Energy Market Experience Learning from the Blackouts: Transmission System Security in Competitive Electricity Markets

INTERNATIONAL ENERGY MARKET EXPERIENCE ENERGY AGENCY This pdf-ﬁle is for the personal use of the purchaser only. The...

Author: International Energy Agency

83 downloads 1269 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

INTERNATIONAL

ENERGY MARKET EXPERIENCE

ENERGY

AGENCY

This pdf-ﬁle is for the personal use of the purchaser only. The information contained in it is subject to copyright by the OECD/IEA. This material is for sale; therefore, the User may not copy, republish, lend, post on any server (unless a multi-user licence is purchased for use on the intranet, in which case the multi-user license terms apply), use on Web sites, transmit or redistribute to any mailing lists any part of the material for commercial purposes, for compensation of any kind or simple dissemination.

Learning from the Blackouts Transmission System Security in Competitive Electricity Markets

INTERNATIONAL ENERGY AGENCY The International Energy Agency (IEA) is an autonomous body which was established in November 1974 within the framework of the Organisation for Economic Co-operation and Development (OECD) to implement an international energy programme. It carries out a comprehensive programme of energy co-operation among twenty-six of the OECD’s thirty member countries. The basic aims of the IEA are: • to maintain and improve systems for coping with oil supply disruptions; • to promote rational energy policies in a global context through co-operative relations with non-member countries, industry and international organisations; • to operate a permanent information system on the international oil market; • to improve the world’s energy supply and demand structure by developing alternative energy sources and increasing the efficiency of energy use; • to assist in the integration of environmental and energy policies. The IEA member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, the Republic of Korea, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, the United States. The European Commission takes part in the work of the IEA.

ORGANISATION FOR ECONOMIC CO- OPERATION AND DEVELOPMENT The OECD is a unique forum where the governments of thirty democracies work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies. The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The European Commission takes part in the work of the OECD.

© OECD/IEA, 2005 No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications should be sent to: International Energy Agency (IEA), Head of Publications Service, 9 rue de la Fédération, 75739 Paris Cedex 15, France.

FOREWORD

FOREWORD Electricity market reform has fundamentally changed the environment for maintaining reliable and secure power supplies. Growing inter-regional trade has placed new demands on transmission systems, creating a more integrated and dynamic network environment with new real-time challenges for reliable and secure transmission system operation. These operational challenges are intensified as spare transmission capacity is absorbed. The major blackouts of 2003 raised fundamental questions about the appropriateness of the rules, regulations and system operating practices governing transmission system security. Despite the considerable efforts since 2003 to address the weaknesses exposed by the blackouts, it can still be argued that the development of these rules and operating practices have not kept pace with the fundamental changes resulting from electricity market reform. More can and should be done. Management of system security needs to keep improving to maintain reliable electricity services in this more dynamic operating environment. The challenges raise fundamental issues for policymakers. This publication presents case studies drawn from recent large-scale blackouts in Europe, North America, and Australia. It concludes that a comprehensive, integrated policy response is required to avoid preventable large-scale blackouts in the future. The legal and regulatory arrangements governing transmission system security can be enhanced. In particular, scope exists to clarify responsibilities and accountabilities and to improve enforcement. System operating practices need to give greater emphasis to system-wide preparation to support flexible, integrated real-time system management. Real-time coordination, communication and information exchange, particularly within integrated transmission systems spanning multiple control areas, can and must be improved. Effective real-time system operation requires accurate and timely information and state-of-the-art technology to facilitate effective contingency planning, system management and coordinated emergency responses. New and existing technology could be more fully employed to enhance effective system operation. Appropriate training is also required to enhance emergency responses. More effective asset and vegetation management can also make a valuable contribution to strengthen transmission system security. An effective policy response should also consider how best to employ marketbased approaches to complement regulatory arrangements to strengthen transmission system security at least cost. 3

LEARNING FROM THE BLACKOUTS

Governments should provide the leadership and drive needed to establish effective, coordinated processes that address the key policy issues in an integrated and comprehensive manner. At the same time, all stakeholders need to work together to address these challenges if we are to avoid unduly exposing transmission systems to the risk of further substantial power failures. This book is part of the Agency’s energy market experience series. It is published under my authority as Executive Director of the International Energy Agency. Claude Mandil Executive Director

