the reactive component, which would take precede~ceover the del~bera~ive one, brings about the following ~ e n e ~ t s ~ robustness, faster response times and adaptabil~ty,The deliberative part of the agent would term goa~-orientedissues. For ~ x ~ p lthere e , is agent by comb in in^ the interface agent and MO mobi~eagents to harness fea other co~b~nations. fer to an ~ n ~ e g setup r a ~of~ at~ least two or more etero~eneou~ agent sy which belong to two or more different agent classes.

I n ~ o ~ a ~Technology ion Application

35

also contain one or more hybrid agents. ~eneserethand Ketchpel 1291 a~iculatec ~ e the ~ ~ y ~otivat~ori for heterogeneous agent systems. The essential argument is that the wosl abounds with a rich divessity of s o ~ ~ a~rreo ~ u cproviding ts a wide range of services for a s i ~ ~ l a wide r ~ y range of d o ~ a i n s .hough these psograms work in ~sola~ion, there i s an te in such a manner that they increasing demand to have them i n t e ~ o p e ~ ~ hopefully, provide ‘added value’ as an ensemble than they do individually. A new domain called a ~ ~ ~ t -software b a ~ ee ~ n g i n e ~ ~has i ~ gbeen invented in order to facilitate t%ieinteroperation of misce~~aneoussoftware agents. A key r e ~ u ~ e m e nfor t interope~ation nts is having an agent ~ o ~ m ~ i c a t language ion (ACL) via which the ‘agents’ can comm~icatewith each other. The potential ~ e for ~ having heterogeneous agent technology are several: ~

S~andaloneapplications can be made to provide ‘value added’ services by enhancin~ in cooperative hetesogeneo~ss them in order to pa~icipateand intero orated because it could obv The legacy software problem may b ‘new leases of life’ by costly s o ~ a r rewrites e as agents ~ntesopesatewith other systems. At the very least, heterogeneous agent t e c ~ o l o g ymay lessen the effect of routine s o h a r e maintenance, upgrade OS rewrites. Agent-based software engine~singprovides a radical new approach to so i~plementationand mainte~ancein gener~l,and software i ~ ~ s o ~in ep a~~ cb~ li a ~ s. ~ ~ ~ e n e s e r ~ tand h Ketchpei [29] e that agent-based sofeware engi ~ o ~ p a r etod object-oriented pro ing in that an agent, like an Q ~es§age-ba§edinterface to its int a structures and algorithms. H ey distinction: in object-oriented p r ~ g r a ~ i may differ from object to object (this is the ~rincipleof po s o ~ a r enginee~ng, e agents use a common language with They h~ghligh~ three ~ m p o ~ questions nt raised by the new agen~-osi~nted so~wae e n ~ ~ n e e n n~ ga r a d i gThey ~ . inciude: priate agent co~mun~cation language? apable of c o ~ u n i c a t ~ in n gthis c o n ~ ~ c t 1e d at commun~cationa s c h ~ t e ~ ~are r econducive s to cooperation.

are availabl~,there a e two ssible ~ c ~ ~ e cto~choose r e s ents handle their own coord ation or another in which g can rely on special s y s ~ programs e~ to achieve coordina~on.The d i s a ~ v ~ ~ of a gthe e former is that the c~mmun~cation overhead does not n e c e s s a ~requireme~tfor the re of a g ~ n ~As s . a consequence, the

various services. They also establish the connect~onacross the e ~ v ~ ~ oand ~ ensure e ~ t s c ~ ~ e‘co~versatio~’ ct amongst agents~

Power System R e s ~ c ~ rand in~

3

Agent

I

Agent

I 1-2 A federated system (adapted from 1293)

General Issues and the Future of Agerzts from t e c ~ i c issues, a~ as mentioned earlier, there is also a ~robl~m which s ~ are looming. They include the following: rivacy: how does the user ensure that a g ~ n ~ sa ~ n much ~ i nn e e ~ e ~ acting on the user’s behalt? Legal issues: ~maginean agent offers some bad advice to other peer a ~ e n t rs e s ~ ~ in ~~ng ~ ~ a ~ i l i to t i eother s people; who is r~sponsible? Ethical issues: agents must limit their searches to appropriate servers, share info with o ~ e r and s respect the authority placed on them by server o~erators.

suppliers, electric generators and distributors will have to

adds to our experien~eand helps us make the next market imp~ementat~on work a little better and more competit~vely.It is believed that to some degree, ~rn~len~~ntation, ~ ~ g i o commodity na~ exchanges will play a key r ele~trici~. This section assumes a ~ ~ e w o which r k has been described in 13 1,321. ~ o m p ~ ipresently ~es having both generat~onand d ~ s ~ i b u t ~facilities on would be d i v ~ ~ einto d separa~eprofit and loss c ~ ~ ~Power e s .is generated by generation co~panies ia tra~smission companies i s sold to energy service com city ~ e l i a b ~Council li~ (N antile associations ( E M S ) will e m e r ~ ein r e l i a ~ i l and i~ s e c u ~ ~ ill promote liquidity and as an i n t e ~ e ~ i a t e this co~petitiveelectri er to all mul~ilatera~ trades, they will provide assurance to traders, that they nee onry about trading because a defaulting contract partner. This framework allows for cash (consists of spot and forward markets), fbtures and p l a m~rkets. ~ ~ See ~ ~~i g u r e11.3, The spot market allows for trading power e other d ~ ~ a t i o e.g. n , 30 minutes) in the next 30 days. Forward contracts c i ~ as specified in the contract from 1 to 18 traders to buy or sell firm e l e ~ ~contracts months. The fittures ~ a r k allows e ~ trade^ to purchase a non-^^ electric^ given ~ o n t hin the future (e.g. 1 to 18 months). Futures contracts pro electricity traders to manage their risk. The planning market is a 10 develop capi~alfor building large items like new plants and trans time horizon (in months)

1.3 Interconnection between the markets

s (for both ~ t u r e and s physicals) for electric energy are ex comm~nand will be an ~ m p omeans ~ ~ tof mitigating risk, An ho~derthe r i ~ h to t buy or sell w i ~ h o the ~ t o~ligationto buy or sell. holder must pay an u ~ - ~ o nremium. t The a ~ o u not f the premium shou potential holders. The worth of an option may vary references, makeup f p o ~ o l i o s(collection of assets is, how does one etemine the value of d in many markets to value options. Its usage assumes many a1 commodity that may not be true about electricity.

362

Power System Restructuring and ~ ~ r e ~ ~ a

The approach taken in this research is to allow CO sed agents to d~veloptheir own valuation f o ~ u ~ as a they e p ~ i c i p a ~inea s i ~ u l a option ~ e ~ markets. er options valuation should achieve higher profit than do s. The computerised agents evolve in a genetic algo valuations are replaced with new agents that are based 011 the succe§s~lideas of the better agents.

As mentioned previously, it is quite likely that regional c o ~ m o d exchanges i~ in which buyers and sellers pa~icipatein a double auction will soon exist. Such e x c h ~ g e sare utilised in other markets and are essentially an extension of the electricity market operating in California, A centralised exchange allows many and varied traders easily to trade a c o ~ o cn o ~ m and o derivatives ~ ~ based on that CO In the cash (spot and forward) market, who indepen~entcontract a ~ i n i s ~ a t(TCA), or . CENCOs and ESCOs cooperate with the IC that the energy transactions resulting from the matched bids do not overload or ren e l ~ ~ ~transmission ica~ system insecure. The ICA monitors and res~ondsto the s y ~ limits ~ e and ~ transmission capacities. The spot market is what we are most familiar with in the electrical and a buyer agree (either bilaterally or through an exchange) upon a ~ u ~ bofe ~r e g a w a t to ~ sbe delivered sometime in the near ture (e.g. 10 MW from 1.00 p.m. to 4.00 p.m. tomorrow). An aptions contract is a form of ins~rancethat gives the option not the obliga~ion,to (sell) a contract at a given price. For e is someone ‘writing’ the contract who, in return for a premium, is ce. See Figure 11.4. Both the options and the d e s i ~ e to d minimise risk. A ~ t h o ~ provisions gh for ~ e ~ i exist, v e they ~ are not (e.g. the delivery point is not located where you want it to be located). The trader ~lt~mately cancels hisher position in the futures market with either a sicals are then p ~ c h a s e don the spot market to meet d e ~ ~ wi ~nd een locked-in via the hhlres contract. ‘Long’ ~enotesownersh~~; to go long figure, long indicates that the trader has pu (call) or the right to sell but) the hture. A trader who write

Information Technology Ap~~cation

3

figure shows how the ‘put’ works. The long trader pays a premium to lock-in a maxi^^^ . The short trader price (exercise price) that he/§he will have prem~~im in return for promising to sell the

1

9

Price

Terminal Price

Using put and call options

value of the option has been the subject of some a e ~ a tA ~ .c and SchQie§put together their formula which has been other commodity markets. Marshal1 [33] states tha r e ~ ~ ithat: re~ ~ - t rate e is ~ The ~ ~ Q interest and constant. The una~rlyingasset pays no as. The u n d ~ ~ ~ yasset i n g is e ~ ~ c i ~priced. nt~y The option is of the European type. ~ ~ c ~costs i Q (for n bwying and selling). of u ~ d e r l y asset ~ g value can be borrowed. restrictions on, or penalties for short selling. holes e ~ u a t ~ for o n valuin ut option is as follows [34]: p = [x.ePip(-r. (T-PI). N(- d2)- s * N(- d l )

where: 2’ = strike price

r = risk free rate (= cumula~ive ~ ~ normal ) ais~bu~ion T = ~ x p i r a ~ date on t = c u ~ e ntime t

~

Power System Restructuring and

344

11.3.2

A g e n ~ - ~ a ~s e ~~ r n p ~ ~ ~ ~ ~ ~ n a l

arket p a ~ i c i ~( ~s u t s~ ~ ~ ~ e r s comp~ex,c h a n g i with ~ ~ time modify their behaviour as time goes along, ~ o s ~ ~Ai~o~ h~o.u gsome h res m a r ~ e res~onses t using control theory, it is g

t~

5

with~ usin~

is relatively smooth s are another search

s

~ n f o ~ t i Technology on Applica~on

iscrete po~ntsin the search ace and selects those gro solve the ~ r o b l e ~ . The basic genetic a1 nithm, as d e s c r i ~ ~byd GoIdberg [35], can be Written as follows a population and set the generation counter to zero. 2. Until done or out of time, do the following: fitness o f each m e m ~ of e ~the popula~ion.

c r e ~ e nthe t generation cou~terand go to step 2.

. Genetic a l g o r i to~ evolve a population of trading agents ~ p t i o n with s Agents

A simple electricity market with four generators that provide is modelled. ~ e n e r a ~ are o ~ d~spatched s to meet demand and a from the aggregate ~ ~ g i cn a curves ~ o f the dispatc~edgene~tors and $34) are offered with valuatio prices of $15, $20, les and the ~ a r ~ e t data. ~A-basedagents then buy and sell the o ~ ~ ~ato n s

Power System ~

3~6

e

5 and ~

~

~

n

lack-Scholes prices. Implicit in the ge~ierationof buy and sell signals is a valua~on of the put options by each of the agents. Hourly demand data for an extended period was prov~dedby a large lity and was used as a source of realistic load data in this s i ~ u ~ a t ~See on. arker price data: Before evolving strategies for ata was nee~edwith which the put option prices were calcul hourly demand data was used in conjunction with the gen~ratorp e the ~ ~ kprice e t in an iterative procedure re~iini§centof witthe suppliers has a unit that is ~ o d e l l e dwith a q u a ~ ~ ~ t i ( Cost = a + bP + cP2). See Table 1 1.1 for the values of the CO cients. The supplier §upplier’s m ~ n i m ~ rod~cespower as long as the market price does not fall below itial cost (which is determined by their minim^ product~onlevel). 1800

1700

1600

1500

1400

3350

0

20

40

60 Demand

Bo

100

140

120

ernand on vertical axis ( M W ) versus time (hours) Generator parameters

LU,

hmex

7.0

12.0

8.0

12.6

Generator 1

a

b

c

pm

100

0.005

100

2

150 200 250

6 7 9 8

0.004

120

pm, 600 700

0.006

150

750

10.8

18.0

0.007

200

800

10.8

19.2

3

4

jl___

The marginal cost is found by taking the derivative of the cost curve ( A = b + 2cp). The m a ~ ~ i ncost a $ curves for each generator are shown in ~ ~ g u 11 r e‘7. Note that each genera to^ has both a minimum and a m ~ i m opera u ~ tin^ level. ( § t a ~ pand s h u ~ d o ~ costs, ramp rates, and minimum up and down time constraints were not G this ~ i ~ ~ ~ a tIfi the ~ n market .) price is below the mini mu^ ~ ~ ~ i n a l

Information Technology Application

7

generator, that generator is removed from consideration and the market price recalculated. This process is repeated until demand is balanced by a set of genera~ors b produce ~e at the discovered price. If price d ~ 5 c o does v ~ not for whom it is p r o ~ ~ to occur after 20 iterations, the market price from the 20th iteration is taken as the ~~~~t price. (Under this simple scheduling scheme, it is possible that a unit could be forced CO produce below its minimum marginal cost but a check showed that this never A brief c~ari~cation at this point may be in order to prevent confusion in the use of the term ‘spot’. The market price is referred to here as the ~ ~ ewith p the i terminology ~ ~ used in finance (i.e. options prices prices); this is not to imply that the hourly market here is the s market (i.e. the ‘spot’ e l e c ~ cmarket i~ as the real-time ele price data for a typical week is shown in Figure 11.8. 3 . Standard deviation of spot price: The standard deviation (s oles formula. For a given hour, sigma is ca~cula~ed used when calculating the B1 period hours prices. The s ~ n d a r ddev~ationof the using a window of the last market price is shown in Fi ut options price data: There are four put options, which c m be bought and sold, aving strike prices of $15, $20, $25 and $30. The market valuation (price) of each o f these is calculated using the lack-Scholes formula for put options, as pres~ntedearlier. Note that the risk-free rate is taken to be constant t ~ o u g h o u the t simulat~onand that T-t is a constant 90 days. This was done to prevent having to ‘roll over’ the options position because the expiration date was reached. ~aluationsfor the put options are shown in Figure 1 1.10. ne can see that they go up and down with swings in the underlying spot price of electricity and that the put options with higher strike prices have higher market valuations, as would be expected. PEP

Each agent in the population buys andor sells the four put options. These agents act according to i n t e ~ a l ~gene~ated y buy and sell signals. These signals are ~eneratedu s ~ ag GA to vary the coe~cientsin a mQdified ~ ~ a c k - ~ c h ocalculation. les ~ p t i o n §could be traded only for peak periods on weekdays, i.e. Monday-Friday, 1l.OOa.~.-4.0O~.m. GA val~ationof options and buyhell signals: The GA is as a string of real number genes. The number of genes is determined by the c on being p e r f o ~ e d by the GA (described next). For these simulations each GA has eight genes, each of which is a real n ~ b e r , The equation currently used by the GA to generate a buy or sell signal is a modified ~ ~ a c ~ - S c h ovaluation. les A signal to buy or sell an option will be generated if the GA valuation minus the market valuation is greater than some tbes dl and d2 in the lack-~choles formula are recalculated using a modifi a’, where CT‘ = (Gene2).CT and where LT is the ‘standard’ calculation deviation of the spot price. A buy signal is genera~ed if [Gene0 * X exp( --r * (2‘ - t ) ) N(- d1)- (Gene 1). S N(- d2)] + (Gene 2 ) is greater than the Market Price. Similarly, if a new d l and d2 are calculate gene^).^ and the Market Price is rea ate^ [(Gene4)-x . e x ~ - r . ( T - t ) ) . N ( - d l ) - ( G e ~ e 5 ) . $N(-d2)f+(Gem7) . then a sell is generated.

-

Power System Restructuring and ~ e r e ~ l a ~ o n

3

IQ

1.7 M a ~ ~ icosts ~ a l on vertical axes ( $ / ~ W vs. )

__ 29

0

20

40

60

1 Bo

100

Spot price on vertical axis ($/MWh) versus time (hours)

120

140

~ n f o ~ a t i oT ne c ~ n o l Application o~

0.

20

Standard deviation of spot price on vertical axis versus time

e and Deregulation ~

Power System ~

37 Valuation for Put 0 (strike=$l5)

Valuation for Put 1 (strike=$aO) 10

6 $

$

4 5 2

0 0

0

0

60 80 100 V ~ l u a t i ofar ~ Put 2 (strike=$25) 28

40

$

20

40

60

80

100

$

0

0

Hours

1.10 Market valuation for the put options ($ vs. hours)

rodu~tion: AAer fitness is calculated, the agen~sare sorted accord~ngto their ss. Reproduction is performed using single-point crossover of two parents selected from the best half of the population using r selection, One child is created and ~ e ~ l a can e sagent in the worst half of the popu Each child’s genes can be mutated in four different ways (bearing in mind that the genes are real-valued): 1, 2.

.

.

2% of the time the gene is r e ~ ~ a c erandomly. d 5% of the time the gene is multiplied or d 10% of the time the gene is multiplied or 1% of the time the sign of the gene is ch

an do^ genes are generated according to the relation: NewGene = ~ e n e ~+iRandom[O.. n 11.(Gen where GeneMin and GeneMm are the max and min values of that gene over the e ~ t i r e ~opulation. (This was tried as a reasonable way of ge~eratingnew genes without disc~dingwh has collectively learned about the ‘re~sonable’ran because the space for real numbers is infinite.) a coefficien~, This process is repeated until every agent in the worst half of the popul~tionhas been replaced. (A variation on this theme is to replace the worst on with randomly generated agents, in an effort to introduce ne 1 and prevent stagnation.)

Results: The GA was able to evolve a strategy that co~~istently made a profit buying ut options in this market. As shown in Fi

~

Infi3mation Technology Application

ent is positive and ~ m p r ~ v over e s the generations, ultimately reaching a value of per trade (with one trade allowed each hour). Figure 1I. 11 also shows the fitness of the worst agent and the average fitness for the whole population. One can See that at the start of the run most agents a c ~ a l l ylose money (make a negative profit) but by the end of the simulation the aver~gefitness has risen to nearly zero. Figure 11.12 shows the best genes from 4 different runs. 2 1.5 1

0.5

0

40

20

60

80

100

Maximum Fitness

-15

.

-20

. 3 Minimum Fitness

5r-----

' 1

1 0

20

40

60

a0

100

Average Fitness

igure 11.11 Maximum, mini mu^ and average fitness over a typical run. The vertical axis m ~ a ~ r e s profit per generation; the horizontal axis counts generations.

na m Solving the optimal power flow (0 associated with

wer

m is ~ n d a to~ the~ ~ n b u~n d~l i nof ~ n open access and is of increasing dereguiated environment of the electricity ension n o n - l i ~ ~optimisation r onal in solving the OPF prob which is difficult to solve. The ~ o ~ p u t a ~ ~difficulties its use in power system o p e ~ t ~ o n s .

gene0 -2.7386 gene1 :-I 4.23i gme2: 4 . ~ ~ 9 6

g~ne5:-6.441 1 gene& 2.6333 g e n ~ 7-6.97 ~ 17 Fitness = 1.7178

The best genes aner 100 generations from 4 ~ f f ~ rruns ~nt

0~~~~to ensure conv as a result, many local

app~iedto the IEEE 30-bus test system under different ~ e n e r a t o ~ resented.