4

ACKNOWLEDGEMENTS

ACKNOWLEDGEMENTS The author of this publication is Doug Cooke, Principal Advisor – Electricity Markets, working under the direction of Ian Cronshaw, Head of the Energy Diversification Division, and Noé van Hulst, Director of the Office of LongTerm Cooperation and Policy Analysis. This book drew extensively from a series of three workshops the IEA held in 2004, which explored policy issues affecting transmission system reliability, performance and technological development. The IEA wishes to acknowledge, in particular, the valuable contributions of the workshop and session chairs and of the many speakers who made presentations to these workshops. The book also benefited greatly from comments received from many other contributors, most notably from the IEA’s Standing Group on Long Term Co-operation, from Peter Fraser of the Ontario Energy Board, and from Ulrik Stridbaek, Jolanka Fisher and Giuseppe Sangiovanni of the IEA Secretariat. Special thanks go to Giuseppe Sangiovanni and Jenny Gell for their tireless administrative efforts to ensure the success of the reliability and performance workshops, and to colleagues from the IEA’s Energy Technology Office, notably Madeline Woodruff, for organising the technology R&D workshop. Thanks also to Muriel Custodio who managed the production of the book, to Corinne Hayworth who designed the layout and front cover, and to Bertrand Sadin who prepared many of the figures and maps.

5

TABLE OF CONTENTS

TABLE OF CONTENTS Executive Summary .......................................................11 Introduction.................................................................23 Chapter 1 Concepts and Context .........................................27

Key Concepts ........................................................................................................................27 A Brief Outline of the System Security Challenge...................................................28 Reliability Standards and Operating Practices..........................................................31 Electricity Markets Help to Strengthen Transmission System Security .............36 Electricity Markets are Changing the Way Transmission Systems are Used...38 More Dynamic Operating Conditions Magnify Challenges for Managing Transmission System Security..........................................................................................44 A More Dynamic Operating Environment May Also Affect the Nature and Frequency of Outages...............................................................................................48 Reliability Standards and Operating Practises Must Change to Meet the New Challenges ...........................................................................................................53 Chapter 2 Transmission System Security Case Studies.......55

Introduction...........................................................................................................................55 Case Study 1: The North-eastern United States and South-eastern Canada...55 Case Study 2: Switzerland and Italy.............................................................................73 Case Study 3: Sweden and Denmark...........................................................................90 Case Study 4: Australian National Electricity Market ............................................99 Chapter 3 Policies to Strengthen Transmission System Security ...............................................................107

Strengthening Transmission System Security Incentives.....................................108 Legal, Regulatory and Structural Framework...........................................................111 Improving Transmission System Security Standards .............................................127 Co-ordination, Communication and Information Exchange ..............................136 7

LEARNING FROM THE BLACKOUTS

Investing in Transmission Capacity ............................................................................142 Investing in Technology..................................................................................................143 Investing in People...........................................................................................................152 Asset Performance and Vegetation Management ................................................155 Applying Market-based Approaches to Support Transmission System Security.................................................................................................................................163 Annex 1 United-States-Canadian Power System Outage Task Force and North American Electricity Reliability Council Investigation Recommendations and Implementation Status..177 Annex 2 North American Electric Reliability Council Draft Functional Model ....................................189 References .................................................................203 List of Tables

1 2 3 4 5 6 7 8

• • • • • • • •

Emerging Conditions Affecting Transmission System Reliability..........39 Some Major North American Blackouts: 1965 to 2002..........................49 Service Restoration in the United States.......................................................65 Primary Causes of the Swiss-Italian Blackout ..............................................84 Automatic Under-frequency Load Shedding by Region .........................101 Inter-regional Power Flows Following the Disturbance..........................102 FCAS Service Delivery During the Event.....................................................103 Examples of Technologies with the Potential to Enhance Transmission System Security..........................................................................144 9 • Examples of Emergency Load Response Programmes............................172 List of Figures

1 • Dy Liacco System Security Model.....................................................................34 2 • Growth in Inter-regional Power Exchanges in the Nordic Electricity Market: 1997 to 2004 .........................................................................................40 3 • Monthly Transmission Loading Relief Requests: 1997 to 2005............41 4 • Power Flows between West Denmark and Norway: 1995 and 2000 ..42 8