11.4.1 The OPF problem seeks to optimise s t e ~ y - s ~power a t ~ system p e r f o ~ with ~ ~ reespec^ to an object~vefwhile subject to numerous constraints. For optimal act dispatch, the objective ~ n c t i o nJ, is that o f total g~nerationcost. ion o f ~ansmissionlosses and voltage level optimi a§

minf(x,

U)

(1 1.1)

subject to

of control variables (these include generator active tap s e ~ i n g sx) ~is the vector o f de~endentv a ~ a b l (1e ~ ) is the ob~~ctive to be optimi generator reactive po are the ~ ~ i e q u acloi n~ s ~ a ~ on n~s power ~ o n s ~ a i ~ t s ;

EP seeks the optimal solution by evolving a population of c a n ~ ~ ~saot ~e ~ ~ i over ons a n ~ m ~ of e rgene~at~ons or i~era~ion f o ~ ~ from e d an existi~gpo~ulation prod^^^^ a new solution by perturbing each component of an exis a ~ o u n t .The degree of optimali~of each of the c ~ d i d a t eso i~easuredby t h e i ~ ~ t n ~which s s , can be

ugh the use of a c o ~ p e t i t i oscheme, ~ the i~dividualsin each p h other. The winning i n d ~ v i ~ ~ a l s the next generation. For optim~sa~~o the more o p t ~ so~utions a~ have a er chance of surviv lation ~volvestowards the global optimal point. ive and the process is ~ e ~ i n a by ~ ea dstopp Rer a s ~ e c i ~ number ed o f iterations or no apprec~ablechange in the best so~utionfor a certain n u ~ b of e ~ge adopted in the present work. The main stages of the EP t ~ ~ ~ n~c l ui ~ ~~ int i~ua ~ ~ ~ a t ~ mutation and compe~~tion are shown in the ~ o w c h oa f~Figure. 11.13.

e s ~ c ~ and ~ nD~re~lation g

ased on the EP me~hodology, an a l g o r i t ~for solving the PF ~roblemcan be e s ~ b l i s ~ e dThe . basic f l o w c ~of~the algo~thmis shown in ~ i 11.13 ~ withr its ~ components described below and in Sections 1 1.4.2 and 11.4.3. s Q l ~ ~An o individual ~: in a pop~~ation re~resentsa candidate ts o f that solution consist of the co~trollableand uncontrol on at all v ~ ~ a b S~ ~e e~ c. ~ ~the c ac ~o l~~~ o ~ v~aa~ ba b~~eeare s ~ p e c i ~ epower d g e ~ ~ r a t (PV) o r nodes other than the slack node, the specified voltage m at all PV nodes and tap positions for variable tap t r ~ s f o ~ e rEach s . candidate solution also stores depe~dentvaria~lessuch as the most recent load flow solution for subsequent use in ~ i t ~ a l i s i nthe g load flow on the next iteratio~to reduce c o ~ p u ~ t i otime n within the loadflow algorithm. Zis~~iuon: Each o f the con~ollableva~ablesof an i n ~ v ~ isd i ~n iat ~ a ~rs e~~ n d o ~ using a uniform random number ~ ~ s ~ i b within u ~ o its n feasible . For example, for the s p e c ~ active ~ e ~ power generation for a PV node i, with acti Pmw, we have

(1 1.2)

1

where U{Pm,, P,, is a ~ i f r o~ ~d numbe~ o~ ~ b e ~ e e nP,," and P,. In additio~to this, one cand~datesoiution will have its specified active power generation for all PV nodes excluding the slack node set to the economic dispatc~solu~ionfor the system active power load as the aggregate active power load of all nodes plus 2% to appr ~ ~ s ~ ~ losses. § s ~This o neconomic dispatch solution is obtained using the

tioons: Each candidate s o ~ ~ t i is o nassigned a fitness to ~ e a s u r its e ect to the objective being optimised. In the case of active and reactive tness of individua~i will be,

M

(11.3)

f i =

J

VP, =

K , ( v ~- 1 . 0 ) ~if Vj > v,- or V, <' V otherwise

otherwise

~ n ~ o ~ aTechnology ~ i o n ~pplicat~oi~

I

initialise Population

I

Make a gradient step

Evaluate loadflow and assign fitness to altered candidates

~

~ 11.13 g Flowchart u ~of EP-OPF ~

m cost of g e n e r ~ t i oand ~ C, is the In tile above ~ ~ u a t ~ o nis, the ~ a x i ~ u ossible i. The term V?, denotes a penalty term on PQ or swi eneration cost of indiv~~ual node j for v ~ o l a preset ~ ~ g voltage limits Y,~’”, Y’”.

represents a penalty on

a reactive power limit. K, and Kq penalty weighting constants, It is not n e c e s s ~ yto impose a enalty on slack node active power ~io~ations as the at at ion stage helps to satisfy this constraint. The EP-OPF algorithm seeks the solution with the m a x i ~ u mfitness. ~ ~ uA new t ~ population ~ ~ of : OFF solutions is produced from the existing population through the mutation operator. A new indivi from each ~ndividua~ p i , where thejth OPF variable in the new ind calc~~~ated as

where x:> denot~sthe value of variable j in pIr. x, is the value of variablej in the parent

~ ( O , ~ is ~ ,a )Gaussian random n u ~ ~ ~ with ber a

me^

of zero and a s

deviation of oJ,The e x ~ r e s s i ~esigned o~ for c,,i s I

(1 I .5) w ~ e r eJi is the ~ ~ eofsi ~s d i v ~ d ui; af~,,, is the m ~ i m u mfitness wit hi^ the po~ulation; xY,x;ltn denote the er and lower limits of variable j ; a is a ~ o s i ~ ~ v ~ tly less than unity; and r is the iteration counter. The term a' ation of~setthe rate of which depends on the value of a (1 1.5) that a solution that has a much lower fitness than th value fora,,; hence it will be moved further by r n u ~ ~ i oton a loc ~~~~~~~: To help in the satisfaction o f the slack node active c o ~ § ~ a i ~allt sunits , other than the slack are assigned R loadi their dispatc~esi s then compared with the total generati of that indiv~dua~. If the difference between them is with the slack unit, then the candi~ateis ~ c e p ~ eIfd .not, the process five a~empts.If with~nthese a~emptsa feasible assignment is n c o n s ~ a i n ~tod force satisfaction by sharing the ~xcessive r ~ ~ a i generators n ~ g as follows. ing the slack node active power in an i n d i v ~ d has ~l e slack unit is unit 1, the total available capacity o f uni

-N i=2

(1 1.6)

~xce$sivegeneration of the slack node is 2

(11.7)

is the SUM of the active power demand an the tran~rnissionloss the value of which is set to that found in the i ~ e d ~ a t prev~ous e ~ y load flow s ~ ~ u t i oofn The loading of unit 2 is then modified according to

(11.8) exceeds the maximum loading o f unit 2, it is cessive gener~tionof the slack node left to be sh

377

Information Technology Application

3

The above proceditre is repeated to modify the loadings of the units 3 to N. After all b: Ioadings of the units are ~ o d ~the ~ slack ~ d node , active power will be on its power limit is viola~~a. : In the corn~etition stage, a s e ~ e ~ ~ ~ o n on from the two ~ ~ ~ h a n iiss ruse n I so~utionss h o u l ~ selection*The selection t e ~ ~ n i used q u ~is a ~ ~ ~ ~ n ascheme rn~nt their co~espondi series of N, t o ~ ~ n ~ m ~ ~ p o nEach ~ ~j ~ d§i v. i d i~isa a~s s i ~ ~ ae wore d s, according to

j=1

where J; is the fitness ofindi

are t

ual i. The opponent r, is c

tes the ~reatestinteger less th 1 [0,l]. The k highe a as ~ the ~~ ~ d~ i vini the d next ~ ~ ~~e ~ ~ ~ r a t i o n .

(I 3 * 10)

-

Power System R

~

~ and ~R e r ~c~ u l a~t i ~ n~

method o f switching which is applied within the load flow stage. When a PY node ha node, it is no longer possible to control the voltage at that bus result the algorith~does not adjust the voltage o f a switched PVnode.

11.4.5

~ r ~ ~ iAccelerat~o~ ent

to the large dimens~onali~ of the OPF p r ~ ~ ~ ee m v o, ~ u ~ o nca ~ ues such as EP can take an ~ a c c e p ~number le of iterations to c ~ n v e e speed of convergence of the EP-BPF algorithm, acceleration t e c ~ ~ ~ ~ rovide an inte~ediater e ~ m a ~ pof i ncandidates ~ to a more optimal ~osition, o nthe popu~ationis moved in led. To achieve this acceleration, a ~ r o p o ~ of of the negative gradient. This is achi the dir~ct~on o ~ ~ hofm1431. As the gradient step forms only choice o f step size is not as critical as it is in a to a constant sniall step size to enswe convergence. The sens~tivityo f the s o ~ u ~ otonc h ~ g e s ent variab~esvaries solution is less sens~tiveto changes in activ than to ~ h ~ g ine s magnitudes and iransfo ap settings. As a result of this, di has a large step size while r step sizes. These variable-

D is pr~vidinga focused local o p t ~ i s a t ~ o whil. n, ation. Reactive power penalty terms are not includ SD f o ~ ~ u l a t i oexcept n for the slack node, which cannot be switch process. The effect of generator node sw~tchin in the load flow routine creates d i s ~ r b ~ c in e sthe solution rocess. These disturb metho~$such as SD diverge or converge to local optima. However o p t ~ ~ i ~ ascheme ~ i o n ese ~ r o b ~ e m are s avoided. o reflect any penalties in the fitness function ( will often discard solutions produced not inc~udedwithin SD p ~ ~ a I t i in e s(11.3). will usual^^ incur gre hrn p e r f o ~ swell on convex ~ r o b l e ~whe s ver, If the solution surface is multi~modalth become trapped in a local optimum. This i s the case wh odelled by non-convex curves such pie ents 139,461. These c u ~ e present s a pr and discontinuities in fie gradient. this gradient, it is possible for the solution to cross the As the step is bas d i s c o R t i ~ uwhere i~ ent i ~ f o ~ a t i oi snno longer v s beyond the local such that if an active power loading of a unit crossed a disco that b o u n d a ~ These . bound~ies,while Iobal EIP ~ ~ e ~~~~h w omutat ~ ~

Information Technology Application

11.4.6

A ~ p l ~ c ~ tSfu ion

The EP-OPF a l g o ~was t ~applied to the IEEE 30-bus test system. Three sets of cost curves were used to illustrate the robustness of the technique. The fust case is where all curves are quadratic 1471; in cases (b) and (c) some of the cost cewise quadratics or quadratics with sine components. Ther~forein are many local optimal solutions for the dispatch p thm cannot d e ~ ~ i the n e global o p t ~ msolution^ ~ e for va~~dating the developed algorithm. lemented using the 6: ~ ~ g u ~ g e ~entiL~m Fro computer. The speci pro~ramwas execu e dthe Append~x, In all cases, the standard a l g o ~ t hand ~ system data are s u ~ a ~ i s in IEEE 30-bus loading is used. as^

are represented by quadratic functions from [ In thi s ~ m a ~ s in e dTable 11.2. The program was run 100 times with the se A p ~ e n ~The i ~ .average cost of solution obtained was $803.51 with the mi $802.62 and ~ a x ~ m $805.61. u ~ The average execution time was 51.4 s o ~ ~ tdetails i o ~ for the ~ i n i m u mcost are provided in Table 11.2. For this case, a solution of $802.40 was reported in [471. This was obtained using penalty functions for generator reactive power limits. The EP-OPF returned a solution with no PV nodes being switched. ~ o w e v ~the r , solution from [47] violates the slack -limit s~ightlyby approximately 1.7 le 11.2. Generator data and cost coeficients for base case (a)

K=-

Bus

P,"'"

r,mm

No.

MW

MW

MVAr

MVA

a

b

C

1 2

50

-20

250

-20

100

-I5 -15

11

15 10 10

I3

12

80 60 50 60

0.00 0.00 0.00 0.00 0.00 0.00

2.00 11.75

5 8

200 80 50

0.00~~~ 0.01750 0.062~0 0.00834 0.02500 0.02500

20

35 30 40

,""

-10 -15

Generation inpu?/outputfunction

Cost Coe~icients

1.00

3.25 3.00 3.00

c,= a, +b,e + C , e 2

6

In this s ~ ~ units y , cost curves were replaced by pie summarised in Table 1R .3 to model different hels or valve-point cise c o n ~ oover ~ units with d~scontinuitiesin cost curves, the ~ n with ~ t st capacity was selected to be the s bus. The average cost of solution $649.67 with the m~n~rnum being d ~ a x i $652.67. ~ u ~ The ~ v e ~ ~a xge~c u ~ i o ~ for the m ~ n i ~ ucost m are ~ r o v ~ in d eTable ~ blem, it failed to con was a ~ p r o x i ~ a t e ~ y

Power System R ~ s ~ c and ~ ~ere~ulaEion ~ n g

3

D has d~~ficulties with n ~ n - c o n vs~o ~l u t i ~surfaces. ~ It is global o ~ t i if~ the u ~ o d i ~ c a ~di e~s nc s~ b e ~ ever, the global o p t i ~ u mwill ing intervals for units 1 entire solution space unlike The voltage profile at the solution is shown in

d ~ ~ o n s ~that a t ethe~

I -

- -

10

5

15 Node

20

25

30

Voltage Profile Solution in Case (b) Generator cost coefficients in case (b) MW

To R4w

a

50

140

55.0

Bus

From

No. 1

2

Cost Coefficients b c 0.70

0.0050

140

200

82.5

1.05

0.0075

20

55

40.0

0.30

0.0100

55

80

80.0

0.60

0.0200

c,?*

~ ~ n e r a t i oinput/output n function C, = U, c b,e -I-

curves of the generators ~ ~ ~ n e c ~ e onent s u p ~ ~ ~ p o s e d loading effects [39,

are pro~idedin Table 11.5, To i l l ~ s ~ r a ~ ~ stics of the po~ulationover the 100 it can be seen that the El?-0

.15 Coizvergence ofthe EP-OPF algorithm in case (c)

thm i s applied to this case its abiIity dependent on the s of n Q ~ - c o ~ v eeost x c ~ ~ ethe s , a b i ~ of i ~the sol~tiQnis great~yreduced. The ~ ~ a l lsearch e l mechan~smsof method for a dirccted local search, ~ ~ o w eperform v ~ ~ , well in these cases. Generator cost coefficients in case (c)

Cost cQefficien~$

ax US No,

1

2

MW

MW

a

b

C

d

e

50 20

200

150.00

2.00

0.0016

50.00

~.0630

80

25.00

2.50

0.0100

40.00

0.0980

Gen~ra~ion i n p u ~ o u ~ cost u t function

C, = a, + b,c + C,

set to an almost pure^^ cost-ba~ed le, may have a less des~rablevol the objective ~ n c t i o nof ~ ~ ~ t i ito is n s~ ~o s ~ ~to ble de a flatter voltage profile. Ideally all load nodes will have a voltage m a g n ~ ~ of d e I per unit. To ~ c ~ this ~ the ~ v e ~ ~ e~ ns c st i o n(I 1.3) was ~ o d ~ to~ e d

Power System Restructuring and Dere

fi

h4 /

VFk=

(11.11)

E

k

KJ(vk -1.0)~ if V, $1.0,

k aP

othe~ise

The ~~r~ VF, denotes a penalty term on a load node k and K,is a constant penalty he ~ ~ ~w ~i ~ ~then ~SD if o ~e u l as t i o n v ~ o ~ w~ ~tr e~also ~ n to the form of VFkabove. With this penalty the 1 a ~ ~to ~m ipn ~t~ i s e the cost o~generationwhile trying to ~ a i ~thet load a ~ fl To ~ e m ~ n sthe ~ teffect e of this change, case (b) The voltage profile achieved is shown in Pi ge level to load nodes 51.54, which is close to th able 11.5. Of the 10 a b e ~ e profile r than that found in (b ~ i f ~ cinu~lr ~o v ~ daidne~q u a ~solutions. ~

ble 111.5 ~i~~~

P, p2

Ps P8 PI, PI, Vi

V2

V5 V8

V,, V13

t,, I,, t55

t3h

solution found by EP-OPF ia case (c)

Case (a)

Case (b)

Case (c)

Case (d)

173.848 9.998 21.386 22.630 12.928

140.000 55.000 24.165 35.000 18.773 17.53 I 1.019 1.048 1.038

199.600 20.000 22.204 24.122 14.420 13.001 1.050 1.061 1.043 1.036 1.100 1.038 I .030 1.MO

140.000 55.000 24.458 33.849 14.518 23.322 1.045 0.952 1.004 1.027 1.044 0.990 1.030 0.940 0.910 0.940

12.000 1.050 1.034 1.005 1.016 1.069 1.055 1.020 0.900 0.950 0.940

1.055

1.OS5

0.980 1.010 0.930 0.930 0.970

1.010

0.980

een widely used in the power in relaying s c ~ ~ r n load e s ~ forecasti

Power System Restructuring and ~ e r e ~ ~ a

~rocess~ng power of the V Q Neumann ~ digital computer with the abili d e ~ ~ s ~ and o n sto y o r d i n a ~ex~erience.ANNs have widely b For e n e r ~ yrnanage~ent,load flow and However, most existing ANNs for electri ~

ions such as load ill be shown in this

11 be s h o that ~ this new ‘ c ~ m ~ I e x ’ to e§ti~ateisba bar voltages in a load flow problem.

e are n n~mbero tota~lingthree lay All the x and the w in rs within an i n ~ e ~[O, a l I]. that w belong^. A set o f erscript of each w uts, dk,for b l ,...,2, ~ o ~ ~ s p oton adset i ~of~i n ~ ~ txj,j=19 s , ...,a, is used as ia ns ard sigmoid function is e loyed and the ~ollowinge ~ u a t ~ ohol

Figure 11,17 shows a cal A” for real nodes and I nurn nodes, rn number of hi is ~ e ~ is ofreely ~ extensib k

1st hidden node

7 A typical ANN for real numbers

1st OULPUL node

1

k -

ni

=1, ... 1 )

i=l

i

1

n

(1 1.12)

i = l , ..., rn

J=I

y f ~ n c t i E, ~ nis~being ~ i n i ~ i s e d 2

to obtain an o ~ t i ~set a lof values o f w usin the ~ h ~ l l - c l i ~ bai n ~~' ,the ~ o l ~ ~ w holds: ~ng

ere 2 = step size = 1.

2

basic e i e ~ ~ noft sthe newly d e s ~ g ~ ~ d ic ~ n c tsay i z~ =~w x, ~ where x is

~so that o at ~

~

t

~

e s t ~ c ~ and ~ n~ ge r e ~ l a t i o n

3

~ o ~ p nl ~~ ~xb exir sand ~ x2, the opera~i~n is clearly show^ in the ~ o w ~ r 11.18.

(1 1.15)

where

j =

Z

ask e

l of the newly ~ designed ~ 'complex' ~

Information Technology Appli~a~on

(11.116)

The new ANN for the complex number format

As this sigmoid function is highly non-linear and complicated, V E needs to numerically. The method is to perturb each w by a very small amount w values of w are kept constant. A new value E is then evaluated. The ratio of of the new E from the old E, due to the ~ e ~ b a t i o gives n , the colrespondi VE. E itself now becomes a complex number and the gradient hnctio

1/2, as defined by the following equation:

(11.17)

versus the In order to test the p e r f o ~ ~ of c e the newly deve~Qped‘co~plex’ co~ventiona~ ‘real’ in handling complex numbers, a simple ~ n c t i o nshown in e q u a t ~ o(1~1.18) is A data set with nine ~~~~g ex les ape ~vailableand ~~0~ uring the training process, we ~ o ~ t i n u o u keep s ~ y track of the total s ~ u ~ e d t from the nine training sets.

0

=

x+-1

x

(11.18)

there are ~ W O~nput For the i ~ p ~ e m e n on ~ athe ~ oconventional ~ ‘real’ on on the new ‘comp hidden nodes and ~ W Qoutput nodes. For the imp there are one input ~omplexnode, three h~ddenc o m p ~ ~nodes x and one are set to one ~nitially node. All values of w for both the mnve itrarily set to 1.5. The before ~aining,i.e. a fair initial guess, and two ANNs d ~ n ~g ~ is ~ h i s ~ of o ~s ~ u ~ error e d of the nine training ite~ationsfor the ‘ c ~ m ~ l e x ’ shQwn in ~ ~ g u 11 r e-20. It can be seen th ile the ‘real’ ANN can only ANN to arrive at a total squared error o f ~ c h ~ e av etotal square error of 3 . 8 ~ 1 0after - ~ 23 000 i t ~ ~ ~ i oAfter n s . the two two networks while the c a value of x = 0 . 2 5 ~ ~ . 2is5 gwes an output of 1.85-jl.4, i.e. should be 2.25-jJ.75. The ‘real error, while the ‘complex’ A” gives an output of 2.3-jl.75, i.e. 1.8 % error. From this illus~ativee x ~ p ~itecan , be conclud~dthat it is better to systems involving complex numbers instead of using a ‘real’ )

te 11.6 Sample values ofthe complex test function

5.1 - j4.9 2.1 -j3.8

x 0.1 -+ j0.l 0.1 +j0.2

1.1 -j2.7 4.2 j1.9

0.2 +j O . l

2.7 j2.3

-

0.2 +j0.2

I .74 - j Z 3.3 j0.9 2.61 - j1.34 1.97 -.j1.37

0.2 +j0.3 0.3 +j0.l

0

~

0.1 +j0.3

0.3 +j0.2

0.3 Cj0.3

Info~at~~ o n~ c ~ nApp~icat~on o i o ~

3

0

0

Error history of two ANNs under training

In order to make a fair comparison, the computer sim~lationhas been carried out again by us in^ thee d~fferentnetwork con~gurations. The same functi as shown in equation (1 1.18) and Table 11.6 have been used consists of two separa~ereal NNs, each consisting of one real input node, node and one real o u ~ unode, t thus t e ~ ‘Two e ~S e p a ~ W ~ es ’ . The sec0 ut nodes, two real hidden nodes and two real The third c o n ~ g u ~ a t consists io~ of one complex hidden node and one ~omplex bjective of this simulation is for detailed reduced by 10 times CO e Figure 1 1.21. It can be seen e b e h a v i ~ ~ofr two separate NNs came there is no crossery poor, as expected i n ~ Q ~ a t ib~tween on the two real error a l ~ o u g hit takes more iter ce, it can be seen fmm both Fi

. However, the time duration of an that of the conventional real , both NNs have more or less the same are ction of this section, complex widely used in electric power systems and, thus, the ‘complex9 design shod ted

Power System R

3 9 ~

~ s ~ andc ~ e~ r ~~~ lna t ~ ion P

whenever ANNs are applied to electric power systems. One typical example of a ~ p ~ y ~ g the 'complex' ANN to load flow analysis is shown in the following section. 0 35 ................................................................................................................................

7

....................

\

...........................................................................................................

.....................................................................................................

..................................