TABLE OF CONTENTS

5 • Difference between Day-ahead Scheduled and Actual Metered Power Flows on all Interconnectors to the New England Independent System Operator Region: August 2005 ..............................43 6 • Parallel Power Flows on a Transfer from Wisconsin to Tennessee........45 7 • Parallel Flows Associated With a 1000 MW Transfer from France to Italy........................................................................................................................47 8 • Frequency Distribution of North American Outages Affecting More than 50 000 Consumers: 1991-95 and 1996-2000 .....................50 9 • Frequency Distribution of North American Outages Affecting More than 100 MW: 1991-95 and 1996-2000...........................................51 10 • Selected North American Outages: 1990-94 and 2000-04...................52 11 • Key Elements of the Ohio Phase of the Blackout .....................................57 12 • Overview of the Regional Cascade Sequence..............................................60 13 • Line and Generation Trips, and Lost Load during the Cascade Phase ... 62 14 • Area Affected by the US-Canadian Blackout...............................................63 15 • Restoration Strategy in Ontario........................................................................67 16 • Initial Load Restoration in Ontario: 14-15 August 2003 ........................68 17 • Load Restoration in Ontario: 14-22 August 2003.....................................69 18 • Overview of the Italian Interconnector Separation Sequence ...............75 19 • Final Line of Separation of the Italian Transmission System from the UCTE Transmission Network ............................................................76 20 • Key Events during the Transitory Period........................................................78 21 • Restoration Progress Following the Italian Blackout.................................81 22 • Impact of Primary Causes on System Security During the Swiss-Italian Blackout ...................................................................................85 23 • Area Affected by the Swedish-Danish Blackout, 23 September 2003 ....93 24 • Swedish Load Restoration Following the Outage ......................................95 25 • Danish Load Restoration Following the Outage ........................................96 26 • NERC Regions and Control Areas..................................................................124 27 • Cost Comparison of N-1 and N-2 Security: Simple Example................129 28 • UCTE Wide Area Measurement System.......................................................149 29 • CAR Visualisation................................................................................................150 30 • The Wire Zone – Border Zone Model ............................................................161 31 • The FCAS Trapezium...........................................................................................168 9

LEARNING FROM THE BLACKOUTS

List of Boxes

1 • East Central Area Reliability Co-ordination Agreement (ECAR) and Regulatory Independence ........................................................................118 2 • Overview of EPRI’s Probability Risk Assessment (PRA) Methodology..... 131 3 • Applying PRA Methodologies: Three Case Studies..................................133 4 • Integrating Technologies to Improve Transmission System Security.....148 5 • Overview of the Australian Frequency Control Ancillary Services Markets....................................................................................................................167

10

EXECUTIVE SUMMARY

EXECUTIVE SUMMARY Modern economies are becoming increasingly dependent on reliable and secure electricity services. The substantial supply disruptions that struck North America and Europe during 2003 clearly demonstrated the fundamental importance of transmission networks for the efficient and secure operation of electricity markets and highlighted their vulnerability to transmission network failures. While large blackouts are by no means a new phenomenon and have happened in the past before electricity reform, these disruptions created considerable public concern, with some claims that electricity reform had reduced electricity system reliability. Growing public sensitivity to supply disruptions reflects the increasing dependence of modern economies on reliable and efficient electricity supplies, and adds to the pressure on governments to effectively address reliability issues. This study explores policy issues associated with the system security dimension of reliability, with a particular focus on transmission system security in competitive electricity markets. System security refers to the ability of a power system to withstand the unexpected loss of key components. This can be thought of as the operational dimension of electricity reliability. This work draws on case studies of recent large-scale blackouts involving the failure of transmission systems to help identify key lessons for policymakers. Issues associated with the adequacy of transmission infrastructure and investment to meet growing demand for transmission services in electricity markets is addressed in the IEA publication Lessons from Liberalised Electricity Markets, which examines key lessons for policymakers from liberalisation in IEA member countries over the last decade.