~

i 11.21gError~history~of three ~ N

11.5.4

s for comparison

App~icationof 'C'omplex"ANN to Load Flow Analysis

with one or more hidden layers is s u ~ i ~ i ein n torder to approximate any conti n-linear ~ n c ~ i oarbi~arily n well on a compact interval, provided sufficient hidden neurons are available [53]. The power load flow problem is by itself a non-linear problem and, hence, it can be ~ a I y s e dwith the h of an ANN, A six-bus network, as shown in Figure 11.22, has been used to test performance of the newly developed 'complex' ANN. us-1, bus-2 and bus-3 are generat~rbuses while bus- 1 is the swing bus. Bus-4, bus-5 and bus-6 are ordinary load buses where the P (active are to be specified. The training example is generated using ~ e ~ o n - R a p h s oalgorithms. n This is just an illus ng real application, the 'complex' ANN will continuou time state of the network in terms of voltage, P and . The details of the network pa~~meters are shown in Tables 11.7(a) and 11.7(b) below:

~ n f o ~ a t i oTechnology n Applica~~on

391

bu

~

i 11.22 g The ~six-bus~ network ~ for load flow computation .7(a) Busbar power for load flow study

Bus bus-1 bus-2 bus-3 bus-4 bus-5 bus-6

PI,,

Qio,

pgen

0

_--

Vsp,

0 0 0

0 0

0.5 0.6

1.05 1.05 1.07

P4

Q4

_--

".._

Ps

Q5

-*-

-__

--_

_--

P6

Q6

Network parameters for load flow study From bus- 1 bus-1 bus- 1 bus-2 bus-2 bus-2 bus-% bus3

To R (P.U.) bus-2 0.1 bus-4 0.05 bus-5 0.08 bus3 0.05 bus4 0.05 bus-5 0.1 b~s-6 0.07 bus-5 0.12

[email protected].) 0.2 0.2 0.3 0.25 0.1 0.3 0.2 0.26

B (P.U.)

0.02 0.02 0.03 0.03 0.0 1 0.02 0.025 0.025

Power §ys~emR e ~ ~ ~ ~ i n b~s-3 bus-4 bus-5

bus-6 bus-5 bus4

0.02 0.2 0.1

0.1 0.4 0.3

0.01 0.04 0.03

14 training examples, shown in Table 11.8, have been generated by the sofhvare for l e ~ i n by g the two ANNs. In this case, the voltage at buse three gener~torsmaintain constant voltages at the c o ~ e s ~ o n d i 3.8 Training examples for the neural networks

0.7ij0.7 0.9+j0.9 0.9+j0.7 0.7+j0.9 0.7t-j0.7 0.790.7 0.790.7 0.7Cj0.7 0.9-kj0.7 0.7+j0.9 0.9+j0.9 0.9+j0.9 0.9+j0.9 0.9+i0.9

0.7J-jO.7 0.9+j0.9 0.7-tj0.7 0.7+j0.7 0.9-tj0.7 0.7+j0.9 0.71-j0.7 0.71-j0.7 0.9+j0.9 0,9+j0,9 0.9Cj0.7 0.7+j0.9 0.9+j0.9

0.7+j0.7 0.9+j0.9 0.7+jO.7 0.71-j0.7 0,7+j0.7 0.7+j0.7 0.9+j0.7 0.7-i-jO.9 0.9+j0.9 0.9-tj0.9 0.9+j0.9 0.9+j0.9 0,9+j0.7

0.97$-j0.089 0.864-jO.137 0.969-jO.101 0.960-jO.088 Q.962-jO.l15 ~.944-j0.084 0.964-jO. 108 0.95 190.086 0.883-jQ.137 0.872-jO.125 0.903-jO.142 0.882-jO. 108 0.894-j0.~40

0.9+j0.9

0.7+j0.9

0.880-jO.116

here fore, inputs to each ‘real9 and three inputs to the ‘complex* two o u ~ u nodes t for the ‘real’ ANN and one the subscnp~refers to the number o n e ~ o remain r ~ ~ n c h ~ during ~ e d the trial test. B

load flow network is learned by the ‘real9 and ‘CO and V5.The ‘real’ ~ o m b ~ n ~ tof ~ oP4, n sP,, P6, Q4 . The initial v a ~ ~ eofs all wei S fan ordinary sigmoid function for ‘real’ [O,I] and it is not suitable for this applica~ion,the sigmoid ~ ~ ~ ~was t islight~y o n mod~~ed to the f o ~ l o w i nform: ~ 4

I n f o ~ ~ ~Technoiogy ion Application

39~

The limit of iterations for both ANNs is set to 230000 as in the case of Section 11.5.3. Figure 1 1.23 shows the variation of the total squared error of the two ANNs with r ~ ~ etoc t the number of iteration.

0.08

0.07 b

U

2 0.05

~

(B

0.04 0.03 -

C~nventionalNN

-

Gomplex PlN

0.02

0.01

0 1

21

41

61

81

101

121

141

161

181

201

221

2 3 Training errors of two AMds for power load Row

After the two ANNs have been trained, they are used to estimate V, under differen~testi~g samples of PI and Q,, i = 4,5 and 6 . There are two categories of testing samples, first set (Cases 1 to 7) being those P and Q randomly selected in between the limits of P and Q i n c ~ ~ d eindTable I 1.8. Another set (Cases 8 to 12) is randomly selected outside the limits of the two ANNs. The P, and QIunder test are shown in to test the ability of ge~e~alisation Table 11.9 while the results are shown in Table 11.10. TaMs 11.9 Test cases €or the neural networks Case P4+jQ4 I 0.77i-j0.82 2 0.72+j0.76 3 0.83-tj0.87 4 0.75+j0.77 5 0.841-jO.81 G 0.88+j0.81 7 0.80+j0.80

Ps+jQ5

P6+jQ6

0.75+j0.79 0.84+j0.73 0.88+jO.S1 0.77-tj0.80 0.72+j0.79 0.82+j0.89 0.82-tj0.89 0.80+j0.76 0.71+j0.77 0.79tj0.82 0.83+j0.87 0.751-j0.82 0.80ij0.80 0.80+j0.80 I

394

Power System Restructuring and Deregulation

8 9 10 I1 12

0.61-1-j0.69 0.92+j0.95 0.58-tj0.69 0.76+j0.94 0.791-jO.87 0.61+j0.57 0.60-tj0.60 0.6O+jQ.60 l.OO+jl.OO l.OO+jl.OO

0.781-jO.67 0.97-tj0.8~ 0.94+j0.68 0.60ij0.60 l.OO+jl,OO

Comparison of the neural networks V, Meal NN

Case

V, /Correct

1 2 3 4 5 6 7 8 9

0.935-jO.109 0.920-jO.1I I 0.924-jO.1 15 0.921-jO.108 0.912-jO.103 0.923-jO.115 0.91790.110 0.922-jO.109 0.932-jO.101 0.923-jO.2 14 0.907-jO.113 0.923-jO.118 0.924-jO.112 0.92290.113 0.923-jO.113 0.919-jO.101 0.897-jO.105 0.920-jO.105 0.974-jO.108 0.923-jO.112 0.993-jO.064 0.914-jO.077 0.785-jO.169 0.928-j0.141

10

11 12

Ys/Complex NN 0.932-jO.108 0.922-jO. 1 16 0.919-jQ.102 0,919-jO.114 0.931-jO.101

0,910-jO.113 0.921-jO.112 ~.922~0.114 0.920-j0,120 0.970-jO.106 0.960-j0,062 0.889-jO.160

o Figure 11.23, although there are on t r a ~ ~ nexamples, g it can be seen is smaller than that of the newly that, initially, the total squared error of the ‘real’ ANN. Actually, after 500 ~ e v e ~ o p e~comp~ex9 d s, the total squared error of the WN has akeady attained a steady-state va d 0.032. AAer 4300 iterations, the ‘complex’ ANN catches up with the ‘real’ and the total s q u ~ e de improving. A ~ e 23000 r iterations, the error is ctually, if we c o m p ~ e e 11.20, one very interesting result can be noted, It seems that the ‘real’ get itself into a m i n i ~ u mafter a small number of can continuous~yimprove itself during the le time for training is about 90 s and 150 s for the real and comp PI1 300 PG is used under ~ i ~ d o 98, w s A l t h o u ~the ~ c ~ a isc$imi~ar. y It is suspected that it is quite easy for er random initial guesses of weights have been trie similar resu~ts. This is one merit of the newly developed ‘complex’ A m . Next, Tables 11.9 and 11.10 are referred to where we want to test the power of pred~ctionof the two ANNs. The first seven testing samples have been randomly selected in be^^^ the limits of t0.7, 0.91 of both the P and Q. The ‘complex’ ANN behaves better in all cases. ~ o w ~ v ewhen r , five alternative testing samples, selected outside the limits, are tried, the ex9 ANN behaves better except in case 9. Et appears that, in general, the ‘CO s more p r e f e ~ ~for l e the present application.

I n f o ~ a t i Technology o~ Application

95

.4 power has doubled Since the 1980s’ formation process in^ has exploded. Process every two years. Today, the Intel Pentium I11 runs at a clock spee Et is expected that by 2002, the chip could run at a clock speed of 3 t the technology could create, sto search and process vast amounts o ’have yet to advance the techno y further to access and interface easily. Tra~it~onal~y, interactio ith a comp~terhas involved mouse or j o y s t i c ~ ~ a c k b aevicc to input information and the us m the system. With the development of virtual re (VDU) to receive output syst~ms7 new intera~~ion ods have been developed that allow the user to computer- generate^, or virtual, environments (VEs). VR can be considered an e x t ~ s i o nof ideas which have been around for some considerable time, such as flight s ~ ~ u l a tand io~ wide screen ~ i n e m aUsing ~ such systems, the viewer i s presen~edwith a ~ c r which e ~ ~ ~ on of the visual field giving a powerful s ite of technologies which permit human resen~tionsof i n f o ~ a t i o nheld in c , a u d i t o ~and tactile stimuli, eac cant extension to the way the users kte shared unders~nding,lead to simulate inacc allowing the user to extract the lessons to be learned without the inherent risk, This alltsws a crte a ~ - ~ ~with m e a computer-generated e~vi~onment in a s i ~ p l ~ , the user to i ~ ~ t ~ rin ‘natural’ m ~ n e rw7 ~ t ~ othe u t need for extensive mining. Pres av~i~able budget and requires high levels of nts in low-cost desktop e technology more of smaller ~ompanies.The strength of VEs is in ion of the n a t u int~ractive ~~ skills of the human. As esktop’ VEs systems, inte~ratingnovel display widely used. The po~ential and a great deal of research is currently develop these technologies into effective useable eaply on a conventional desktop large-screen display for mult~~user pa~icipation.A l ~ o ~ not g h always re~evantto use, i ~ e r s i v erepresen~a~~ons can involve the use of head-mou~teddisplays tactile gloves, and other devices to enhance the effect. Ap~licationsrange from simulations cal items ~ ~ a n g ~from n g buil~ingsto mole~ulars ~ c ~ r e to s )more abs~act such as the disp~ayof large amounts of t ~ ~ e - v data ~ ~ (e.g. n g analysis of world lex databases) or illus~a~ing intan~ib~e concepts [54].

11.6.1’

Types of

A l ~ ~ o u githis dif~cultto catego~seall V systems, most con~guratio gory can be ranked by the sense sion or presence can be regard

Power stem ~

3

the user is focused on the tas in hand. ~ to be the pr~ductof several param ity, s t e r ~ s c o p ~vic field of regard and the in i s ~ ~ and ~ ~theo level n of imersi factors ~ v o i v e ~ .

~

e

~

~andcDere ~

e presen~e ~ i is o

i

n

~

of the display. For VE will incr~asethe

~

terns are not re

T

systems (adapted from [SS]) Main ~ e ~ t u r e s

Scale Sense o f ~ i aw~cn~ss

~e~i~i-High

LOW

~ Low ~

Field of regard La Sense of immersion

t

o

~~ e~ ~ ~i ~ m

Medium Low Low Low ~ o n ~ L ~ w Mediu~-High

Vision i s the m a i ~sense th shade, etc. V

~

High High

High

d e ~ i ~ e rhave s co~cen~ate ed such as 3-D graphics, vwi ~ i s very s c~o ~ p ~l e xand~ concen~ateson ~ n t e ~ r e t a tof i oi~n f Q ~ a t it~~naist

ective (i.e. what is seen varies from per so^ to tter at sim~atingvision than non-imme~ive ' forcing concentration on the virtual I on-po~able.There are also three related betow.

II.~.6 Cave Gave is a small room where a computer"generated world is ro~ect~on is made on 0th the front and side walls. This soluti ce collective VK experience because it allows different people to share the same e x ~ e r i ~ nat the same time. It seems that this t ~ c ~ o l o g i c solution al i s p ~ i c u I ~ appropriate ly for cockpit simu~a~ions as it allows views from differ~ntsides of an i ~ a ~vehicle. i n ~

~ cameras ~ d at ~~lepre§ence systems immerse a viewer in a real world that is ~ a p by~video a d ~ ~ location ~ a n and ~ allow for the remote m ~ i ~ u l a t i o n ~ ~ a ~ o Telepresence r§. is used for remote exp1orai;ionlmanipulation of hazardous e n ~ ~ r o ~such e n as ~ sspace and u n d e ~ ~ ~ ~ ~ e r .

of real e n v i r o n ~ e ~with ts logies of ' ~ u g ~ e n t ereality' d allow for the view s ~ p e ~ ~ ~ virtaaI o s e dobjects. As a matter of fact the 's view of the world is sup~lementedwith virtual objects and items whose mean in^ is aimed at e ~ c h i the n~ ation content of the real e n v ~ o ~ e n t .

has been used by the military and by space scienti p h ~ a c o l o ~ i s~~os l, e c u biologists 1~ and theoretica~~hysicists into its domain. Simply speaking, the t e c ~ o l por ~ o v i ~ hei e~ ical attributes of scientific models. It is a too2 on of data to a new dimension, to the point at which the user, i teracts with, or is e n ~ l by f ~the model that has been c r ~ ~ t e d . to accelerate scientific ~nders~anding by enablj~gthe

VR to v~sual~se and ~ r o t o imaginative ~ e p ~ o j e ccan ~ s h o ~ the ~ n Iif~cyc~e of make them available much earlier than would otherwise be the case. n ~ e ~ ~ tmedia, i v e where a wide variety of ~ x p e r i ~ ccan e s be create exp~oreat their Q W pace, ~ choosing their own ~ a t h w a y ~ ~ As the technologies of VR evolve, the a ~ p l i c a ~ of ~ns i s assumed that VR will reshape the interface b e ~ e e n p by ~ ~ f e new ~ nways g for the c o ~ u n ~ c a t i o ofninformat

Information Technology Application

399

and the creative expression of ideas. Note that a VE can represent any 3-D world that is either real or abstract. This ~ncludesreal systems like buildings, landscapes, spacecra s c u ~ p ~ rcrime e s ~ scene recons~ctions,solar systems, and so on. Of special jnteres~are the visual and sensual representation of abstract systems like magnetic fields, turbulent flow s t r ~ ~ t ~molecular es, models, mathematical systems, auditorium acoustics, stock behaviour, population densities, and any other artistic and creative work of abstract nature. These virtual worlds can be animated, interactive, shared, and can expose behaviour and ~nctiQnality. "hough still relatively new, VR has already been put to use in a number of different, iniiovative ways. In the world of industrial design, engineers are using CO simula~ionsof prQto~pesto speed up the time required to take a new product from the drawing board to the productioii line. In the world of science and medicine, doctors are computer-simulated pathologies to determine the outcome of ~o~entially risky cedures before these procedures are actually p e ~ o ~ on e d il e ~ ~ i n e e ~architects ng, and interior designers are using VR s realistic, computer-genera~edsimulatjons of proposed environmen~.The can then be ~ o d i ~ eindreal-time based on client input, zoning ordinances, ~ e s t h e ~ i c concerns and budgetary considerations. In the world of weather fore casting^ VR is being used to predict weather p a ~ e and ~ sto where a storm will make 1 create h ~ r i ~ a m n e o ~ wh~ch ~ ~ can s accu~atelyd s astronomy students tour and when. In the world of higher education, galaxies and physiol students tour the innermost workings of the human body. A VR sim~lationof a CO pip~workla~out,for example, could allow access, main~nance and safety aspects to be examined at the design stage, more effectively than by mode~~ing. It imi~ediateIypermits the evaluati~nof routing and accessibili~~ thereby avoiding expensive, t~~e-consuming correction during or even after c o n s ~ c t i o nT~h e r ~are many portunities that have yet to be explored. romising In the ~ n ~ o ~ a t iage, o n VR has been identified as one of the de~~elopment areas. There is a constant improvement in marketing per of both quality of appl~cat~ve VR systems and receptiveness of potential customers, T h i s is due to decrease of the cost of VR systems and devices, (2) the rmance r e l i a b ~ l iof ~ the t e c h ~ o l o (3) ~ , the extKeme~y ed from VR use in its various forms and purposes such as gh the t e c ~ o l o g yis mature enough to have d~fferentappli~atio~s, there resolved for its use for practical app~ications. The sensational press cov associated with some of these t e c ~ o l o ~ has e s led many ~ o t e n ~ iusers a l to overe e the actual capabi~~ties of existing systems. Many of them must a~tuallydevelop the t e c ~ o l significantly o~ for their specific tasks. Unless their expertise includes ~ ~ o w l e of ~ gthe e human-machine interface requi application^ their res~ltingproduct will rarely get beyond a 'conceptual ~racticalapplications. Current VR products employ proprietary hardw There is little doubt that incompatibility between different systems is restricting market growth at present. It is probable that as the market matures, certain s t ~ d a ~will d s emerge. The premise of VE seems to be to enhance the interaction between people and their systems. It thus becomes very important to understand how people perceive and inte events in their environments, both in and out of virtual represen~tion of reality,

~ u n d ~ ~ ~questions n t a l remain about how people interact wi the SYs~ems,b v h e y may ce and a u g ~ e n cognitive t p ~ r f o ~ a n in c esuch e n v ~ o n ~ p l o for y ~i n s ~ c t i o n~, i n i n and g other ~ ~ o p l e - o ~ e

The t ~ e system ~ a consists ~ of an infrared camera, shown in Figure 1 1.24, a shown in Figure 11.25. The ~nfrareddetectors inside the camera are cooled argon to and they sense ds p ~ c ~ in m the range betwee while floppy disks and h i g ~ - s ~ e e d proce~singon a PC,

QISS

are o ~ f e r ~tod

Power System RestruchJring and Deregulation

402

DMK with pannin~tiltinghnctions can give the absolute c o o r d ~ ~ tof e seach grid point on the power e ~ u i p ~ while e n ~ the thermal system can give the rea~-times u ~ a c tee ~ p e r a ~ r e of that grid point. When this information is fed into a tailor-made ~ - b a s software ~ d e, a 3-D thermal image can be displayed and manipulated. The major prob~emhere is with the co~espQndence between the DMK and the thermal system, i.e. matching eve^ point sensed by the DMK lo a c o ~ e s p o n d i point ~ g on the thermal image.

2 6 Laser-based ~ s t a n c e - m e a kit s~~~

of the system and the procedure of calibration. Sobel [62] and G e ~ e ~ ar o p t ~ s a t i o nfor c a ~ e r acalibratio~.A comp~hens survey of the ~ ~ t e r aand ~ r ediscussion of r n ~ ~ ~for o des~ e ~ r o nci c ~ ~was ~ p ar ~ §s ~ ~ t e d Lenz and Tsai [64]. It has been found that the s i ~ p l e cs a~~ e r amodel and its calibrat~onpro ti~isation. At the same time, its level o f for our application. A pin-hok camera of a standard e ~ e c ~ o ncamera. ic Let in the ~ n ~ v e r 3-D s a ~world coordin coord~~ates of the image point on the thermal i to I(Xp 5)is modelled a standard approach i prQ~ecti~e coordinates. a 3x4 matrix, is known where the element, t34,i t inside the transfo scaled to ' 1'. Let c ~ ~ ~ d i nvector, a t e together with the two ation of the ~ a ~ s ~ o ~ a t i o n .

(11.28)

with known, (xwj9ywi, ine the 11 u n ~ Q w nt to have more points so an ~ v e r a g i ntechn ~ n is larger. In other ~ ~ r dthe s , fQllQwingn sets

n = 6 are e ~ o u ~ toh

where

( tl1 f 1 Z t13 814 121 822 t 2 3 t 2 4 131 t 3 2 f 3 3 >'

f the least sum of squares over the n number o f ~ a ~ ~ poin b r a ~ ~ ~

From ~ ~ ~ e reight i ~ calibration n ~ ~ , points are

Y w j Xwj

( I 1.24)

Information T ~ c h ~ o A ~ ~ g~ y~ ~ c a t i ~ n

xwj

Ywj z wj

1

Power System R e s ~ c and ~ D ~ enr ~ ~ l a ~ i o ~

4

object has the same spatial resolution with respect to the original one. Inte~olation surface temperat~eis by means of a similar process. The grid points are generated in appropriate sequence by the two ~ ~ - c o n stepper ~ ~ ~ ~ e m o t o ~ ,For each of the n number of 8, within the specified ran^^, there are m n ~ b e r of s 8, ~ i ~ h another in specified range. Hence, the grid points can be viewed as elements of es where n x m = M , each representing the x, y and z coordinates of ely. For the (ij)grid point where i = 1, ..., n-1,a n d j = 1, ..., rn-1, nts, namely (i+lj), (ij+l) and (i+lj+I)> are conside ing ( i j ) , (i+lj) and (ij+l) and the other hav (i+Xj+l). The equation of the first plane is given by the followin

(1 1.27)

ints are created on relevant planes. The surface te in equat~on(1 1.27) can be oint from the three vertices are I L

(11.28)

The ~ e ~ p e r a of ~ rany e point on the three sides of the tri ~ n t e ~ ~ of~ the ~ t two ~ overtices, n i.e. the two end points a ~ ~ ~ t i omatrix n a l ~~nsist~ g r must be s ofnnine the e q ~ a t i (1 o ~1.29). ~

(1 1.29)

7

~ f o r m a ~ i oTechnology n Application

11-7.4 I ~ p ~ e ~ e~ ~ ~~ ~ a t ~i op ~ ~

e

The competitive electricity market: raises utility cost consciousness. n o ~ a i l yassociated with equi~menti n v e § ~ e and n ~ continuous niai system. Power trans~orm are one of the most expensive elements in the s ~ s ~ e m . ~ d e n t i f i ~ a ~of ~ oany n hot ots, i.e. potential faults, could provide benefits inclu extended ~ ~ n s f o Ii ~ e r es, reduc~ion in risk of failures and i maintenance s ~ a t e ~ i e A s , t r ~ s f o ~room, e r shown in Figure 1 1 2 8 , housing three 1500 kVA 11 kV/380 V ~ ~ s f o ~ ewas r sused , for implement in^ the developed system.

1.28 Three 1500 kVA ~ ~ n s ~ ointa~typical e r t r ~ s f o ~plant e r room

in the x direction, E200 m, rection, while the s u ~ a c ete from 24.5s"C to 44.8"C with a resolution of 0.2"C. It takes about 3 seconds to record the and o ~ e n ~ t of ~ oa n ,thus needing more than I ~o~~ the w ~ o p~oe~ i by ~n There is not enough. inates and surface t e m p ~ r a ~of r e each of points belonging to a con 91. ustration while the full i s shown in Figures 11.31 and 1 1 X . The x, y and z coordinates measured in metres, are t ~o o ~ d ~of~ an a ~image es the a b s o ~ u ~ int with respect to the c o ~ r ~ i n a st e ~ s of~ the e ~ DMK. In the ~ e o m e mode, ~ ~ athe ~ 3 surface of the ~ a n s f o i ~s as e ~shown in Fi 1 1.3 1 without any information on temperature. It should be noted that the 3-Ea surface i s not totally identical to the real surface in this situation. The reason is that part of the

Power System ~ e s t ~ c and ~ i n ~ lation

9 T r a ~ ~ ~ o rNo. m e3~under i ~ a g i ~ g

ower t e m p e r a ~ e , All these fe

The c a ~ i ecover has a

rdinates versus surface temperature

4m) 2.5

2.5 2.5 2.5 2.8 2.8 2.8 2.8

Y m 1.7 1.8 1.9 2.0 2.3 2.4 2.5 2.6

z(m) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1

TC) 43.8

37.0 31.4 41.2 27.2 2-72 27.2 27.5

0 c_

E -0.2

cc

N

-0.4

3.5

surface ~~~~e~ a ~ ~in the f geometrical ~ ~ e rmode

410

ower System Restructuring and ~ ~ ~ e ~ l a t i o n

~ u ~ h e r m o rthe e ~ user can h e l y adjwt the viewing angle to concentrate on m y p ~ ~ c u l apart r of the ~ v ~ ~ o n mfor e n ele ~ ~ e ~ o ~ r a cpmh sy mode. The 3-0 i n f o ~ a t i o nof all compon play. Ths designer can thus fly around improper placement of equipm the 2-D draw~ngsconv~ntional~y suppli has been com~~ssioned, regular thermal t spots in the equipment can i l ~ ~ a of~ these ~ o hot ~ sspots can b A point to be noted is that a sk because any technically proficie d the 3-D t ~ e ~ o g r ~ s .

In this chapter, we have considered four hot topics in i n f o ~ a t i otech ~ & p ~ ~ i ~ a ~ ni oa ~ n es ,l y ,~nte~~igent agents, evo~ution ~ r o g r a ~ m i n gvi, n ~ n e ~ ~ r rk s , ~ ~ se of ~ e r i v a t ~financial v~ instrum I: and useful tool for manag~ngrisk in l ~ c a btoi ~valuing ~~ e Scholes to set the market valuation of put options on e l e c ~ ~ ~ en used to evolve a ~ e n whose ~ s fitness was ~ ~ a § u r by e dtheir

d has failed in the

s for ~ a i n t ~ n a n c this e, e optical and thermal

Iiifomatioii Technology Appl~ca~~on

devised, resu~tingin a new gradient ~ c t i o for n back propagation. In order to demonstrate d ‘complex’ ANN is superior to the col~ventiona~ ‘real’ that the newly de novel t ~ c ~ ~isqcarried u e out for load flow simple app~ication power network consisting of six buses. It conc~udedthat the ‘complex in two aspects. Firstly, the ‘complex’ will not the conventional ‘real trapped in a locd min Secondly, it seems that there is an improved ability to cases not fal~ingwithin the trainin ~~

The authors would also like to thank IEE and IEEE for granting p e ~ i s s ~ otonEeproduce the materials contained in references [4,61] and [9,1 13 respectivley.

fa: The load Row data for the system is that of the s t ~ ~ a r d

hes ll,lZ315 and 36 are in phase tap-ch tap pin^ ranges of &lO% with a step size of 1%. The busses is 0.95 p.u, while the upper limit is 1.05 p.u. generation nodes have an upper limit of 1.10 p.u. ~pulationsize is set at 20 and the The number of tournaments Nt was set at tant M in the ~ ~ efunction s s in (1 I ) is d e t e ~ ~ by e dthe set o f ?%“he value o f a in (11.5) was 0.9. limit c h ~ k i was ~ g started at Row when a previous solution w not used in the load flow in iteration 1 when it was. For all cases the weightings within the r ~ ~ e vtoa the ~ t case are K , = ~ 0 0 0K, q = 10,000 where QS,&is in

.

ing gradient acceleration is 50 % in all cases. d in EP and in comparison studies include penalties identical to those described in [43] for voltage violations with weighting of 20. ~enalties for act~veand reactive power limit violations e slack node of this For case (d) ngs of 30 and 10 ~esp~ctively. e, the penalties for voltag~v i o l a t i o ~are replaced by penalt~es of the form of Vt;;. in ( weighting K,. of 10. The SD step size for all cases (a)-(c) is 2.0 for active 0.001 for v o ~ t a gfor ~ ; case (d) the step sizes are 0.00~ for active po er tap and 0.001 for voltage. Q-limit treatment for all nodes other e switching d within the load Row routine. the slack node is ~ ~ d l by

[I]

M.P. Wong, ‘ ~ ~ ~int~lligence ~ i a and l neural network applications in power system^', Invited Paper, Proceedings of the International Conference on Advances in Power system Control, Operation & ~ ~ n a g e ~IEE, e n 1943, ~ , pp.37-46.

41

Power System Restructiiriiig and Deregulation

[2] S.B. Lau and K.P. Wong, ‘An artificial neural network approach to transient stability

[3]

141

[S]

[6]

[71

181

[9] [IO]

[I 11

[I21

1131

1141

[lSl [16]

assessment’, Australian Jourrral o f Intelligent Information Processing Systems, Vo1.3, No. 1, 1996, pp.75-85. R.P. Wong and S.U. Lau, ‘An artilicial neural network approach to modelling generator fuel cost characteristics’, Journal of Institution of engineer^, Singaporc, Vo1.36, No.6, November 1996, pp 7 1-77. W.L. Chan, A.T.P. So and L.L. Lai, ‘Initial applications of complex artificial neural networks to load-flow analysis’, IEE Proceedings - Chzerufron, Transmission arid Distribution, Vol.147, No.6, November 2000, pp.361-366. K.P. Wong arid E. Tsoi, ‘Genetic algorithms approach for the evaluation of trade-off between economic and environmental costs in power dispatch with multiple fuels and pollt~hnt~’, Proceeditzgs ofthe International (hference on Advances in Power System Control, Opemtron & Mmagement, IEE, 1995, pp.553-558. K.P. Wong and S Y .W Wong, ‘Combined genetic algoritl~mls~niulatedannealing/fUzLy set approach to short-term generation scheduling with take-or-pay fuel contract’, IEEE Transactions on Power Sys%ms,Vol.1 I, No.1, 1996, pp.112-118. K.P. Wong and S.Y.W. Wong, ‘Hybrid genetic/simulatetl annealing approach to short-tcrm multiple-fuel-consta~ed gencration scheduling’, iEEE Trmsacfiom on Power Systems, Vo1.12, No.2, 1997, pp.776- 784. L.L. Lai and J.T. Ma, ‘Genetic algorithms and UPFC for power flow control’, Interntrtional Journal on Engineering intelligent Syrlenw, vo1.4, No.4, CRL Publlshing Ltd, UK, December 1996, pp.239-242. J. Yuiyevich and K.P. Wong, ‘Evolutionary programming based optimal power flow algonthni dispatch’, IEEE Transacfion,son Power system.^, Vo1.14, No.4, 1999, pp.1245-1250. K.P. Wong. A. Li and T.M.Y. Law, ‘Advanced constrained genetic algorithm load flow method’, IEE Proceedings - Generafron, Transmission and Distribution, Vol. 146, No.6, November 1999, pp.609-616 Derek W. Lane, Charles W. Richter, Jr. and Gerald B. Sheble, ‘Modcling and evaluating electncity options markets with intelligent agents’, Proceedings of the International Conference on Power Utiiity Deregulation, Restructuring and Power Techriologiec 2000, City University, London, IEEE, Apiil2000, pp.203-208. L.L. La], H. Subasinghe, N. Rajkurnar, b. Vaseekar, B.J. Gwyn and V.K. Sood, ‘Objectoriented genetic algorithm based artificial neural network for load forecasting’, Lecture Notes in Computer Sciencc, LNCS, Springer-Verlag, Xin Yao et al. (Editors), May 1999. K.P. Wong and C.C. Fung, ‘Development of a fuzzy-logic-based control algoi-ithm for the commitment o€ energy sources in an integratcd energy system’, IEEE Conference Prvce~dings First Australian and New Zedand Crinference on Intelligent Information Sj)stetris (ANZIIS93), Decembcr 1993. pp.432-436. P.C.K. Luk, L.L. Lai, T.L. Tong, ‘GA optimisation of rule base in a fuz7,y logic conirol of a solar power plant’, Proceedings of the internufional Covference on Power Utility Deregzilatzun, Restructuring mid Power Technologies 2000, City University, London, E E E , April 2000,221-225. L.L. Lm, ‘An expert sysrem used in power system protection’, IFAC Symposici Ser~es,1990, Mo.8, Pergamon Press, Oxford, pp.489-494. J Bradshaw, eci., Sojware agents, MIT Press, Cambndge, Mass., 1997. 61

I3

Information Techiiology Application

J,S. Rosenschein and C. Zlotkin, Rules of Encounter: Designing Conventions for Au~omated Negotiation among Computers, bridge: MIT Press, 1994. Barbara ~aye§-Roth,Robert van Cent, Rembert Reynold, M. V a u ~ a n Wescourt, ‘Agents application’, IEEE Intelligent Systems & their Applical 99, pp.23-27. A. Brooks, ‘A Robust Layered Control System for a Mobile Robot’, ~ E E EJ a ~ r n ~of l Robotics and Automa~ion,V01.2, 1986, pp.14-23. J. Ferber, ‘Simulating with Reactive Agents’, in E. Nillebrand and J. §tender (Eds.), Many Agent ~ ~ ~ u ~ aand t i Artificial on Lqe, ~ s t e r d a m10s : Press, 1994, pp.8-28. R,A. Brooks, ‘Intelligence without representation’, Artscial Intelligence, Vo1.47, 1991, pp.139-159. M. Wooldridge and N.Jennings, ‘Intelligent agents: Theory and practice’, The ~ n o w ~ ~ d g e eering Review, Vol.10, 1995, pp.i 15-152. Nwana and M. Wool~idge,‘Sohare agent technologies’, British Teleco~m~n~cations ology Journal, Vol.14, October 1996. J. Bates, ‘The Role of Emotion in Believable Characters’, Com~i~njcations of the A V01.37, 1994, pp.122-125, A. Newell, A. (1982), T h e Knowledge Level’, A r t ~ c ~ Intelligence, al Vol.18, 1982, pp.87127.