Electricity Reform Creates New Challenges for Maintaining Transmission System Security................... Electricity reform has brought more efficient use of transmission systems and greater regional integration of power flows resulting from inter-regional trade. The many benefits flowing from electricity reform are discussed in the IEA publication Lessons from Liberalised Electricity Markets. Greater integration has helped to improve overall transmission system security by permitting more effective reserve sharing. It has also helped to reduce transmission system security costs. Growing demand for transmission capacity to accommodate inter-regional trade means that transmission systems are increasingly being run at or near their security limits. 11

LEARNING FROM THE BLACKOUTS

Electricity market reform has also brought unbundling and independent, decentralised decision-making. As a result, decisions affecting network operation and performance that were once made in a centrally coordinated way within vertically integrated utilities are now made by many independent market participants. Decentralised decision-making has fundamentally changed utilisation of transmission networks. Previously stable and relatively predictable patterns of network use have in many cases been replaced with less predictable usage, more volatile flows and greater use of long-distance transportation, reflecting growing inter-regional trade. New patterns of transmission network use are creating a far more complex and dynamic operating environment, with real-time monitoring and management by system operators becoming more and more crucial for maintaining transmission system security. At the same time, unbundling has reduced system operators’ capacity to manage system security through coordinated actions across the value chain. The emergence of regional markets that span multiple control areas has added to the transmission system security challenge by increasing each system operator’s exposure to the operational decisions of other system operators and to potential failures outside their control area. In this more integrated and dynamic operating environment, an event affecting a relatively distant part of a transmission system may have greater potential to spread and severely disrupt the supply and operation of electricity markets.

A Comprehensive and Integrated Policy Response is Needed .............................................................................. Despite these fundamental changes, development of regulatory arrangements, reliability standards and system operating practices relating to transmission system security generally have not kept pace with the new real-time system operating challenges resulting from electricity market reform. The major blackouts of 2003 and 2004 raised searching questions about the appropriateness of existing regulation, rules and practices in this new operating environment. Examination of these matters is being undertaken in the wake of these disturbances, but the focus has been relatively technical. Failures of the kind experienced in 2003 and 2004 also raise more strategic issues for policymakers. The new challenges are interrelated, requiring a comprehensive, integrated policy response if they are to be successfully addressed. An effective policy 12

EXECUTIVE SUMMARY

response must address the many policy dimensions of transmission system security in competitive electricity markets including: legal, regulatory and structural arrangements; technical standards; operating practices; coordination, communication and information exchange; training; application of technology; asset management; and vegetation management. It should also consider how best to employ market-based approaches to complement regulatory arrangements to strengthen transmission system security. All key stakeholders whose actions affect transmission system security must be involved in this process, because unbundling effectively makes maintaining system security a shared responsibility that needs to be precisely allocated among the parties involved. Key stakeholders in this context will include governments, system operators, transmission owners, regulators, and market participants. Governments are well placed to provide the leadership and drive needed to establish an effective and coordinated process that addresses the key policy issues in an integrated and comprehensive manner. Inadequate responses to these issues are likely to encourage overly conservative management of transmission capacity at the expense of efficient inter-regional trade, and possibly leave interconnected networks unduly exposed to the risk of further substantial power failures.

Legal, Regulatory and Structural Framework...................... Legal and regulatory arrangements provide the foundation for establishing effective governance and incentive structures for transmission system security in competitive markets, replacing previous substantial reliance on reliability obligations placed on local, vertically integrated utilities. The case studies highlight the failings of poorly defined governance arrangements and the importance of clarifying roles, responsibilities and accountabilities within a legal framework that provides comprehensive coverage and enables effective enforcement. An effective legal and regulatory framework should: ●

clarify individual and shared responsibilities for transmission system security;

●

align accountabilities with the new functional responsibilities resulting from unbundling;

●

ensure the boundaries of authority to act are specified for each party, and that parties have sufficient authority to undertake their responsibilities within those boundaries; 13

LEARNING FROM THE BLACKOUTS

●

provide strong incentives for effective coordination and information exchange, within the value chain and across systems spanning multiple control areas, reflecting the shared nature of responsibility for aspects of transmission system security;

●

create transparency and objectivity, given the potential commercial implications of system operators’ actions in competitive electricity markets;

●

strengthen coverage, accountability and enforcement, where necessary, to help reinforce incentives for providing appropriate levels of transmission system security, and to build the credibility of the regime;

●

be applied consistently across an integrated transmission system; and

●

balance market requirements for access to transmission capacity with transmission system security requirements.