P. Maes, (ed), Designing A ~ ~ o n ~Agents: m o ~ Theory and Practice from ~ngineeringand Back, MIT press, 1991 A. Chavez and P. Maes, ‘Kasbah: An agent marketplace for buying and selling goods’, ~roceedingsof the First ~nternationalConference on the Practical A p p l i ~ a t ~of~tn~ ~ ~ i ~ i g e Agents and ~ u l t i ~ A g eTechnology nt 1996), London, April 1996, pp.75-90. P. Waynes, ‘Free Agents’, Byte, March 199.5, pp.105-114. M.R. Genesereth aid S.P. ~etchpel,’ S o ~ w aagents’, r~ Communications of the ACM, Vo1.37, 1994, pp.48-53. G. Wiederhold, ‘Mediators in the architecture of future information systems’, IEEE Computer, V01.25, 1992, pp.38-49. C. ShebIC, energy in a fully evolved marketplace’, ~ o r t hAmerican Power S)imyosium, te University, KS, 1994. C. SheblC, ‘ d operation in an auction market structure’, Paper p~es~nted at the 1996 I E E E ~ E SWinter Meeting. Baltimore, ND, 1996. J. Marshall, Futures and Option Contracting: Themy and Practice, South Western P u b l ~ ~ ~ n ~ USA, 1989, lives, Prentice Hall, USA, 1997. Genetic Algorithms in Search, ~ p ~ i m i z a t & i o ~Machine Learning, A I

~~~

ccessive linear p r o g r a ~ i n gbased OFF solution’, O ~ t ~ m a ~ ~ e q u i r e ~ and ~ n ~Challenges, s IEEE Power Engineering nlinear p r o ~ a m m i nal~ rithtns and d~coinposit~on strategies for OPF’, Optimal Power Flow: Solution Techniques, Requirements and C h ~ ~ l e n ~ e s , IEEE Power ~ n g Society, ~ 1996, ~ ~pp. 10-24. g

Power System R e s ~ c ~ r i and n g Deregulation

[38] J.A. Momoh, S.X. GMO,E.C. Ogbuobiri and R. Adapa, ‘The quadratic interior point method solving power system optimisation problems’, IEEE Transact~onson Power Systems, Vo1.9, AUSS~ 1994, pp.1327-1336. .P. Wong, and Y.W. Wong, ‘Genetic and genetic/simulated-~ealingapproaches to economic dispatch’, IEE Proceedings - Generation, Transmission and Djstribution, Vol. 14I , No.5, 1994, pp.507-513. E401 IEEE Committee Report: ‘Present practices in the economic operation of power systems’, IEEE Transactionson Power Apparatus and Systems, VoLPAS-90, 1986, pp.1768-1775, [4 11 D.B. Fogel. Evolutionary Computation; Toward a new Philosophy in Machine Intel~~gence, IEEE Press, 1995. . Wong, and A. Li, ‘A technique for improving the convergence characteristic of genetic algorithms and its application to a genetic-based load flow algorithm’, Simulated Evolution and Learning, J.N. Kim, X. Yao, T. Furuhasi (Eds), Lecture Notes in Artificial ~ n t e ~ l i ~ e n1285, ce Spring$r-Verlag, 1997, pp. 167-176. 1431 H.W. D o m e 1 and W.F. Tinney, ‘Optimal power flow solutions’, IEEE Transactions on Power Apparatus and Systems, Vol. PAS-87, 1968, pp. 1866-1876. P. Wong and J. Yuryevich, ‘ E v o ~ u t i o n ~ ~ p r o ~ a m i n g - b a salgorithm ed for vironmentally-constrained economic dispatch’, IEEE Transactions on Power Systems, V01.13, No.2, 1998, pp.301-306. [45J K.P. Wong, A. Li and M.Y. Law, ~ ~ e v e l o p ~of ~ econstrained nt genetic algorithm load flow method’, IEE Proceedings - Generation, Transmission and Distribution, Vol. 144, No.2, 1997, pp.91-99. [46] D.C. Walter and G.B. Shebk, ‘Genetic algorithm solution of economic dispatch with valve point loading’, IEEE PES Summer Meeting, 1992, Paper No.92 SM 414-3 PWRS. [47] 0. Alsac and B. Stott, ‘Optimal loadflow with steady state security’, IEEE T ~ u n ~ ~ on c~~on Power Apparatus and Systems, Vol.PAS-93,1974, pp.745-751. [48] J.M. Zurada, Eds. Introduction to Artijkial Neural System, Info Access and Distribution Pte Ltd., Singapore, 1992, pp.1-3. [49] L.L. Lai, ~n~elligent System Applications in Power Engineering - E v o l u t j o ~ aP~r o ~ r a ~ m i n g and Neural Networks, John Wiley & Sons, Chichester, 1998. [SO] T.T. Nguyen, ‘Neural network optimal-power-~ow’, Proceedings of the Fourth In#ernationa~ IEE, Pub No Conjerence on Advances in Power System Control, Operation cft ~anagement~ 450,November, 1997, pp.266-271. [5 I] T.T. Nguyen, ‘Neural network load-flow’, IEE Proceedings - Generation, Transm~ssionand No.12, January 1995, pp.51-58. ~ i s ~ r i b u ~ jVo1.142, on, [52] W.L. Chan and A.T.P. So, ‘Developmentof a new artificial neural network in complex space’, Proceedings of 2nd Biennial Australian Engineering ~ a t h e m a ~ Confeuence, ~cs Sydney, July 1996, pp.225-230. 1531 J.A.K. Suykens, J.P.L. Vandewalle and B.L.R De Moor, ArtiJicicaE Neural Networkr for ~ o d e l l ~ and n g Control of Non-linear Systems, Kluwer Academic Publishers, Boston, 1996. [54] Virtual reality: personal, mobile and practical applications, IEE Cfflloquju~, Digest No.

in the design of an immersive system’, IEEE ~ o ~ p u Graphics ~er and Applications, Vol. 14, 1994, pp.55-59. [56] S . Kalawsky, Exploiting Virtual Reality Techniques in Education and Training: Technological Issues, S I N 4 Report Series, 1996.

Information Technology Applica~~on [57] S.G. Bumay, ‘T.L. Wi~liamsand C.H. Jones, E&, Ap~ljca~jon of therm^^ I~iagjng, Wilger, 1988. [58] A.T.P. So, F.H.Y. Chan and A.W.C. Kung, ‘A real time system for the diseases using computer~zedthermo~aphy*,Biomedical Thennolop, pp.27-35. [59] E.H.Y. Chan and A.T.P. So, ‘Application of thermography in advanced consumer elec~onics’, Frocee~ingsof the Infe#arional Symposizdm on Consumer Electronics, Beijing, C E October 1992, pp.337-340. y technique of high-voltage electrical [60] Niancang Wou, ‘The infrared t h e ~ o g ~ p hdiagnostic equipments with internal faults’, Proceedings of P O ~ . ~1998, ~ ~IEEE, N 1998, pp. 110-115. 1611 W.L.Chan, A.T.P. So and L.L, Lai, ‘Three-dimensional thermal imaging for power equipment monitor~ng’,I.. Proceedings - ~eneration,Transmission, and ~ i s € r i b ~ # ~Vol. o n147, , No.6, November 2000, pp.355-360. [62] 1. Sobel, ‘On calibrating computer controlled cameras for perceiving 3D scenes’, Artificial Intelligence, Vo1.5, 1974, pp.185-198. ~ o r ~ h o ~ , [63] D.B. Gennery, ‘Stereo-camera calibration’, Proceedings of Image Und~rs~and~ng 1979, pp. 101-108. [a]R.K. Lenz, and R.Y. Tsai, ‘Techniques for calibration o f the scale factor arid image center for high accuracy 3D machine metrology’, IEEE Bansactions on Pattern Analysis and ~ i a c ~ j n e Intelligence,Vol.10, NOS, 1988, pp.713-720. E651 O.D. Faugers and 6. Toscani, ‘The calibration problem for stereo’, Proc. of ~ ~ Miami, P R 1986, pp.15-20. E661 R.M. Taylor, W. Robinett, V.L. Chi, F.P. Brooks, W.V. Wright, R.S. Williams and E.J. Snyder, ‘The nano~anip~lator:a virtual reality interface for a sca ~Mellin~ microscope’, Computer Graphics,Vo1.27, 1993, pp.127-134. [67] G.M. Herb and C.A. Shaffer, ‘A real-time robot arm collision avoidance system’, IBEE T r i a ~ a c ~ Robotics ~ o n ~ ~and ~ u t # ~ a t iVo1.8, ~ n , No.2, 1992, pp.149-160.

r Lsi Lei Lai City U n ~ v e r sLondon i~~ UK

UK

Utility c o ~ p ~ n i~e~s e s e n t te hd e ~ ~ e l on v ethe ~ I ~ t e ~be t ~us i ~c s s es ~ ~ w the a ~ n~ ~ se as ~ e~ u t i c k ~as y possible. ~ e c ~ available ~ l ~ for ~ Intenlet y applications is difficult, bus~ness o ~ ~ o ~ n i towards ~ i e § the Internet will a d v a ~ t a ~The e . u t i ~iin~d u s has ~ a~waysbeen w ~ t i n g so that they can be purchased easily. Waiting for the In could take a long time and c only result in loss of

12.2.1

at Is fheI ~ t @ ~ n e ~ ?

A p ~ ~ ~ofthe c a ~Internet ~ n to Power ~ y s t e onitoring ~ and Tradi~g

7

catego~cs:c o n ~ e r c ~and a l n o n - c o ~ e r c ~ a~xamples l. of commercial use are p u b ~ ~ c s information. , financial data, roduct a d v e ~ i s e ~ e n tand s are publication of papers, references, on-line ~ t o ~ a and ls .The ~ ~ ~ risnnote only t able to d ~ s ~ i static b u ~in static i n f o ~ a t i o ncan also be dist~butedin the form of active We ~ependingon information requested, or in pages such as search en in response to ~ueriesfrom the I n t e ~ e user. t gages are pages in which changing data is constantly received. Such pages can contain online music, radio stat~ons,video or real-time data updates.

The Internet allows compiiters to talk to each other via a cable or wireless CO order to allow computers m i n g differearl operating systems to communic l a n ~ u a gor ~ , ~ a n s ~ ~ s sprotocol, ion is ~ e ~ u i rThe e ~ .most comm on the I n t e ~ ei s~TCPIICP. The use of a protocol ensures that a user information on the Internet regardless of the computer, operating rryste ~ n f o ~ a t i oonn the Internet to be universa~~y accessibl vided in a f o ~ a that t can be displayed success~llyan veloped to allow data to be received in a d o c ~ e n t are s plain text d o c u ~ e n t sCO presen~bIelayout [2,3]. allow software to display the text in a ~ o r m a ~ elayout. d For active pages, languages such as J cript or Java allow software to be included in a added interactivi~. s o ~ a r products e used to display Web doc ~rowsersbecause they assist the user in browsing or surfing the I most common Internet browsers are Internet Explorer by M ~ a v i g a t by o ~~etscape.

12.2.3

mat Would

thout the Internet?

met most comgu~erswould be s ~ d a l o n e network. These c o ~ p uwould t ~ only be able to access in the local area network ( ~ A This ~ ~ i n f.o ~ a t i o nwould h themselves or ~ a n s f e ~ efrom d a physical medium suc within the LAPS. It would not be possible to access the latest news or obtain up-to-date a com~uter.~ ~ e n e vaesr o ~ a r ceo m p o n ~on t a comp ed sofhvare version would need to be available at its location. ical medium to be supplied to location of the c o ~ p u t e woul r ional costs when compared providing i n f o ~ a ~ and on s over the Internet. The I n t ~ ~pro~ides et ~ u ~ t i p ~ese of e ~ e c ~ o n i cn f o ~ a t ~ ~ n , ~amuals,m a g a ~ i ~ e~s t, o r ~ ah~ s , s, f r e q ~ e ~ tasked iy qu~st~on on c o ~ ~or~progra~ming t ~ g problems and many more which WO ava~lablc.

Power System Restructurhg and ~ e r e ~ ~ a t i o n

i n d u s has ~ been utilised by the power ~ n d u for s ~s ~ e ~ l i n ~ cing ~roductivity.The best power plants are not the plants with their computers. The best power plants will be the ones w h ~ are c ~using the right IT tools and using them appropriate^^. There are many benefits to the power ~ n d u by s ~accessing the largest resource of IT tools, the Internet and some of them are listed below: e r e ~ l a ~ of ~ oenergy n market formation on power privatisation available for customers esentation of private e ~ e r g ysupply companies ice c ~ m p ~ i s for o n energy custome~ -up to electricity suppli cing supply chain costs ised supply chain by t of remote e - p ~ e r s h i p s tomer relationship management d control for manag~ngpeak demand energy ~ ~ c i n g power systems component monitoring an component control me expert advice for problems which have een expe~encedon other sites n-line c o n s u l ~ c (ye - ~ o w ~ e d gimprov~ng e) ~nowle~~ g e~ a g e ~ e n t . automation for continuous energy supply oni it or in^ to operation, e.g. in case of point f a i l ~ e 1 marketplace in the energy sector floors or NetMarkets for e l e c ~ s~ppliers c ~ ~ ce auctions and negotia~ions between energy d i s ~ b u t o and r supplier Ability of governmental regulators to monitor energy companies on-line rchase o f ~ ~ h i ornspare e ~ parts from a wider range of s u p p ~ i ~ r ~ g of raw materials such as oil, coal or gas -line ~ a r k e ~ ~ acontrol c e s ~nvento~es

for teleco~municat~ons ice provider (ISP) services o f c o ~ o d i t i e and s equipment

~pp~ication of the Inte~etto Power System ~ o n i ~ o rand i ~ Trading g

19

w Can I in^ the I n f o r ~ ~ t I~Need? on In order to find the information required within millions of Web sites a search engine can be used. Most ISPs provides the Internet user with a simple search facility to search by cat ego^ or keyword. The number of Internet search engines is constantly Companies which are provid~ng a free Internet search facility are advertising as a source of income. Search engines are constantly combing amount of accessible Web pages trying to index the information they conta~n.This indexing job is done by a parr of the search engine called a Web crawler. If specific k e ~ o r d sare used for searching through the accessible Web sites, these k e ~ o ~ are ds matched against the index and lists of pages containing the keywords are ~splayed.

Generally, the Intemet can be used to gather or publish information on all t can be c a p ~ e din electronic format. ut finding the right information with quality describes the problem of the usability of the Internet. One of the m with connecting to the Internet is slow technology. If Internet users have slo old computers or old software, the usability of the Internet is not high. Keeping technology updated requires i n v e s ~ e nin t hardware and software upgrades. ~ o ~ e r cusage i a ~of the Internet can only be effective if such investments are met. But there are more parameters which affect the usability of the Internet and which cannot be influenced by i n v e s ~ e non t the client side parameters like: which search engine to use, what keywords will give the best search results and which Web pages contain the information required? ~ ~ e ~ ~ broken links, which are connections b e ~ e e nWeb sites where the target site has been removed, f r a ~ e n t a t i o nand repetitive or duplicated contents will reduce the usabi~~ty. Increasing Internet usability is the ultimate objective for many commercial users. Therefore, keeping a cooperative database of practical keywords, laces of interests and bookmarks on an internal Web page will increase productivity and reduce Internet ~ u ~ n g .

12.3.I

~ ~ i e nUse t ~for ~ ~c e s e u r ~ h e r s

Originally, in addition to the US military effort, universities created the Intemet to share infQrmationon research p r o ~ a ~ eIns other . words, the I n t e ~ itself e ~ has been a research p r o ~ a m m ebetween universities in the USA. With the Internet in place, research p~ojects can be continued where other research projects have stopped. This is p ~ i c u ~true ~ l for y open governmen~land ~ n i v ~projects, s i ~ avoiding duplication of rese private research projects are executed behind closed doors for economic reasons although there are exceptions. esearch software projects which are sponsored by universities or the public sector for evelopment on the frontier of technology are often open source and accessible to ch projects often benefit from the input of hundreds of con~ibutingp r o g r ~ e r s from all over the world. One example of such a collective effort is the Linm operating system, It has been ~ e v e ~ o p by e d an countable number of c Q n ~ i b u tand o ~ ~ a ~into r ~ d a very stable and reliable system. Most importantly, its source code is freely available on the I n ~ e ~ e ~ .

4120

Power System Restrur;turing and

The ~nternetis the ideal medium for publish~ngi n f o ~ a ~ i owithout n h a ~ n to g pay high rates to commercial publishers. Everybody with I n t e ~ eaccess t and Web space c r e s e ~ c hresul~sor join newsgroups to exchange research ~ n f o ~ a t i o nA. ~ ~ a g i n a btopic ~ e is avai~ablewithin the never-ending lists of newsgroups. ~ e w s g r o ~ p s allow researchers to publish and discuss their results with c o ~ p e ~ eaudience. nt Whenever research problems accme be found in the Internet news~oups.Sc query and collaborate with colleagues and access or share s o ~ a r eand i n f o ~ ~ a t i omade n available on remote machines across the I n t e ~ e t .

12.3.2

~ ~ u c u t ~Use ~nal

The Internet can be used to access information on schools, universities, scho~ars~ip9 ~ n by sharin fellows~ips9 and others. It is able to improve inte~ctionsb e ~ e insti~tions i n f o ~ a t i o nabout events, projects, timetables, resources and act~vities,wh~chmay pro the usability of resources, such as sharing transportation or avoiding overcrowding in the local s w i m m ~ gpool. The Internet can even help reduce the ~ ~ i costs n gof schools, e, ent to be bought in bulk and shared by several institution§. als can be made available to students on-line, which saves on material costs, cannot be lost or left at home and allow students to get pr~pared.F ~ h e ~ o rthey e, allow p ~ t e students ~ ~ a to ~ gather more detailed i n f o ~ a t i o nabout a course up for it. There are several on-line training courses avaiIable on the I n ~ e ~ eThey t . allow people who live in ~ e m o locations ~e to continue their education after leaving school. With the help of an on-line tutor, which monitors the progress of students remotely, queries can be sent and answer~dvia e-mail within minutes, On-line exam~ationsand o n - ~ i nmu~tip~e-choice ~ questions generally follow such studies, includ~ngthe publishing o f exam~nationres~lts.

12.3.3

Inte~net

Prior to ~ ~ i a ndecision g on which product to purchase, extensive i n f o ~ a t i o fact sheets and general opinions can be analysed. The I n ~ e ~ ~e nt a ~ lusers es t product or component performance by being able to access dir competing companies. Good starting points for obtaining lists of CO sp~cialise~ on-line magazines, virtual ~xhibitionsor virtual shopping centxs.

usinesses can compete with on-line quotes for services and goods to a ~ c pto s ~ ~ b l e cu~tomers.They can show detailed statistics on their busine5s p e r f o ~ a n c eto a ~ a c t pot~n~ial i n v ~ s ~ and o r ~shareholders, ~usines€escan pu~lishi n f ~ ~ a t i oand n co~~icate via a secure Internet connection and firewalls to improve c o ~ u n ~ c a between ~ o n remote

Ap~li~ation of the Internet to Power System Monitoring and Trading

~ u l t i m ~ dmeans ia the simui~neoususe of more than one medium. A single medium can be text, image, video and sound. ~ ~ i ~ t i m edevices dia are able to play music, animated images, motion p i c ~ r e sand videos. But multimedia technology is not just about playing multiple media, it also includes storing, transmitting and presenting information from multiple sources. Uses for such technology include e n t e ~ i ~ e nvideo t , conferencing, video on demand (VOD), close circuit television (CCTV) and distance learning. These are many different formats in which multimedia contents can be stored. The most common ones found on the Internet are listed in Table 12.1. 2.1 Common multimedia types Category Audio, Sound, Music

Movie, Video Images, Photos

Extension (MIME type) AIF AV1 M3U MID MP3 SND WAV AV1 DV DVD MIV MOV MP2 WE MPEG MPG BMP GIF P E G JPG TIF PGX WMF

This list s ~ a r ~ s only e s a fraction of available file types. There are many more file types for images, sound or movies and new ones emerge constantly. If the Internet browser receives a multimedia file, it identifies its contents by the type, which is related to the file extension. Once the content i s identified, the browser executes somVare to use the file as intended by the originator. In cases where the brow~er does not include software for opening a file type, e.g. MOV (Windows browser invokes a helper appl~cation,e.g. a movie player. For an unknown or a browser plug-in or an external helper application may be required. and ability of the Internet to distribute multimedia content, c ed music and video occur. Without copy protection it is ve convert music or video tracks into an Internet distributable format. But there are obvious advan~gesfor selling music electronically over the Internet, e.g. no record c o ~ p no ~ y ~ i n t ~ ~ e d i a ~low e s cost, , large audience and many more.

Internet on-line services such as Internet banking and account managing, s h o p p ~in~ ~ virtual hopp pin^ malls, live news and trading floors, just to name a few, have become very popular. On-line shop~inghas become very popular for light goods (sma~lpos~agecost) music, videos and books. Several large supermarket chains are trying to push ping for food and groceries by introducing a fixed delivery fee and delivery times. Such business is not time critical and can be accomplished continuous connection to the Internet. Time-c~ticalon-line services, such as trading floors, require a con~inuousconnection to the In~ernetin order to follow and react to market changes. They can only be succes the used IT infhstructure can handle real-time data transmissions and if cont~ngency are in place in case of technical failure.

Power System Restructuring and ~ e r e ~ ~ a ~

2

Trading floors and on-line auctions are a very promising development on the Internet. They allow multiple sources and end users to meet in a c o ~ o virtual n business without verbal c o ~ u n i c a t i o nor travelling.

12.3.7

S~pport for Professionals

~ e ~ e r e n c i ni gn f o ~ a t i o nplays an important part in many professions9 p a ~ i c u l ~ the ly legal, medical, scientific, financial and information technology professions, Internet was estabiished, these professions relied on an extensive amount of published papers, e.g. books, journals and reports. The production and dis~butionof paper reports can be expensive and slow, resulting in i n f o ~ a t i o nbeing unavailab~ewhen it is required or only available to those who can afford it. Since the introduction of the Internet, such i n f o ~ a t i o nis readily available to everybody. People who are ork king in a fast moving e n v i r o ~ e n tsuch as the IT sector require flequeni updates. The I n t e ~ provid~s ~t a medium in which updates can be made available to everybody quickly,and cheaply; therefore, the latest technology and manuals can only be found on the Internet. The Internet user can be hidher own doctor or lawyer. But there is a danger. Using the ~ ~ f o ~ a twithout i o n the necessary experience can sometimes lead to wrong conclus~o~s. This is especially true for self-analysis of illnesses. Some ~ n f o ~ a t i oon n the Internet should only be used for the purpose of improving i n f o ~ t i o nfor a specific group of professionals. It might be useful to know which illness matches the symptoms but a professiona~shou~dcompile the final conclusion. ~nternet~based business analysis solutions for the utility marke~laceprovide utilities and e n companies ~ ~ with~large-scale sop~isticatedanalyses of their data and allow them to extend access to these analyses to a larger number of users. As a result, users will have -click access to energy and power plant data to s u p p o ~their energy ~ a d and ~ g t management and for improved business decision-making capabilit~es. n for the power Internet forums have been successfully used as an ~ n f o ~ t i osource utility industry with regards to IT-related questions, such as the year 2000 (Y2K) issues. There is a vast amount of power utility related data already available on the ~ n t e ~but et some of it is poorly organised and difficult to find. Energy i n f o ~ a t i o ncompanies a d ~ e s s this shortfall because they specialise in the collection of information related to power utility research. ~ n ~ companies e ~ e ~ have created virtual conference room, where visitors can review case studies, survey results, white papers, reports and studies, meet with staff consul~nts, and p a ~ c i p a t ein an on-line survey.

12.3.8

The Power Industry and the Infernet