Regulatory arrangements need to be independent, with regulatory processes characterized by transparency, objectivity and consistency. An institutional framework based largely on industry self-regulation is no longer considered credible by market participants in competitive electricity markets. Concerns have been raised about the independence and objectivity of such arrangements. Conflict of interest could compromise the development, application and enforcement of system security rules, or lead to inertia as competing interests are unable to resolve rule-making issues in a timely or effective manner. Independence could be addressed by allocating regulatory responsibility for transmission system security functions to existing industry regulators. However, economic regulators may not possess sufficient technical competence to effectively verify and enforce compliance with transmission system security requirements. A careful balance is required to ensure independence and competence. Structural arrangements that promote independence, transparency and objectivity will reinforce the legal and regulatory regime. Independence and objectivity can be strengthened by ownership unbundling of system operating functions from contestable components of the value chain. Ownership unbundling would help to reinforce incentives for transparent and nondiscriminatory behavior that are more closely aligned to efficient management of system security and market-based dispatch. It would also remove any related conflict of interest that could influence operator interventions and possibly adversely affect transmission system security. Ownership unbundling of system operation functions would enhance the real, and perceived, credibility of system operation in competitive electricity markets.

14

EXECUTIVE SUMMARY

Exposure to legal liability can help strengthen incentives for effective system operation. However, the nature and scope of legal liability needs to be clearly specified and consistent with system operators’ functional responsibilities in unbundled electricity markets. Excessive or poorly defined liability may encourage conservative practices which could undermine effective emergency responses. This is a particular issue in the context of load shedding practices. A regime of limited liability would allow for an appropriate balance between commercial discipline and encouraging appropriate actions to manage an emergency event. Transmission system security can be strengthened through greater aggregation of regulatory and system operating functions across integrated transmission systems spanning multiple control areas. A single independent system operator with a holistic, real-time perspective of an integrated transmission system and with sufficient resources to manage credible emergency events is in a strong position to effectively respond to such events in a manner that minimizes their overall impact. Similarly, a single regulatory structure would facilitate transparent, objective and consistent monitoring and enforcement across an integrated transmission system. However, there may be legal, technical and organizational challenges that create practical barriers to aggregation in some cases. Effective coordination of system operation and regulatory activities becomes particularly important in the absence of more aggregated structures.

Transmission System Security Standards............................ Operational standards applied to manage transmission system security have changed little since the introduction of electricity market reform, with great reliance placed on the N-1 standard1. The standard is typically applied in a deterministic way that does not take account of the probability of a failure occurring or the impact of potential failures. The blackouts raised questions about how these standards are interpreted and applied. Improved compliance with existing standards may not address the fundamental appropriateness and effectiveness of those standards in reformed electricity markets. Applying more onerous standards would generally increase the level of transmission system security within an integrated transmission system, but at great cost. 1. A power system can be described as being N-1 secure when it is capable of maintaining normal operations (ie. reliably delivering electricity of a given frequency and voltage subject to technical limits) in the event of a single credible contingency event, like the loss of a transmission line, generator or transformer.

15

LEARNING FROM THE BLACKOUTS

Probabilistic methodologies, such as probability risk assessment2, could be used to enhance existing standards, providing a more flexible and adaptable operational standard that better reflects more dynamic, real-time operating conditions. Probabilistic approaches could be refined to incorporate a measure of the potential cost and benefits associated with a given level of system security. This could enable more responsive management of system security reflecting the value markets may place on secure electricity services. Technologies that improve real-time system information and that enable operators to directly control power flows over particular transmission lines would increase the potential for using probabilistic methods to augmenting existing standards.

Coordination, Communication and Information Exchange .......................................................... Operating practices need to more fully reflect the dynamic and regionally integrated operating environment resulting from electricity reform. In particular, operating practices need to be flexible and adaptable to permit effective real-time responses to system emergencies and disturbances. Contingency planning and emergency responses need to be undertaken from a whole-of-system perspective, reflecting the shared nature of responsibility and actions required to maintain transmission system security, particularly in integrated transmission systems spanning multiple control areas. Effective coordination, communication and information exchange is vital in this context. Interaction between control areas needs to move beyond day-ahead information exchange and exception based coordination closer to real time. Coordinated real-time security assessment and control is required. Multilateral agreements covering all control areas within an integrated transmission system need to be implemented to ensure, among other things, holistic operational contingency planning and to provide agreed protocols for coordinated action in the event of an emergency situation. Some progress has been made in these areas, however, scope exists to enhance and strengthen these arrangements. Coordinated management of transmission system security needs to be supported by real-time data exchange and communication. Accurate, real-time information is required to support effective contingency preparation and 2. Probability risk assessment (PRA) involves identifying initial events and possible sequences of related events that can lead to system failures, and estimating the total probability of these events occurring. PRA essentially involves calculating a measure of the probability of undesired events on the transmission system and a measure of the severity or impact of those events. PRA has been successfully applied in several industries with complex engineering systems that are exposed to low probability, high consequence failures.

16

EXECUTIVE SUMMARY

emergency responses. Information also needs to provide a holistic picture of an integrated transmission system, to support effective, coordinated management of system security. Greater harmonization of data standards would improve the quality of information, and promote more effective information exchange between system operators within an integrated transmission system. Communication strategies also need to be developed to ensure effective information dissemination to other key stakeholders, such as governments and the community, during emergencies and system restoration. Governments should examine existing coordination, communication and information exchange arrangements with a view to improving and strengthening them as appropriate.

Investing in Technology........................................................ Development and deployment of new and existing technologies has considerable potential to improve system operating tools to enhance transmission system security. Technology can improve the accuracy, quality and timeliness of information and support the development of more accurate and dynamic system modeling. This, in turn, would support more flexible and adaptable contingency preparation and promote greater real-time situational awareness. Technology may also improve effective operator control over power flows, which would permit more flexible operation of transmission systems and more effective real-time responses to alleviate congestion, manage emergency situations, and enable timely service restoration. Furthermore, it offers the potential to facilitate real-time coordination and more holistic management of system security in transmission systems spanning multiple control areas. A number of new technologies are under development which will contribute substantially to improve transmission system management in terms of increasing flexibility and security. Integrated deployment of technologies has the potential to substantially improve transmission system flexibility and resilience while enabling more innovative solutions to improve reliably, power quality and market efficiency. Technology, however, should not be viewed as a ‘magic bullet’. An appropriate balance between automatic and human control needs to be maintained, such that technology supports effective system operation and does not become a substitute for it. 17

LEARNING FROM THE BLACKOUTS

Regulatory or institutional barriers to efficient development and deployment of cost-effective technologies need to be identified and removed. Effective processes will be needed to test and validate new technologies, to help reduce uncertainty and deployment risks. Governments should support the development of such processes. They could also consider providing assistance for promising areas of research and development, where appropriate. Any assistance should be provided in a framework of strong cooperation with industry to ensure that outcomes are efficiently and effectively deployed.

Investing in People ............................................................... Highly trained and experienced personnel are required to manage transmission system security in the more dynamic real-time operating environment created by electricity reform. More attention has focused on emergency training in the wake of the 2003 blackouts. Training programs are being implemented based on emergency simulations to sharpen these skills. However, new training programs may be needed in this new operating environment. Competencies required to successfully manage transmission system security in liberalized electricity markets need to be identified and training programs reviewed to ensure that they equip system operators for the more dynamic real-time operating challenges associated with electricity markets. Joint training programs should be developed to facilitate more effective coordination among system operators managing an integrated transmission system. Training could also be extended to other relevant professionals involved in supporting secure system operation. These could include personnel managing related information technology and other technical or engineering staff. Training might also be extended to other parties whose actions could affect transmission system security, such as generator plant operators and technical regulators.