Can the Internet really stand up to its promise to increase productivity and p r o ~ ~ ~ i and lity ost for the power industry? is not much evidence to support this s ~ ~ m e nbut t 9with connection ownership, this will change in the near future, to increase productivity by using the Internet to find the right ~ ~ ~ o ~easily a t ~ando n quickly and to attract ~otentialclients to ~ommercia~ Web pages to in~rease~ r o f i ~ b ~ l i ~ .

~pp~ieation of the Internet to Power System

nitoring and Trading

23

~ f ~ c i eWeb n t page desi should start from the need of the clients (the c o r n find the ~ f o ~ a t i or o nproducts they require ( i n ~ e ~ e d i acontent) te and move to t (e-commerce tool). Making money via e-commerce requires the real business on-line. ~ ~ c r e a s ~ n 1 barrier to purchase goods on-line, is one of the than just i n d u s ~news. It ices tailored specifically to the power i n f o ~ a t i o nto the average I ~ t e ~ e t professiona~~ to satisfy a need, or to solve a problem. For example, low optimis~ngthe c o ~ b u s t ~ oprocess n or reducing shipment delivery tim shipload c a p a c i ~using other Web sites. ~ n c r e a s e~~a f f i cto ~ e pages b will increase its p o ~ u and ~ at h ~ e ~f o r eits value. To t visitors, the Web page should contain added conte increase the rate of rec Web sites should provide easy access to e ~ e c ~ i cia l erspec~ive.~ o w e rin ~ i s ~ b u t Qcon~ac~ors, rs~ engineers~purcha$~n r e s ~ ~ c for e s mmu er electrical industry ~ ~ o f e s s i o ~ a l s . s the ~ n t e ~ can e t support the core c o ~ p e t i rgy-trading p l a ~ can ~ obe~ designe~to receive orga~~~sations and use these biddoffers to sched to the ark et inch resources locally, A d ~ i t i o ~~anl c t i o n provided § data, s e ~ l e ~ e nbilling t, and p ~ b l i ~ h iof~ gpricing and trading i ~ f o ~ ~ t i o n i ~ f o ~ ~portal. t ~ o Such n a platform can be designed to give market ~ a r t i c i F ~ ~ t to access the m ~ k e24 t hours a day, seven days a week. ~ t i l i ~ ies are developing e l e c ~ o n ~ c Internet. In illing will reduce the utility CO on-~ines e ~ i c efor r ~ s ~ d e n t ~ a ~ and their ~ s w e r Qs utility c ~ m p a n and ~ ~ s1 les, This will allow po offer a c ~ ~ n ~ cst o r s ~ ~ t c Qe ~ e~c t ~ i t wiring. ing l n t e ~ etechno~ogy t adv iders (ASPS). The idea is to provide so is instalIed on a rem ter with ~ n ~ e access. ~ e t This avoids the probl and ~aintainedor u p ~ a ~ e d ess to expensive and highly speci the standby time and cost of , for ~ a i n i n g~ u ~ o s can e s be lea I" of people a ~ e n ~ This ~ g is. one

12.3.9

Recent ~ ~ p r o v e m eon n t the ~ Internet

Since the computer ~ a n u f a c ~ r i nindustry g and the Internet are ~irectlyrelated, faster and cheaper computers will constantly cause expansion of the Internet networ~.This, in itself, is a positive deve~opmen~ as long as the transpo~tionnetwork is at least expa~dedat the same rate. Therefore, a constant improve men^ of the Infernet i n f r ~ s t ~ cis~ nr e c e s s a ~to ensure a continuous quality of service. With more and more users using and publishing information on the Internet, more and more information becomes available. It is c e ~ a i n ~not y the case that every Web site on the Internet on a specific topic contains valuable i n f o ~ a ~ i o n§ometimes . i n f o ~ a ~ i oisn duplica~e~ or even wrong. This makes it sometimes difficult to find quality contenf for serious researchers without wasting time visiting ~ i ~ e r eWeb n t sites con~ainingeffec~ively similar information. ~~)nsequently~ the more iiifo~ationthere is on the lnterner the more diluted the quality of content on a common topic and the more difficult it becomes to find quality content. But the Internet is constantly i~provingits search engine$, which are now using iiifomiafiorr-refiningprocesses with artificial intelligence (AI). The processing power of PCs has been doubling almost every two to three years and with the new generation of niultimedia extended processors (MMX), Web p include images, sound or video clips. This development has increased the Internet sur-fig, sometinies caused just by the graphicai design and ~ n c ~ i o n a ~ i ~ ( i n ~ e r a c ~ i vand i ~ )nOQ the content of a Web site. People visiting interactive ~ e sites b can interact with their contents for fun or for business ~ ~ ~ o sThe e sInt . improved the appearance and i n t e r a c ~ ~of v iWeb ~ sites and has cre ortant issue su~oundingthe Inte et. A system w h i c ~is not us business. When the I ~ t e WRS ~ ee ~ s ~ a b ~ i s hsee~~, u rwas i~ rs did not intend to use it fo esses started to e m e r ~ eon need for securily arose. There are several reasons why h c o ~ p u t eSome ~ . o f the most common ones are clienl/se A simple securi~y-rel~ted bug in browsers can allow ~ n f o ~ a t i oWith n . the latest 128 bit en rotocol, hackers will have at least ion, ifit is within their ab~~ity. the Int~rnetwhile on the move is one o f th e~ ~~ o ~ra bhand-~eld l~ e . d e v ~ such ~ ~ as s w r i s ~ ~ ~are c ~c ue ~s e n t ~available y for accessing the ~ n t e ~ ~ ~ .

Currently, the PC is still the most common way to access the K ~ t ~ ~ e ~ . few years, Internet access will be drama~icallyincre TV set-top boxes, game consoles or video teleph ~ n ~browser ~ ~ ene tling I n t e ~ eaccess t at any location. WAP or ~ ~ ~ e r r i e t - e n~elephone$ ~ b l e ~ are cu ntly being pushed as a far, there is still a lot of convincing and iiii~roveme~~ts to be done until WAP phones will

~ p ~ l i c ~ toE i othe n Internet to Power System

onitoring and Trading

take a s e ~ om~asr ~ eshare. t They suffer from b ~ d w i d t hres~ictionsand small dis Pal~topsin c o n ~ a slook t quite promis~n~, since thek display is reasonabIy idth restrictions. C ~ ~ e n t l~y , A phones P and palmto Ie to mobile phone text ~ e s s a g i n ~ ~ omising technologies for accessing the ~ n t e ~ine the t ~~e are as Internet display units, if connected to the ~ n t e ~via e t a teleph c o ~ ~ c t et-top i o ~ b ~ will~ sit eb e ~~e e nthe TV and e teiep~onec o ~ e c t i o nand will eliver services s ~ ascdig~ TV or video on demand. game consoles in the late 1990s, game consoles are a ~ o ~ ~ e r e T~ternet.They need to be equipped with a s ~ ~teci l ~ require a telephone c o ~ e c t i o n Already . some et connection from game consoles for offer full Internet browsing. already have a display, ted to a ~ e l e ~ line ~ o and ~ e keyboards and the abWy to ady. They also have l o ~ a t of ~~n far the costs of o ~ e r s h arid ~pm lds (e.g. in fridge doors) have se to some respect. 12.4.I

Access. to the ~ ~ t e r ~ e ~

Access to the ~ n ~ e ~ n ~ t

Pi

or s~all/medi~m"size connection to an ISP. a ~ ~ e stos ,the ~ n t ~ ~ e f a fast connect~onwith all i t s us ternet, making it the ~ ~ ~ or middle ~ e

-8. The Internet as a three-tier connection

d

Power System ~

426

e and ~ e r e ~~l a t i o n

The Internet is building on a client-server rela~ionship model, where the client’s Internet browser connects to an Internet Web server. For the browser to an operating §ystem or platform needs to be ins~al~ed on the client’s PG. to Web servers. Operating platforms supply the basic structure of the computer environ~entsuch as convenient access to all p e ~ p h edevices ~l installed on the computer. mile it is possible to create an Internet browser for a specific computer hardware layout, the m ~ ~ of~ computer ~ d ehardware combinations would require a browser for every possible option, Therefore computer software is generally written for operasing systems. The most comMo~o~eratingsystems for client n i a c h ~ e sare icrosoft’s Windows, Linux and Apple ~ a c ~ n t o s hMacOs. ’s The more applications these are available for any platform, the more popular this latform becomes, Therefore, most home or o ~ i c e - ~ a computers se~ will have one of the ~reviouslymentioned operating systems. servers have different criteria for choosing the ~ l ~ t onf which o ~ they reside. ad applica~~on support is necessary €or home and office CO r e l ~ ~ b iare l i ~the major criteria for Web servers. O p e r ~ ~ i nplat g Solaris, W i n d o ~ sNT and Wewlett Packavd are focused on secure access res~ction$and secure ~ e m management. o ~ Access res~ctionsinco orate ~ ~ l ~ i - u scapabilities er nt levels of access, e.g. Web users cai only acces g. Secure memory management incorporates nt levels of memory access, e.g. every op ning in its own address space and will not conflict with other pro~rams system in case it crashes. b allow access issue for Web servers is s e c u ~ i ~ , ile ~ e servers ody, access res~ctionsapply to all other areas on the serv~r’shard disk.

12.4.3

Web Clients

client is a piece of software that is able to receive ~ f o ~ a t i for o ndis p u ~ o s e s Web , clients are used to access information pub~ishedon an Internet-enabled Web server via a URL. Web clients do not exist in isolation since ey have to access a server for i ~ f o ~ a tretrieval. i o ~ They are part of the clie~t-serverMO 1, ~ h i c is hs h in ~ 12.2. The Internet utilised the clien~-servermodel because of nature.

are 12.2 The client-servermodel

The most common Web clients used for displaying i n ~ o ~ ~ a ton i o na c browsers such as Internet Explorer or Netscape ~ ~ v i g a t oWeb r , cl~entsfor

~

Application ofthe Internet to Power System Monitoring and Trading

427

info~a~~ storage o n can be found in Internet search engines. A Web client must be able to understand the format of the remote i n f o ~ a t i o naccessed for successful p r o c ~ i n gIf, . for example, a Web site containing Chinese writing is accessed, the browser must have the reqLiired fonts installed. If rcal- me data should be displayed in a Web client, the client has to have the capabili~to receive data updates and display changes ~ ~ c o r d ~ n gThe ly. m u l t i ~ d eof data types and constantly emerging new technologies and standards forces companies building such clients to release frequent updates. Users of Web clients should always try to update client s o f ~ a r in e order to access new Internet tecIino~ogies.

12.4.4

Web Sewers

Accessible UIpLs must be located on a dedicated Web server. Only Web servers which are enabled for ~nternetaccess are accessible by the Internet user. Basically, a Web server is a computer with Web server software such as Apache Web Server, Internet Information Server (IIS), Personal Web Server (PWS) or any other Web server software. This software allows other computers to connect to a specific port (normally port 80) and display the contents via a Web browser. 12.4.5

Web Protocols

Web servers are able to understand several protocols. A protocol is a method computers use to communicate with each other. There are several types of protocols. Different types of protocols are required for different tasks, e.g. Web page access or file transfer, The most common protocol used over the Internet is a combined protocol called part is rcsponsible for the c o ~ u n i c a ~ i oand n the IP part is required for identi~cationof computers. In order to address uniquely any Web semer on the Internet, a unique token is required. This has been realised with telephone numbers in mind. Therefore, a Web server can be addressed by a set of numbers, the TP number. It can be used in the browser as a hexadecimal, octal or decimal number. Its most common appearance is decimal and it looks like this: 123.456.789.012. Since such numbers are difficuit to remember, a more friendly way has been d ~ ~ c l o p e d , called a domain name. The domain name allows the use of friendly names such as h ~ : / / ~ . w i l e y . c o instead m of 199.171.201.14. Good domain names are limited and most of them have already been occupied. Some of them are available on the t ark et for bidding, which i s very similar to personalised car number plates. Recent court rulings have tried to discourage domain name hogging by forcing individuals to release branded and trademarked company domain names so that the companies can represent themselves on the Internet without paying millions of dollars. 12.4.6

E-~~il

The Internet owes parts of its pop~~arity to the e-mail system. E-mail i s an electronic means of sending a message from one computer to another in an organised fashion. E-mail services are offered by an ISP. Mail accounts can be created from ISP e-mail providers such as CompuServe or AOL. E-mail is the fastest and cheapest way of sending messages

Power System Restructuring and ~ e r e g u l a t i o ~

to any location in the world. There are specific protocols for sending and receiving e-mail messages. The protocol used to send e-mail messages across the ~ ~ t eis~~ he et’ ~ i m p ~ e rotocol ( S ~ T ~The ) . protocol used to receive e-mail messag~sis the Post Office versions of these protocols have been improved in robustness and P2 or POP3. If e-mail c o n ~ i n smore than just text, e.g. a ~ ~ c h m e n tanother s, ~ r o ~ o cisorequ~red. ~ llows do~nloadingor uploading of files on remote machines and is called rotocol (FTP). It is a~~omatically invoked if an ~ ~ a c h r n ei sncopied ~ to a hard disk. If, for i ~ s t a n e ethe ~ graphics adapter driver software requires upda~ing,i.t is more than likely that it is available on the Internet. Generally, there will be more than one location, called FTP site, for ~ownload~ng. The most used protocof for Web browsing is the ?TP.This protocol carries ~ n f o about ~ ~the~originator ~ ~ nof the i n f o ~ a t i o nand the information itself It is able to tell the b r o ~ s e of r which type (e.g. plain text or cQ~pressed) and f o ~ a(e.g. t ~ T ~JSP, L ¶ ) the ~ n ~ o ~ isa so~ that ~ othen browser can play it correctly. Free ~ I i t e ~ e ~ - b aesed mail sewices are available, e. g. from HotMail o

62.4.7

Internet Security

I n t ~ ~~eetc ~ r i is t yn e c e s s a ~to protect cornpurer resources against the risks and threats that arise as a result of a connection to the Internet. esign of the Znte~etoriginated from the idea of cQnnec~ingcomputers b e ~ e e n s etc. f i r com~unicationand owle edge-shar~ng purposes. There was no reason for a n ~ ~ to o consider ~ y s~botagingthe connectio~~, since only a selection of trustwo people h sical access to the computers connected to the I n t e ~ e t sharing ¶ se~isitive rese~ch ation. Tilerefore, security issues were not part of the i n i t ~ a~~ ~design. t Since more and more users have access to the Internet and its utilisation for business and ~ ~ ~~ansactions c ~ a has i grown, Internet security has become a r i m a ~concern. The rcasons for ~xplQiting or sabotaging the Internet are man~fold. One of the major security concerns is caused by the fact that data is: transported as text, allQwi~g easy access for third parties. This risk is mos acceptable for non~§en nce the Z n ~ is~a ~ ~ t ss these lines if avo st option for se~sitived afford a ~ o n ~ ~ n u ocable us

home workers to have access to sensitive eompany data from any location. These r e q u i r e ~ ~ ~have n t s persuaded many companies to open up their private Intranet to connect to the public Internet. Protec~~nga private network and shielding it from h~ckers without restrictin commun~catio~ to remote users can be achieved with a firewall. A firewail w r ~ ~ to e ncombat unauthoris~access to files or uiider~y~no p e r ~ ~ ~ systems. ng on the company policies, only selected services are granted access to the outside world. Figure 12.3 shows how an Intranet can be protected with a firewall. Local computers are able to connect to each other and to the Internet, but remote coinputers with Internet access

e

Power System R e s ~ ~ c ~ rand i n ~g e r ~ ~ l a t i o n

43

recipient needs to receive the private key, which can be intercepted. Private key generators produce only one key (A) for encryp~ionand decryption of data.

! Trans~ission

1

across the

I Internet

nusing a private ~ key ~

Data ~

t

~

~

~

The alternative is to use a private/public key pair. In this case, a message c e n ~ ~ t with e d the public key, but only d e c ~ t with ~ d the ~ ~ i v a key. t e This allows p u ~ l i s h i nthe ~ public key to many people, who are able to send ~ e s s a g e sback to the e the publisher of the public key is the hing the public key, the ~ e s s a g can ~sb blic key, only the p ~ v a t ekey can dec public and private key lic key generators always pro where A i s used for encryption and B for

I

Secret readable data

I

I ~rans~ission

I f

I

across the Internet

~ e c r y p t ~ owith n private key

.5 Data encryption using a public and private key pair

result irm using less server re sour^^$ so t

age is l~adedfaster. Th aved, e.g. the use of c s ~ n t e r c h Format ~ g ~ (GIF). The GIB; ir image ~ompressionratio. One of th

o p ~ balanc~ ~ ~ aof ~image quality a

Application of the Internet to Power System Monitoring md Trading

Server 1

Server 2

Server 3

Network load balancing with three servers

ua

33

software is ~ m p o ~ a nThe t . answer is Java. Java has been develo~edwith the Internet in mind. It is not exactly i n ~ e ~ r e t eordn o ~ - i n t e ~ r e t ebut d ~s o ~ e w h e in r ~the m i d ~ l e ~ because the source program code is compiled into byte-code, process. Java byte-code is i n ~ e ~ r e by ~ eadJava Virtual Machin p r o ~ e s s o r ~ s ~ien~si ~ c~ ~ i during o n s run-time. This mechani different platforms if an a p p r o p ~ J~ ~ e g languages are enabling Web pages are i ~ p foro on-line ~ ~ and re ~ntgractiveWeb pages are required if feedback from the Web user is relevant.

12.5.3

at Is ~ a v a S c r ~ ~ ? a p r o ~ ~ i n i ~n ga n ~ a which g e is exe ages that provide a means of adding As shown in Figure 12.9, JavaScrip

in the Web ~rowser.It is one of

age from a Web server, the browser inte~retsthe JavaSc~~pt code for page i n t e r a c t ~ v If, ~ ~for . $elect~b~e list of products and ces, ~ a v a ~ c r i pcan t keep track of the running of all selected from the list. Such interact~vi~ cannot be accomplis~edwith basic cript v e ~ s a t i i allows i~ Web pages to be created without JavaScript does not necessarily need to be emb eb pages in JavaSc~iptmight defeat the ML pages by adding interactivity inste Computer 1

nment for JavaScript applications

de in J a v a S c ~ is ~ tvery similar to writ in^ code for a Java is how events for executing code sections are trigg objects, e.g. a button, to trigger code, which might ~ ~ l ~ u laasubtot te d show the result in a ~ o p - u pmessage window. whenever unknown p r o ~ a m m ~code g is executed ca~tionmust be ta Internet browser executes the JavaScript code in an encapsulated env~ronmen~~ preventing access to system reso~rces,e.g. the hard disk. ~ h e o r e ~ ~ cita lshould ~ y be just as safe to

~ p ~ ~ i c aof~ the i o 1n ; i t ~ ~toe Power t System

onitoring and Trading

as it is to execute applets. ut holes have been f o ~ ind some browsers’ Java security ~ ~ p l e m e n ~ ~allowing i Q n s cleverly written JavaScript code to access files with own location and name. Nevertheless, the execution of Java code in et the Internet browser can be turned off when brows in^ JavaScript or an a ~ p ~ within u n ~ s t e Web d sites.

opment p r o ~ a m i ~ i n1 g

creation of a Java source file, this file can be compiled into Java b ~ e - c o d evia a Java compiler as shown in Figure 12.10. ptain text Java instructions

Execution of bytecode in Java Virtual Machine (JVM)

“Prograrn,class”

2.10 Java code com~iiat~oii and execution

Java c Q ~ p i lcan e ~ ~ o ~ l o a d from e d the n ~ ~from e tvcarious 1 numero~sc ~ ~ e r c Java ~ a ldevelop~entplatforms on the m ~ k e t swhich all ~ e v e l o ~ ~ ae n t er ~ e b u g ~ ~than n g n o n - c ~ ~ ~ e r cones. i a l Since Java in~oducedby icrosystems, it i s one of the most reliable sources for tutorials, ~ o ~ ~ p i land e r sother Java resources. It can be accessed via the h ~ : /.jav/a.sun. ~ corn. Java byte-code can be executed on any different computers. There ing ~ a n ~ a Any g ~ c. o ~ p u t e rwhich has a Java le to execute java byte-code. This means that softwar and compil~donce. This is a real advantage ~ s i ~ n ewritten d, terms of dis~ibutionon the Internet.

E n v i r o ~ e nfor~ a Java ~pplication

Power System ~ ~ s t ~and e ~ n g

of which have a

, thus ~ a ~ i itn safe g to mn a p p ~ ~ ~ s can p e r f o ~if it is live on the Ente w h i it~was ~ ~ o ~ ~ ~ d .

~

n

~

~for ar Java ~ applet ~ ~

n

t

shows the e n v i r o ~ n

A p ~ l i c a ~ ~ofo the n Internet to Power System Monitoring and Trading

access to c o n ~ ~ e n t i a ~

be seen as a web if lines ~ e p ~ e s e ple who are §pending time on th noth he^^ are browsing or surfing the Web.

7

Power System ~ e s ~ ~and~ ~ i n g

438

all parts are combined into a single H L page. The static part can include ~eneral~ n ~ o ~ aand ~ i logos. o n The dynamic part can be a table where data is q u e ~ e a da~base,formatted and enclosed by ~~~L tags. Such dynam~ccreation of WT can be achieved via a CGI or servlets. Therefore it is not unlikely that initially being d as a set of templates with the conten~sadded via creation, ~ e n ~ r a t i o a dynamic Web page is illustrate

ases are the most used dat ases are categorised s are hiera~chical,relat r e e f f ~ ~how t s the the data is stored and ~ e r e ~ othis ~ n c ~ i o ncan s access the data. Examples of a hi ata files as shown in Figure 12.14 [6,7]. ~~t~ h i ~ ~ ~ c~ ~a i t caa ~~ data a ~is stored ~ ~ , in a tr data can be found near the root

es t ~ b j ~ c t - o ~ i e n(00) t e d ~ ~ t a b a scombine of i n f o ~ a t i o ns ~ o r in e ~the table. For ex ction could be the calculation of moving averages.

g l ~ n g u a gGalled ~ ~ ~ c ~ r e d ‘seque~’~. SQL is imple~ented s t a n ~ ~said d to be the ~ ~ t a ti si not o ~ always guaranteed ~ ~to a ~ ~a ofor w ~ ~~ a ~spe ~r e s s ‘

Application of the lnternet to o w a System Monitoring and Trading

.14 Most common database types

nce the SQL q ~ e r yhas been defined and coded, it needs to be sent to the for execution. Database vendors have their own version and i r n ~ l e r n e ~ ~ t of ion their database manager and query optimisers. Therefore, a common c r o s s - p l a ~ ~ o ~ database connectivity standard for Java has been introduced called Java database connectivity (JDBC). JDBC drivers have been developed from JDBC’s pred~ce§§or, ODBC, and are available for almost every database. JDBC comes in different levels of d a t ~ b ~ ac~essibili~. se For examp~e,Level 1 JDBC drivers a C bridges for databases where only an ODBC driver exists and Level 4 can access a database directly and are generally written in pure Java. When the database and the SQL application reside on the same computer, and no server exists, the database model is called a two-tier model with the first tier being the ap~lica~ion and the second tier the database as shown Figure 12.15 .

JDBC driver via LAN

Two-tier JDBC driver connection

Power System R ~ ~ ~ ~andc Deregulation ~ r i ~ g If the database is located on a server the appiicat~onaccesses the database via server software. Such server software can be accessed via an ordinary http request. IVcan be written in any CGI executable language, e.g. Per1 or C*, or as a pure Java application, e.g. servlets or EJBs). ~erver-s~de soflware generally contains parts of the business logic of the database. Business logic is, for example, pre-programmed SQL methods for accessing a database or invoking transaction scripts as shown in Figure 12.16. In most Web applications the third tier is to be regarded as the connection to the database, since applications cannot be granted direct access to the database across the Intemet for reasons of security. Therefore, whenever a database is accessed across the Internet, an appropriate CGI, a servlet or EJBs must be coded. There are several software companies creating ‘off-the-shelf client-server software for data presentation for the client and database access on the server. ........ Client ............................ ............................... ~

f

SQL queries sent to servlet driver via WAN .....

/

I

Figure $2.16 Threetier JDBG driver connection

12.6.4

Web Pages with Functionaliiy

Web pages can include knctionality, e.g. collecting data typed in by users and its validation using JavaScript or VBScript. DHTLM is a collective description of mixing the ~nc~ionality of a scripting language with Web page interactivity.

1

~ p p l i ~ ~ tof i othe n Internet to Power System ~ o ~ ~ i t o rmd i n gTrading

12.6.5

Web Pages with Integrated A

~

~

~

i

~

a

~

i

~

~

~

A new trend into leased ~p~lications on the Internet can be noted. Many small CO which c m o t altlford to develop an application on ltaeir own, have used es for e ” ~ o ~ e rfor c eshopp~ng ~ on the I n t e ~ eand t trad

.7 This section aims to give readers who are not Web developers a quick b a c ~ ~ r o u n extended Markup Language (X ). XML is primarily used to define wit atting i n f o ~ a ~ o n . IS the most suppo~edf o ~ a t t i n gl a n ~ a g eby browsers on lacks e x ~ ~ s i b i since l i ~ , the tags which are used must be defined within the the race began b e ~ e e nmajor Web browser m a n u f a c ~ r e ~style , defin ~ n ~ Q d uasc ~a matter d of competitive advantage, L weakness is that the tags are used for formatting and only little about what the information is. XML can describe the stored i n f o ~ a t i o nclearly,

12.7.1 The s h o ~ c o m in i ~H~ ~ haw accelerated the introduction of X L. One ofthe major is that XML does not contain tags which relate to if~erencesb e ~ e e nTITML and X OCUmentS is assign the f o r r n a ~ i ~ofg the data. The forrnatt a constant format data elements, represent~gspecific e document. By i n ~ o d u c ~ nnew g data e entire ~ d u s ~ i are e s able to interchange i n f o ~ a t i o nin a suitable format. Since ocuments contain the data elements, a new type of document is re ion about its repres~ntation.Such documents are called stylesbe in~o~at~ will o nbe given later. Stylesheets can change the way X i n ~ to be changed, only the stylesheet requires ~odificat~on, browser, If the ~ o ~ a t tneeds separat~ngthe maintenance between data and f o ~ a t t i n gor content and layout.

12.7.2

Reasons for

document i s accessed, a plain text editor can be used to access the data, d y be able to readlwrite the files. Plain a d v a n ~ gis~that in years to come e v e ~ b ~ will files are pla~forrnand application independent. This means that it is not n ~ e s s to a ~use files have been created in order to read the ~ n f o ~ a t ~ o n . is conversion can be saved if data creation s document file created in the 2980s word^^ or rd 2.0) 50 years later. This gives XML a truly universal and timeless data s cr~ss-platforrndata ~ c ~ i v i and n g com~atibi~i~y problems,

Power System Restructuring and ~ ~ r e ~ u l a t i o ~

42

split by type of data and subsequen~lydisplayed depending on meaning, This is because different parts of the data can be identified which enables different applications to utilise it in different ways, e.g. searching or summarising. A data element starts with a tag describing the meaning of the data, e.g. , and ends with a terminating tag, e.g. 4NAME>. XML data smctured in such a way is referred to as being well farmed. ~ ~ ~f~ Q t~ ~a~a g : ~ ~ a ~ data is presented in a hierarchical format. Hierarchical formats have the advantage of faster drill-down for more specialised information or move-up for more generalised information.One of the major disadvantages is that they suffer from data duplication. r

~

~

r

i

~

i

~

~

~

~

~

~

~

~

XML data can be formatted for display by using a stylesheet. Stylesheets define how a specific element is displayed, e.g. on a screen or printer. This enables the user to reuse the XML data for different views or presentations by applying different stylesheets. As well as displaying XML data, stylesheets can be used to convert XML data into different formats such as LaTeX or PDF.

Inline ~ ~ ~ ~ t Q n g s ~ XML allows the inclusion of other files containing XML. This results in manageable chunks of XML data. Files containing XML data chunks can then be included in one or more XML documents, reducing the amount o f data duplication. s~i}~pi~~ allows users to define a tag set of their own. Some rules with regards to its layout are Iisted below: L requires one Iarge container element, which encapsulates sub-elements. All open tags must have a corresponding closing tag, e.g. GIHi>. All sub-elements within a hierarchy must be closed in reverse order. Outer elements containing sub-elements can only be closed if all sub-elements belong to the outer element are closed, e.g.

. Attribute values for tags must be in quotes, e.g.

. The same data can be formatted in different ways by introducing different ways of represen~ingelements. Once the data has been generated in we~i-forma~ed XML, it can be reused by different industries.

12.7.3

Separation of Content and Layout

Information contained in static Web pages may change form time to time, challenging Web page designers for fast and reliable update mechanisms. M a i n ~ i n ~ nthe g fl~xibilityof static Web pages is therefore one of the major design issues driving the in~oductionof new strategies and technologies. HTML pages contain content and layout within one document. Content is the information displayed on an HTML page; it can be in the form of pXain text, tables, charts, graphics or others. Layout is the presentation of the HTML page; it is embedded as HTML markup tags and is not explicitly displayed to the viewer since the browser translates the inarkup tags into positioning information.

3

Application of the Internet to Power System Monitoring and Trading

Classic HTML pages contain both content and layout in the same file, causing dif~cu~ties since common layout needs to be replicated for all pages if changes are needed. For example, if % large company changes the layout of its Web pages a modification of each Web page is required if they were written in static HTML markup l a n ~ a ~This e* problem can be avoided if content and layout are separated. The t e c ~ o l used o ~ for the separation could be achieved with XML for content and extended Stylesheet Lan (XSL) for layout. Figure 12.17 shows the relationship. Details on XSL will be given later. By separating content from layout, Web design can be split among specialised teams such as graphical experts, script programmers and site managers. This allows each component to be reused and versioned, reducing maintenance complexity.

Content Repository

3 SQL query objects

SQL query objects

.I7 Rendering of XML data with an XSL stylesheet for NTML display

~ e p e n d on ~ gthe XSL stylesheet ou ut formats such as WebTV, WAP, PDF or others could be c o n § ~ c t ~Since d . XSL stylesheets are in principle XML documents, they can be converted by another XSL stylesheet into a new XSL stylesheet as shown in Figwe 12.18.

Change management system based on XSL stylesheets

Power System ~ e $ ~ ~ c t u r and i n E~ e r ~ ~ l a t i ~ n

44

and XSL can be transfe~edbetween mul~ple~ l a ~ f o ~ s , ming languages. This protects the technology inves stored in plain text and therefore always accessible. This m proof and allows new ~ e r g ~ tne gc ~ ~ l Q g itoe srely on a simple but co~prehens~ve data structure. ~ a y Voa l~i ~ ~ a t i owith ~ DTD

e definition (DTD) file c o n t a ~ ~layout s ru validate XML to avoid invalid or inc XML files require to be well formed so d o c ~ e n contains t rules, with w ~ c an h order to pass a DT validation test. Such rvles con parameter data types, andatory or optional tags or data. u~ilisationcan assist error analysis within an XML dQcument.

12.7.5

~~lesheet~

se of a stylesheet is to display X L data in a format specified by the stylesheet. allow the same XML data to be d i s p l a ~ e din di€ferentways. Since a data view is d ~ p e n d e ~on t which information about the data is requ~ed,s t y l ~ s h e eoffer ~ an enormous flexibility not matched by RTML. If an HTML do in table format, and a transposed view is required, a new HTML creat~dand this will caus ation of data. If the data is documents need to be up using a set of s avoided. Regardless of which view is required by the us o c u ~ e n tcQntaining the data. This ~ a n g e m e n toffeps since only one data source is involved. If the user nee n be used. One XML document can be f o ~ a in~many e ~ different ways just s of stylesheets, which can be used with the most common browsers. The WOmost frequently used are cascading stylesheets ((3s style 1ang~Iage(XSL). Stylesheets allow ~odificationof thousand^ of concu~entlyand consistently, and this makes the r e d e s i ~of Web sites much simpl~r.

font styles to XML elements. its a ~ i b u t e ssuch , as name, weight, sizes, f o r e g r o ~ dcolour, backg ~ ~ a g r a spacing, ph and many more. CSS stylesheets were introduc means o f extending the style properties of L has a set of pre-defined eleme t contains the same element n will result which might not result in the

headings. If a CS§

Application of the Internet to Power System

nitoring and Trading

~ have any pre-de~nedeie ents in a browser, will not e x ~ e r i e n cthese

SS gets its name from the fact, that the stylesheets can be cascaded. This m e ~ that n~ more than one stylesheet can be applied to a data source.

d o ~ u ~ eand n t format it for the purpose of creating a static ~T~ on the Inkernet.

document for publi~hing

links to another document. This is acco~plish define how individual parts of a d o c u ~ e n are t h ,XSL or XLL documents. L hy~erlinkswith the difference that are more ~ e x ~ b ~ e . ow a connection to entire documents inks allow a more ts, XLinks allow mu~ti-direct~ona~ 1 the links allow running in more than one direction. They allow every element to become a link not just pre-defined elements. ~ P o ~ t e allow r s links to arbitrary positions in an XML document, a re~erencing,footnotes, end notes, interlinked data, connections between parts of remote documen~and other more complex document nav d to link by reference rather than by exact location g a series of reiationsh~~s among information held in allows ~iiipoin~ed links to other XML documents. XLL links c L since they allow for one-to"many, giving user more choice.

The increasing complexity of large electric power systems has resulted in a greater need for ~ ~ ~ n t e to n ~ensure c e a reliable supply of power. ~ o n d i t ~ o ~ - b a smain e d ten^ d ~ § ~ b u t on-line ed HV condition ~ o n ~ ~ have o ~ been n g the current trend. In Non with the construction of m internaeional irport, new power substations have been built to meet the huge energy d e m a n ~The ~ capacity of the existing distributed mon~tor~ng system, which is based on one-to-one ~ o ~ u n i c a ~was i o considered ~, i n a ~ e ~ uand ~ t et~ereforea completely new design concept was tried. The schematic block diagram of the new~y developed system is shown in Figure 12.19.

Power System Restructuring and Deregulation

446

12.8.I

R ~ q u ~ r e m o~fnAirport ~ s Stibstation

An international airport, currently the largest in Southeast Asia, was constructed and was opened in 1998, A number of electric power substations for the new terminal building and associated i n ~ a s t ~ chave ~ r ebeen constructed. A detailed study into one of the numerous substations revealed the shortfalls of the existing distributed on-line monitoring system because the substation there had been too remote from the maintenance centres. The engineers in charge of the transmission network in China Light & Power Company Ltd (CLP) very often need to know not only the real-time status of power equipment but also the security and fire safety of the substation. Furthermore, in consideration of a more efficient operation of the system in the future, personnel in other organisations, such as the Airport Authority, Fire Services Department and other operation and maintenance departments within CLP, may need to gain access simultaneously to the important information within the substation. The original information system needed to be enhanced and extended to tackle the fire safety and security requirements. Therefore, the idea of remote vision for substation monitoring has been employed. This enabled engineers and relevant staff to sec on their remote display monitors the real-time scene of the indoor environment of the substation at different office locations or at home during standby duty. Intruders and fire outbreak in terms of smoke emissions can be detected immediately. To allow simultaneous access to information by all parties concerned, the old method of using modem-based peer-to-peer communication has been abolished and replaced with an In~e~et-based client-server concept. MC

- Micro Controller

PC - Personal Computer Cap.

I

[email protected] The whole Internet-based monitoring system

- Capacitor

I

Application of the Internet to Power System ~ o n i ~ o and ~ nTrading g

The substations, though having great impact on the integrity and normal whole airport, are normally unmanned. Existing substations are equipped panels that retrieve signals from smoke and heat detectors. False alarms are fr~quently encountered and this leads to wasting resources as the fire services are only able to discriminate them when they arrive at the remote sites, Illegal intruders must be and prohibited from entering such substations at any time. To accomplish mentioned above, a remote vision system was developed.

te V i $ ~ ~ n V cameras are installed at different locations in e off-the-sh Figure 12.20 shows the structural schematic diagram of the remote vision sys is to cover all internal areas as completely as possible. For example, the eight locations of the airport substations being monitored are the fire panel, control roam, 11 kV switchgear room, 132 kV switchgear room, substation entrance, 132/11 kV transformer bay, cable basement 1 and cable basement 2. Each camera is equipped with the functions o f zooming and tilting. The video signal from each camera is wired back to a tai ‘remote control and multiplexing box’. The on-site PC controls each box via the prhter port. Through this box, the lighting contactors of the eight locations can be e de-energised based on commands from a remote server. This is to ensure ~ l l ~ i n a t ~level o n €or each camera to grab a satisfactory real-time image of each location. Via this box, the video signal of any one camera can be selected by an image ~ a b b e card r on a time-~ultiplex~ng basis. F u ~ e ~ o r the e , PC is c a ~ u n i c a t i ~with g all o microcontrollers in the existing distributed monitoring system. In addition, control si for p a ~ i n gand tilting each camera can be output from the box. C o ~ u n i c a ~ i obetween n ~ c e is accomplished by a modem. the PC and the CLP m a i ~ t e ~ centre On the sofhvare side, the on-site PC has two modes of operation, namely the re mode and the real-time mode. The regular mode i s active during normal operation. The onsite PG s e q ~ ~ e n ~grabs i a ~ ~images y from the eight cameras at a ~ ~ e ~ u e o€ n c5yseconds per frame. The value of the average grey level can be used to assess the overall ~lluminat~o of the site and the lighting system of the site can be switched on and off acc~rding~y. The average grey level of this updated image is further compared with that of the previous image, which was grabbed and saved onto the hard disk 40 seconds ago. If t ~ e i~s ea significant change in the average grey level, the two images cannot be compared d ~ e c t l y and the system will regard it as an error and wait €or another 40 seconds. ~ ~ e the ~ updated image is subtracted from the previous image so that any significant chan nsidered significant, the on-site PC will first of a11 save the two relevant images onto the hard disk for later reference and then inform the ma~ntenance centre by producing an alarm at the server. On top of analysing the images, the on-site PC saves the real-time images onto the hard disk at a frequency o f two sets per how. There are two levels of operatian being selected by the server, namely the coarse level and the fine level. Under the coarse level, images of size 320 pixels x 200 pixels are transmitted, resulting in a transmission cycle of only 48 seconds for the eight images from the eight respective cam er^. If the user finds anything unusual, the fine level can be

~

Power System Restructuring and

switched in, resulting in a transmission rate of around 35 seconds for each image of size 640 pixels x 400 pixels. The user is able to fix any camera ‘on-line’ and p a ~ ~ i l ~ z othat om p a ~ ~ c u lca r~ e r aThe . compression algorithm for these images is ‘ s ~ d ~ d with the quality factor set at 15 o/o so that the file size of coarse-level e-level images is around 30 kb. There are two f a c t o ~ ~ a n s m i ~ s ~rate, o n namely the quality factor and the speed o quality factor is the optimal value based on experime improvement is limited. If an ISDN link is provided from the s su~$~ation, the ~ansrnissionrate wilf be su~stan~ia~ly improved. This remote vision system requires neither spare contacts nor a d d ~ t ~ o n~ as~d u c e r sIt. can be used to prevent theft as well, General ins~ectionof the s such as c ~ e c k i ncleanliness ~ and quality of ~aintenancework. Is can be grabbed as images so that the user at ce centre can confirm. whether the are false or genuin~ in the aetivat~$zone. re~evan~ camera to see the existence of smoke remote vision system can be used to monitor external contrac necessary in the substation. ~ q u i p ~ ein n th a ~ r d o u areas s or areas withou~ ce, such as confined spaces or equipment rooms with live conductor^, can be monitored by this system. During major overhauling or fault h a n ~ l i n the ~, ~ a i n t e ~ ~ c e ~ a n a g e is r able to visualise the equipment status through the ~ i s ~ mon ~ay i n s ~ c t i o n to s the site engineers, Site problems encountered can be effici eration of the site staff and central management personnel.

Remote vision system

A p p ~ i ~ a ~ ofthe i o n lnternet to Power System ~ o ~ i t o and ~ nTrading g

smission of energy requires constant monitor ~ e ~ a n epower n t supply. ~ o n i t o r ~ nofg such occasions in which vital changes of important paramete used to predict a po~entialproblem with the equipment. Such monito sudden loss of sub-sta~ionequipment leading to unexpected power cuts. uncharacteristic behaviour of one element in the supply chain can i p e r f o ~ a n c eof other equipment or might even cause d ~ a g e This . is even worse if sporadic m a l ~ c t ~ oofn one device leads to damage in another device. Such sporad~c m a l ~ n c t ~ oofndevices can only be detected if continuous monitoring is practised. heref fore constant monitoring of equipment will improve power system reliabil reduce maintenance costs because devices can be replaced before they cause damage. tice, it is imposs~b~e to monitor every unit since they are geograph~cal~~ $is ~ u r t h e ~ o rite would , be too expensive to keep qualified personnel in remote locations for 24 hours a day, 7 days a week. Therefore, remote monitoring of e q u ~ ~ ~with e n the t help of c o ~ ~ u t ehas r s been practised for some time. Com~uter$are the perfect a l ~ e ~ a t i vtoe monitoring personnel since they are able to monitor constantly and accurately, detecting even the smallest changes in critical parameters. ~omputerscan be placed at important points of a substation. The price of computer hardware has been failing continuously for a number of years, making remote monitoring with computers effective and economical viable. Most of the hardware r e ~ ~ i r for e d real-time data collect are well established and robust. Once the connected and con~gured~ device ers for each hardware component make i access and control its ~ n c t ~ o n aonl ian ~ abstract level. Device drivers for disp ling cards or other hardware, for example, are genera~lywritten in a low-level programming language such as assemble^^ C or C3-i- to increase processing speed. Using ing l a n ~ a g e will s increase processing speed, since the ked for a specific type of processor and o p e r a t ~ ~system. g dency plays an important role when it comes to deciding which compoiients to buy. Not all companies can afford to update their device drivers in good time if 8 new or changed processor or operating system is in~oduced.Exp compon~ntssuch as U 0 sampling cards can be rendered useless if devic atible with the latest processors and operating systems. Therefore if h ~ d w a r ei s purchased in large numbers from a manufac~rer record of a continuous supply of device drivers. Once the ~onitoringco~puterscollect data, access to the data by users ~ u s be t gra~ted.Compu~erscan dis~ibutedata in many different ways. ~ a s i c a ~there l ~ , are two distinct network conste~la~ons. LANs, where local computers are connected via a local where remote ~ o m p u t e rare ~ connected by means of l o n g - d i s ~ c e g access to substation data via the I n t e ~ e trequires an I n t e ~ e t connection. uters act as data collection points for a remote power devices i s accomplished by sequen~ia~ly converting their analogue signals into digital ~ n f o ~ a t i ovia n 640 conve~ers.Such AIL) converters can be found on standard PC I/ s. The

Power System ~

50

~and ~ ~ r ~s ~ u l a t ~~ o

typical range of an A/D converter can be dJ2V or ASV, which requires a p p ~ o p ~ a t e conversion of the analogue signal to match the A/D converter’s input range. The AfD conversion sample frequency depends on how many data sources are conv~rtedand on the required data accuracy. It can be quite low (4 data t r ~ s f o ~ a t i o are n s planned. If, for example, spectral analysis or other data-intensive ~nsformationsare part of the overall monito~ngprocess, the s ~ p l i n gfrequency must satisfjr the mathematical constraints of the ~ a n s f o ~ a t i oused. ns In order to avoid loss of accuracy or injections of harmonics into the analogue signals, NI) c o ~ v e should ~ e ~ be placed as close as ~ o s s i b ~toe the source. Once the ~ n a ~ o ~ e signals are converted into digital info~ation,~ansmiss~on will not cause loss of accuracy. Figure 12.2 1 shows how substation components exchange data via a LAN.

re 12.21 Collection of data in a local PC