Asset Performance and Maintenance.................................. Transmission system security is fundamentally dependent on predictable and reliable asset performance. Innovative maintenance practices should be applied where possible. Condition-based monitoring, for example, could be used to target and optimize maintenance efforts, to reduce maintenance down time while also minimizing the risk of component failure at least cost. Coordination of maintenance cycles across integrated transmission systems 18

EXECUTIVE SUMMARY

would further improve transmission system security by minimizing the impact of scheduled line outages on available transmission capacity. The case studies highlight the need for effective verification and enforcement to ensure that equipment is properly maintained and operates predictably, particularly during emergency situations. Maintenance requirements and performance standards should be reviewed, with a view to clarifying standards and strengthening verification and enforcement regimes as appropriate. Protection settings also need to be reviewed to ensure that they provide an appropriate balance between equipment protection and reliable system operation. Regulators need to be mindful of the potential impact of regulatory decisions on incentives for efficient asset management. Regulatory outcomes need to allow adequate cost-recovery for prudent maintenance.

Vegetation Management...................................................... Contact with trees is one of the most common causes of transmission line failure and was a key cause of the major blackouts in North America and Italy in 2003. Effective vegetation management is critically important for maintaining transmission system security. Transmission owners should establish formal vegetation management plans based on accurate information about the vegetation under and adjacent to transmission lines. Field inspections of vegetation conditions should occur on a frequent basis, with the inspection schedule based on anticipated growth. Easements should be maintained in accordance with environmentally sound control methods that minimize the potential for fast-growing trees to take root under or adjacent to transmission lines. Vegetation management standards need to be reviewed, with a view to creating clearer, more consistent and enforceable standards where possible. However, considerable variation can occur in vegetation and climatic conditions affecting easements, which may practically preclude absolute prescription in this context. A balance will need to be struck between flexibility to accommodate local variations and certainty to permit effective verification and enforcement. Regulatory arrangements should also be reviewed to remove potential duplication, resolve conflicting requirements, and to streamline regulatory approval processes for undertaking vegetation management. Vegetation 19

LEARNING FROM THE BLACKOUTS

management is an expensive exercise and regulators need to ensure that regulatory arrangements permit recovery of prudent expenditures.

Market-based Approaches can Enhance Transmission System Security.............................................. In general, markets encourage more flexible and responsive use of the transmission system, which has the potential to complement system operator management of transmission system security by reducing pressures on transmission resources at times when systems are congested and operating at or near their security limits. Market-based mechanisms also allow more efficient, innovative and better targeted provision of transmission system security at least cost, and help improve the flexibility and efficiency of transmission system security management. Significant benefits have already been achieved from market-based procurement of certain ancillary services. Commercial arrangements have led to a substantial reduction in the overall cost of ancillary services for maintaining transmission system security in United Kingdom, United States, Scandinavia and Australia. They have also allowed for more coordinated and transparent management of shared system security responsibilities in the Nordic market. Opportunities to extend market-based approaches for purchasing ancillary services should be examined. For example, market-based procurement could be expanded to include a wider range of ancillary services, such as more effective coverage of network control ancillary services like reactive power. Potential may also exist to use more dynamic market-based models, like wholesale spot market auctions, which have the potential to substantially improve flexibility and efficiency by more closely aligning resource procurement with real-time requirements. Another possibility may involve moving toward cost allocation based on the causer-pays principle. The Australian frequency control ancillary service market has made considerable progress in these directions. Substantial benefits may also be achieved from more effective harnessing of demand response. Demand reductions in response to high prices are likely to occur when transmission systems are operating close to their security limits. Such responsiveness would help reduce pressure on system security and improve reliability by improving the balance between generation and load. To date, efforts to harness demand responsiveness have typically relied on administrative approaches such as demand management or ad hoc public appeals. More effective market-based approaches need to be developed. 20

EXECUTIVE SUMMARY

A variety of interruptible contracts with large industrial users are already used in North America, the United Kingdom, Scandinavia and Australia. However, considerable potential remains to be tapped. Innovative financial products can support more effective harnessing of demand response. For instance, retail products could be developed that offer consumers differing degrees of service delivery. Such products would allow large customers to choose the level of service they are willing to pay for. They could also be used to identify consumers that are more willing to experience service interruptions. This could support more efficient targeting of certain emergency interventions, such as load shedding, to protect system security at least overall cost to the community. Governments and key stakeholders need to consider how to promote the development of market-based demand response to support transmission system security. The Nordic and United Kingdom markets have made considerable progress in this direction. However, there are limitations to the application of market-based approaches in this context. System security exhibits some characteristics consistent with a public good, and it is possible that a pure market solution may produce insufficient resources to maintain system security. The strength and depth of competition to provide services may also have implications for the deployment of market-based mechanisms. Application of market-based mechanisms will also be affected by the quality and timeliness of information, and by market participants’ ability to access information. Effective real-time metering, information management systems and control equipment are required. However, installation of these systems can be relatively expensive and coverage may therefore be limited to relatively few market participants. This could represent a practical barrier to rapidly deepening and broadening participation in the short-term, particularly in the context of demand response. These and other practical, technical and institutional issues need to be carefully considered in the context of developing and deploying market-based mechanisms and products to help manage transmission system security. The combination of these issues may place some practical limitations on deploying market-based mechanisms at this time. Regulatory approaches will continue to be needed to ensure that transmission system security is not jeopardized. However, market-based mechanisms and products offer considerable potential to enhance and complement regulatory arrangements to strengthen transmission systems security at least cost. 21