The raw data received via the LAN from the ~ i c r o c o n ~ o ~needs ~ e r sto be converted in such a way that it can be sent to the ~ a i n t e n ~ centre. ce It requires a f o ~ athat t is e extendable, in case new components are added to the monitoring requiremen~s, received from different sources needs to carry add~~ional information such as the name of the source, Its location, date and time, scaling factors, units and many more. There are different possibilities on how to encode this addi~ionalinformation. The most configurable and extendable formatting standard, which is widely accepted, is XML. It is compatible with all opera tin^ platforms since it is contained in a plain text file, for e x ~ p i e : 20 c/PowerAngle> CPowerRating Unit = kVAr 200 c/PowerRating>

ADD~icationofrhe Internet to Power System onitoring and Trading

$a tation data has been collected and stored in local PCs, it needs to be published. se o f the case study is to grant access to the substation data for all responsible parties. Such parties may be the p e ~ o ~ofe the l electricity and security CO brigade or other remote experts and advisers. In order to publish i n f o ~ a t i o nover the a Web server connected to the Internet is mandatory. is study, there are several different ways of d i s ~ i b u ~ i nn~f ~ ~ a t on i o the ~ Internet, such as:

Static Web ~ a ~with e s a dynamic applet and data polling. Static Web pages with a d ~ a m i applet c and data streaming. s ~entionedprevious~yin brief, static Web pages are not really suitab~efor cons~ntly require the data to be embedded within the document. If the ,the d o c u ~ e nneeds t to be changed r n ~ u a ~ erefo ~ y . fore, fast data ented by static Web documents. D ~ a ~Web i c pages we one way of publishing changing data over the a new static Web page is generated and ~ran§mi~ed to takes place, Such g e ~ e ~ t i oofnWeb pages can be do (CGI) and an executable program located in the Web k of the CGI inte~faceis to instan~iatethe C y the user. A p r ~ g used r ~ for the CGI c programming language ut must be compiled or i n t e ~ r e ~by ~ lthe e server. updatin~~ n f o ~ a ~ oi o the n ~~~e~ i s s u i ~ for b ~slow-chang~g ~ data suc weekly events.

eb page

~ ~ n e r on ~~ed request by the CGI clfthe server.

~ ~ c ~created l i Web y page via server CGI

pages me not limited to the CGI. They can be S e ~ l e t sare w r i ~ e nin Java and executed on the Web server. They fbnct in terms of r o b u s ~ e s way to ~ ~ I ~ pro ~ aams~ but e are d e ore flexible and re~iab~e security. If a Web page is generated d y n ~ i c a l l yon a server with ~ e q u e n data t chan every 10 m ~ u ~ ite is~ likely , that the client browser might display obsolete info The ~ r ~ ~ l with e r nserver-side- enerated Web pages is that the browser b e c o ~ e sava~~able on the server. ~rowsershave no

Powcr System Restructuring and Deregulation

452

if new data has become available on the server. ~ h e r e ~ ~the r euser , is r e ~ u i ~ etod~ ~ l the o a ~ con~inuouslyby selecting the refresh or reload option in the rowser. are even more options to generate dynamic Web pages. Internet ~ r o g ~ a ~ i n g l ~ g u a g such ~ s as active server pages (ASP), s ~ r v side e ~ i n c ~ ~ (d e~ S or ~ J) a v a ~ c ~ and pt

tic Web pages can contain a

e a ~ " data ~ ~update ~ e request via repea~~d r e ~ u e to s ~the ~ server ~

d

d

~

Application of the Intemet to Power System Monitoring and Trading

~

~

§3

12.24 M Real-time ~ c data updates via continuous connection to the server (data s~reaming~

The ~aintenanceoffice is connected to the remote power substations using a standard t~~~communication connection, as shown in Figure 12.25. ~ependingon which ~arameters are ~ o n i t o r ~ind the maintenance office, different data update ~e~hnologies need to be considered. In the case when all measurements taken from the appliances within the substations are within their set tolerances, transmission of averaged n i e a s u r e ~ ~ ~might its be suf~icien~. In case a fault QCCWS, all measured and locally stored data from a defined point could be transmitted. In order to receive continuous data transmission, data streaming is requi~edfor fast real-time data updates.

~ u b s ~ ~ 1i o n

Substation 2

2.25 Connection between power substations and the maintenance office

If remote expert advice or an on-line reporting is required, data transmis§ions of real measurements can be transmitted across the Intemet. There are several ways o f displaying r e a ~ - t i ~data, e e.g. as numerical values in an analogue or digital display or time series graph, as shown in Fibare 12.26. At present, browsers do not have a built in ~ n c ~ ~ o n a l i for supporting graphical representations of data. Therefore software extending the browser's display capabilities must be used. Applets could use the browser's client area for

Power System R e s t ~ ~ ~and r i~n~~r e ~ i a t i o n

454

drawing lines, shapes or colours. Such shapes offer the basic ~ n c t i o n a l ri e~ ~ u i r efor ~ controls capable for displaying real-time data.

igital Display

Analogue Display

Time Series

.26 Different controls to display real-time data

other impo~antaspect of working with applets is that they are able to connect back to the server from which they were loaded to retrieve new data updates, regardless of whether data polling or data streaming is used. Dispiay~greal-time data in an applet is roughly a two-step process, as shown in Fi The first step is to transmit the applet from the Web sewer to the browser. step is to transmit data to the applet,

12.8.3

~ o n ~ ~ ~ ~ i n ~

A few examples will be given on the mo~~toring of power station equip men^ such as circuit breakers for the prevention of major faults and supply i n t e ~ ~ t i Q ~ s . The SF, gas pressure measurement history over d a circuit breaker (C J a n 1996, ~ ~ was ~ ~ e s e n t efor 12.28. It can be seen that there has been a very serious SF, gas leakage problem with the CB and the system was successful in giving a warning to the ~aintenanceteam on 17 December 1995. The gas topping exercise was compie~edon 18 December 1995 to avoid a major failure ofthe CB. A method was develope~to measure the travelling of s based on looking at the c ~ ewaveforms. ~ t Figures 12.29 and 12.30 show the me X 32 kV CB which is used to switch a 132 kV, 80 reactor. From the figures, it can be seen that the closing time for the CB is 125 rns while the ~ippingtime is 50 ms.

Application of the Internet to Power System ~ o n i t o r i ~and g Trading

c4”

U,

3.7.

3.6.

3.5

3.4

IS

Time Is

re 12.29 Current waveform for closing of reactor CB

Power ~ystem~

Time is C u ~ e nwaveform t for tripping ofreactor C

be air c o ~ ~ r e os ~~e or a~~ time i n ~ (Ton) and idlin are sbown in ~ i 12.31. ~ If rToff ~is short, this ~ o r n air~ will ~ meet ~ ~the ~lower e limit ~ very quickly

storage tank. As a result, prec

Air compressor odoff timing

e

s

~ andc ~ e~r ~n~ ~~a t i o n

Application of the Internet to Power System ~onitoringand Trading

57 ~

St

er

Electricity deregulation is creating a free electricity market which is differen~from count^ to country. For each ~ e s t ~ c t uutility, r e ~ the market operator provides the essential service nction. Electricity ~radingin Europe will change ~ a m a ~ i c a ~asl ythe wholesale and retail markets open up to competition. Competition between utility suppl~ers will bring bene~tsto end users only if each competitor has the same access to ~nfo~nation regard~ngpower pricing and distribu~ion,To keep the energy marke~lacec o ~ p e ~ i t iit~ e , d i s c r i ~ ~ n atransparent ~o~, and easily accessible for each compet~tor. ng is not confined within a country’s borders. Many countries are to ne~ghb~uring countries so that a ~entralisedoperated can have a key role [9]. That kind of power exchange will have to offer a re~iableand efficient exchange information between the market participants by operating a r ~ ~ ~ a b l e , highly d~s~ributed and low-cost informa~ionnetwork. If the open energy market is to succeed, all participants must be wired into a s~andard data exchange Infrastructure that must be platforni and language ~ndepen~ent. Tl-tesefore the Internet, with its ~ l a t and ~ olanguage ~ independence, is the choice for h o s ~ on-line ~g wer traders require fast reaction to market changes. They nee to control their trades across all current bids, offers and iiegotiations by means of a mouse-c~i~k and r ~ ~ u irealre time ~ a r ~ ien ft o ~ a t i o n ,including market depth as well as vital news i n f o ~ ~ a ~ i Furtherxiore, on. anonymity during negotiations and tools for t analysis of marke~~ o n d i t i ~re the relevant ~e~uircments. er exchange with its large nuinbers of v ~ ~ a b l rn The complexity of the es predi~~ion of market trends rediet. Therefore, pa~icipan~s must be awa eters to s u p p o decision ~ making in the daily offer~ngpro xchange can s c h e d ~ ~ enough e capacity to meet all requi~ ibe different kinds of auctions natory and unifoK~auction system [lO]. An ideal power exchange r e ~ u l a ~ ~and o n reserve in ~ a ~ a lbased l e ~ on auctions. The dispa~ched regula~i~n is the capacity to maintain real-time g r e s e ~ ei s the prov~sionthat can res the market situation and follows t pa~~cipants have to ensure high p r o ~ t a b i l i ~

nd a clear re~at~onship between the value of a re, pa~icipantscan use an ageni with a specific be and trading systems coal services and new tools and technologies for controlling, ~cheduling~ ~ l e c ~ i ~c ~i ~ ~ e fin~e~~igent 5 r e , agent tech~ologyhas been develo~ power arke et as described in Chapter 11. Complex distributed system enefit the ~ ~ i t e r a c ~between ~ o n intelligent s ng of electricity. ln~eiligent agents per s in an on-line auction [ 131. As mentioned previously, agents for buying or selling electric^^ r e ~ r e 5 e n t ~either ~ g generators or consumers. In order to use agents to

458

Power System R e s ~ c and ~ D ~ enr ~ ~~ l a t i o n

advantage, each agent needs to present a unique economic and strategic behaviour model. These mode~s are based on human behaviour with respect to different tra env~onments,For example, agents can show an ;anxious buying and selling behaviour, greedy behaviour or relaxed behaviour to emulate market p~icipants. There are several ~nte~et-based simulation environments for exp various power exchange mechanisms avaiIabIe on the Internet [ 141. allow pa~icipantsfrom different locations to compete in the open market. This is advan~geousfor the training of personnel, who are able to try different buying and selling strategies under changing market conditions without causing interfkrence on a real trading floor. With the help o f more advanced trading platform models, differen~ auction types, e.g. uniform price, single and doub~e-sided auctions, and di~erent c o n s t r ~ i ~e.g. ~ ~transmission , losses, line capacity and stability limits and congestion s i ~ ~ t i ocan ~ s be , explored. The ultimate objective for each si~ulatjonwill always maximi~eprofits from trading energy.

The first step in building a trading platform over the I n t e ~ e is t to gain quality ~nternet access with enough i and width to serve all clients at a ~easonablespe Internet access cannot be achieved by telephone. It is necessary to rent or buy a dedicated r with a reliable ISP, which offers a 24-hour, '?-day customer service. nce a reliable Internet ~ ~ ~is es~blished, t i oserver ~ s o ~ a r must e be pure ing Web services, Currently, the most common Web servers are IIS ~ i ~ r o s o fApache t, Web Server from Apache and Web Logic. There are many so~tware co~panieso f f e ~ gcompetitive Web server so~utions, which can also integrate ecommerce packages. ing a reliable trading platform across the Internet i not trivial. A ~ ~ r n a must be to ensure data security and data ~ t c ~ t yata . security across the Internet has constantly been improved by the int~oductionof better and faster s e c ~ t y algo~thrns.The most used and trusted method is secure sockets layer (SSL). rela~ive~y simple to i lement and does not require changes to any existix~ Data integrity can be achieved by buying a database fiom a major vendor. Such may include startup consultancy and customer support. It is i m p o ~ to ~ tde database in such a manner that the database § ~ c ~will r edeliver optima^ performance. are sent to clients in XML format, conversions fkom table ~ o ~ a t abase response times. eref fore, the choice of database layout should match the d i s ~ b u ~ data e d format if possible [ 151. Figure 12.32 shows a simplified block diagram of a between clients connected to a trading platform. On connection to the L page conta~ningall the required fields to n submission of a transaction, the Web s tr~sactiondetails, which should be validated for c o ~ e c ~ e s ~ ase. If invalid data is contained in the ~ s a c t i o changes will be rolled back to restore the da suction servers can be purchased for keeping

Application of the Internet to Power System Monitoring and Trading

As with many real-time auction and trading platforms, data update§ are sent to the ceivd data updates via XML allow faster data updates, since n to the browser cIient area to avoid the generation of pages. ~ u ~ e ~ omore r e clients , can be s y n c ~ o n o u s ~~yp d a ~ because ed small portions of XML data are sent across the Internet, saving precious b ~ d w ~ ~ t h . It will take an entire p r o g r a ~ i n gteam to create a real-time auction platform from s to finish. There are several s o ~ compa~ies ~ e offering complete solution pac cornrn~rce and on-line auctions. I n t e ~ e t applications have different r e q ~ i r e m cunknown ~~~ to desktop a~pficatians. ~ e q u ~ e m e such ~ t s as sc c o n t i n u ~are ~ of great i r n ~ o ~ for ~ c~ e e applica~ions. b ~ e b - b a s se o~ ~ a r e for highly scalable products require a great knowledge of r n ~ ~ t ~ - t ~ e envir~nrnenzs aded and parallel process~ngarchitec~res. Client Computer I

time data via an

~ommunicationsaechitectuee

application of ~ ~ t e ~ e t ysaem ~onitoringand is a very e area, e examples have been the benefits derived fr e obvious. er, it can be seen that much work remains to be done. One area is system security in the openaccess e ~ v i r ~ n ~ e n t .

structuring and ~ e r e ~ l a t ~ ~ n

also like to thank E E E for

ission to r e p r o ~ ~ c

ymond ~ r e e i i rl n~t r~o ~ ~ c ~ itoo nthe I i i t @ ~for e ~Eng~neer~, 1998. Dynamics ~ T O’Reilly ~ L ~ s Teach Y o u ~ s e ~ 4~inT ~ L 1899, ~ e a c ~Press. pi~

Chan, A.T.P. So and L.L T r a ~ a c t i #on ~ sPower Sys ~ o ~ e e(If~thei Kn~ ~ ~~r n Q t i o n ~ l ~ # w @Techn~~og~es r 2000, IEEE, April 2000, pp.47 1-475. er Academic ishers hers^ 1999. Sheblh, ~ h a 6:~Agent ~ ebased ~ Econo~ics,in Power ~ y s ~~e~~s st ~ c ~and r iEconomics ng fl2] ~ ~ Liu, Naili ~ Song, nSacques Law ~ C

~

n,~ ‘New ~ e t h for o ~ ~

access fees, 162, 165, 167 active reserves, 25

battery charging, 28 bench~ark,116,125,128,X57,163 bid prices, 23,98, 176 bilateral con~acts,24,G 1, 5, 158, lG7, 168, 1 bilateral model, 96

black-start capabiIi~,93, 19 198,199,218

~ e n control ~ a ~s y s ~ e ~12 s, central utility model 52

148,259

autononiy, 355,356,359 a u ~ ~ " r ~ ~ l127, o s 128 ~~s,

back-to-back thyristors, 269,273 ba~ancin~ ~ ~ k eG8,78,85, t , 113

c o ~ p ~ ~ i t ixii, o n ,1,2,4,5,8,9, 11, 15,

Index

62

304,329,330,332,334,347,356, 360,373,377,420,457 competitive ~idding,1,65 competitive framework, xi, 110,353 co~petitive~eneration,2,3,4, 107 competitive metering, 114 competitive trading, 24 compu~tionalintelli~ence,xxi, 353 condition mQnitoring,129, 132,295, 300,304,312,313,320,322,328, 445 congestion manage men^, xiii, xxi, 58, 69,70,71,75,78,79,86,88,89,90, 94,95,97,99, 104, 178, 180, 5, 198,200,209,215,216, c o ~ ~ e s t i o~ n~ a n a g e ~markets, e n t 93,94 contract market, ]I0,6 I, 68, 179 contract path allocation, 57

damper, 273,274 data pol~ing,45 1,452,454 data security, 458 data s ~ e ~ i n451,452,453,454 g, database, 136, 137,319,321,408,419, 420,437,438,439,440, 458 d a y - a ~ ~61,69,79,86 ~d, day-ahead market, 71,78,90, 176, 178, 191 delivery time, 86,421,423 demand de~and ment, 1 IS demand-side bidding, 68 deregulatiQn,xii, xi& xiv, xviii, xix, 1,2, 5,6,7,9, 10, 15, 19,45,48,50,52, 52,55,57,58,64,70, 71,73, 108, 111, 116, 119, 133, 140, 153, 161, 167, 171, 173, 175,202,217,218, dere~latiomof energy market, 4 18 desalination plant, 38,49 discrete wavelet ~ a n s f Q338 ~,

dissolved gas analysis, 296,323,329 distrib~tedgen~ration,13, 16, 17,20,21, 22,23,25,26,46,48,99, 108, 144, I64 d i s ~ b ~ t gene~ation ed tec~ologies,13 distribution auto~atiQn,127, 128, 147, 148,151,418 distrib~tiQn co~panies,4,63,64, 110, 111, 113, 115, 116, 117, 119, 154, 175,302,316,318,353,361 distribution loss, 63 district heating, 21 d i s ~ r b ~reco~nition, ce 341,350 economic dispatch, 53,77,78,82, 109, 121,133,374,414 eddy currents, 325,326 elasticity, 59, 192, 195, 196,209, 215, 220 electrical d ~ s c h ~ g296,300 e, electricity and gas networks, 1 1 1 electricity dis~jbutionindustry, 111 electronic auction ~ a r k ~ t10 s, e-mail, 354,420,425,427,428,429 embeddedcost, 57,58,186,187,189, 190,194

embedded generators, 112 embedded systems, 128 emissions-free e l e c ~ c i 19 ~, energy function, 206,385 energy mix, 6 energy policy, 16,48 energy purchase cost, 113 energy storage, 5, 13,259,2 270,285 nzglish auction, 55,56 equilibriu~point, 68,69,84,97,206, 207 ethernet, 348 e v o l u t ~ o n comp~tin a~ ~ ~ Q l u t ipro on~~ ex ante market, 61 ex post market, 6 1,73 excitation capacitance, 27,223, 32, 35 expert s y s t e ~ s353,355 ,

Index

faci~i~tors, 359 fiber optic communication, 147 fiber-based ~ansmission,142 file types, 42 1 financial markets, 78,88,94,97, 171 financial ~ ~ s m i s s i rights, o n 95 first rejected o~fer,55 flexible AC transmission system, 162 flicker, 266,33 1,342,346,347,352 f o r c e - c o ~ u ~ t converters, ed 278 forward markets, 71,86,95, 106, 178, 361 fossil fuel, 3,4, 6,45,53 Fourier transform, 336,347 free space lasers, 141 ~equencym~dulatio~, 144 fuel cells, 10, 12, 13,20,2li, 26,99,330 hll graphics in~er~ace, 134 ~ ~ rmarket, e s 8,68,74, I0 362,364 fuzzy diagnosis, 323,325, 328 fuzzy logic, 38,49,341,412 g a ~ i n g50,78, , 83,88,91,92,95,98, 99,107 gas industry, 165 gas turbine technolog~,173 generation companie§, 22,67,72,73, 175,361 eneration mix, 11 156,180,412 genetic algorithm, xix, 49, 360, 362, 364,365,367,370,410,412,414 GIF image, 421 g a v e ~ e n~nte~ention, t 16,45 graph theory, 246,25 1 green c ~ ~ i ~ c a 17 tes, green energy, 20 278,279,280,283

63

harmonic distortion, I3,26,331,346, 348 harmonic instabilities, 34 head-rnounted displays, 3 hedging, 65,95,360 hedging contracts, 65 hidden nodes, 384,386,388,3 hot spats, 297,400,407,410 hour-ahead market, 158, 176, 178 HVDC, xvii, 73,260,263,264,266, 274,277,278,279,280,281,286 hybrid agent, 355,358,359 hydro, 3,5,6, 12, 13,20,68,72,73, 105,174,229,259,280,330 IGBT, 262,263,264,269,278,280 immersion, 395,396,397,400,405 incipient faults, 29(j9323,329 incremental cost, 53,57,83,84,85, 88, 90,91,92,99, 196 incremental cost allocation, 57 Independent Power Pro~ucers, independent system operator, 2 104,121, 175,217 inelastic load, 65,92 inequality cons~aints,198,21I, 212, 373 i n f o ~ a t i o nt e c ~ o l o2,54,59 ~.~ infrared detectors, 297,401 infrared irnager, 400 i n ~ s ~ e~ ~l ~r nei n1g , installed capacity, 22,25 223,23 1 266 intelligent electronic devices, 139 interface agents, 355,356,357,358 inte~ationalf i n ~ c i n ga~encies,124 I n t e ~ e txiv, , xvii, xix, 1 14, 1 18, 140, 141, 143, 144, 145,358,416,

458,459,460 auction, 55,56,60,61,65,67, 82,84, 90,91,95,96,98, 105, 108, 109,

Index

46

193,194, 195,362,413,457, baiidwid~¶ 43 1 ~ d u ~ a ~143,354,399,420 ~on,

ricing, 23, 57, 99, I 187 ~ ~clearing, ~ 65,71,85, k ~ 87,t89,90, 91,96,177 219,231,234,236

~ a r k ~anspar~ncy, e~ ~ e g a w mile a ~ al~ocation,57

in~er-zana~ ~ o n g ~ s ~ i88 on, inves 307 9

335 mother wavelet, 337,338,

mutation, 40,41,372, 373, 375, 3’76, 377,378

159, 160, 162 ~ a ~ g i costs, n a ~ 3, 53, 58,210,240,242

nodal p r i ~ i 59,73,88, ~~, 166, 167, 187, 188

Index

~ ~ o ~ - d i s c r i m i nauction, a t o ~ 55 n o n ~ ~ ~ ~s ye srt ~s ~~397 s ,~ e

-.-

power pool, 4,22,82,86,87,%3, 100, 109,159,176,179,182,183,18 185,292 power quality, xiv, 21,25, 21”79 127,

154,231

ly, 116,117, 119,1

150

Index

(I

real~t~me markets, 78, 86 r e ~ e s s anafysis, ~ ~ n 116 latory body, 110,33 r ~ ~ l aincentives, t o ~ 293 r e l i a b i ~benefit, i~ 189, 190

sation, 260,261,271,272, 275,276,282,285,28 service ~rovider,xiii, 111, 156, 162, 163, 164,170,288,289 se~ement,55,63,69,71,79, 177, 423 shadow prices, 96

units, 127, 129, 132 s i m ~ ~ ~electricity e o ~ s market, 87 single-p~aseloads, 27,46 smart agents, 355 smart m e t e ~ ~61 g, social welfare, 54,8 sment, xiii, 115, 117, 125,316 42,46,49,330,399, solar collectors, 38

rew wall, 428,429 ~ ~ s w o429 r ~ ,

§ y s t e ~ y n a ~ ~xxi, c s 80, , 101

Index

system marginal price, 23, system opera to^, xiv, 51,53,56,59,60, 61,65,69,73, 103, 115, 120, 121, 139, 154, 157, 158, 166, 168, 177, I78,192,X93,194,195,210,331 system-wide blacko~ts,155

UHFradio, 144, 149, 150 ~ b u n ~ ~ ixii, n g50, , 52,53,73, 1 uncons~inedschedule, 65 unified power flow controller, 275,33 1

t~e-or-pay,412 tap-chang~,261,277 telecommunicat~on~ n d u s153, ~ , 154 telephone n e ~ o r k 1, 14 thermal heating t e c ~ o l o37 ~, thermal limit, 58,59,66,259, 283 ~ h e ~ o g ~400,410,415 ph~, therrnovision cameras, 297 thyristor cQn~olled reactors, 266 t h ~ scontrolled ~ r series capacitor, 271, 285 tier supplier, 1 12 time of use, 135,190 tournament scheme, 377 t r ~ s i e nenergy t margin, 206 t r ~ s i e nst t a b i ~ ixvii, ~ , xx, 139,206, 219,285,412 ~ ~ n s m i s s i oaccess, n xvii, 5 1, 175, 184, 191, 197,200,216 transmission channels, 1 ~ ~ s ~ s scharge, i o n 58,90,95, 165, 168,199,211 transmission loss, xiii, 57,60,65,72, 105, 120, 165, 186, 191, 192, 196, 197, 198,204,214,247,257,373, 374,376,458 tr~nsmissionmodel, 8 tr~smissionopen access, xiii, 2 16,37P transmission pricing, xxi, 58, 105, 168, 169,187, 191,218,221,246 ~ ~ s r n i s s i protocol, on 417,427 FTP 428

103,104,108,109,1177,180 UNIX, 136,426 uplift charge, 55 usage charges, 162,16 use of system charges, 27, 72, f 11, li 15

mission revenue, 162, 16 transmission system expansion, 162, 163,170 two-tier system, 120

valley load time, 241 vertically integrated, 8, 50,58,64,72, 77,153,155,156,157,163,16 178,210,360 vertically integrated utilities, 77, 153 virtual e n v i r o ~ e n395 ~, visual display unit, 395 voice activated messages, 1I voltage collapse, 140,260 voltage control, X4,26,80,93,1 194,284 voltage dip, 117,333,350 voltage sags, 13,331, 332,334,335, 348,349,350 voltage source converter, 280 v o l u n t ~system operator model, 158, 160,161,162,163

WAN, 134,139,358,431, wavelet transform, 336,337,339,350

458

static, 417,433,437, Web server, 426,427,4 451,454,458 web space, 420,437 website, 75, 114

Index

8

LPE ca~ies,3 13 ay, 59, 178, 198, 199, 3, 14, 17,20,21,22,26, 45,49,53, 147,259,280,330,349

zonal price, 71, 166, 167, i! ricing, 90, 166, 167, 18