INTRODUCTION

INTRODUCTION Modern economies are fundamentally dependent on reliable and secure electricity services. Electricity makes an essential contribution to economic performance, international competitiveness and community prosperity. The growing importance of the information technology and communications sectors for economic activity and prosperity is serving to reinforce reliance on high quality electricity services. The Electric Power Research Institute (EPRI) has estimated the total cost of power disruptions to the information technology and communications industries in the United States at around $52 billion in 2003, and at around $100 billion for the economy as a whole or 1% of gross domestic product per annum3. Growing public sensitivity to supply disruptions reflects the increasing dependence of modern economies on reliable and efficient electricity supplies. Such sensitivity adds to the pressure on governments to effectively address these issues. The substantial blackouts that struck North America and Europe during 2003 clearly demonstrated the fundamental importance of transmission networks to the efficient and secure operation of electricity markets and highlighted the vulnerability of electricity markets to transmission network failures. While large blackouts have happened in the past, these disruptions created considerable concern among policymakers, practitioners and the general public about transmission system security and its implications for the efficient and reliable operation of electricity markets. In response, the IEA initiated a project to examine key issues affecting the reliability and performance of transmission networks serving competitive electricity markets. Project goals included to: identify and analyse the key issues affecting the development and performance of transmission networks in competitive electricity markets; promote understanding of these issues among policymakers and regulators; and facilitate debate and exchange of views between stakeholders about these issues and how best to address them. Three workshops were held in 2004 exploring various aspects of these issues. The first workshop focused on transmission network reliability in competitive electricity markets. Key issues addressed included: lessons from the 2003 supply disruptions in North America and Europe; potential technological responses to improve system operation and reliability; regulatory

3. EPRI (2003a).

23

LEARNING FROM THE BLACKOUTS

arrangements to strengthen transmission reliability in competitive markets; and regulatory and organisational options to improve transmission system operation. The second workshop focused in more detail on the technological dimensions associated with improving the reliability and performance of electricity transmission and distribution. Key issues addressed included reviewing technological options to improve reliability and power quality; improving and expanding effective system control of integrated transmissions systems, and more effectively integrating distributed and intermittent generation technologies. Opportunities for improving research and development were identified and potential for greater international collaboration to promote development and appropriate deployment of technologies were discussed. The third workshop focused on transmission network performance in competitive electricity markets. Key issues discussed included: the policy context for improving transmission performance in competitive electricity markets; the challenges and options to improve transmission reliability, planning processes and outcomes; encouraging timely and efficient network investment to improve transmission performance; regulation to reduce risk and strengthen transmission performance in competitive markets; and the potential for greater integration of transmission networks and competitive electricity markets to strengthen incentives for superior transmission performance. Further information, including all presentations, background papers and summaries of outcomes are available on the IEA website (see http:// www.iea.org/Textbase/work/workshopsearch.asp). This publication focuses on transmission system security in competitive electricity markets. It draws primarily from the first and second workshops, and from official inquiries into major blackouts in Europe, North America and Australia in 2003 and 2004. Chapter 1 provides a brief introduction to the nature and scope of transmission system security issues and considers the implications of electricity market reform for transmission reliability and system operation. Chapter 2 presents four case studies drawn from events in North America (August 2003), SwedenDenmark (September 2003), Switzerland-Italy (September 2003), and Australia (August 2004) where transmission failures led to substantial supply disruptions. Chpater 3 explores some key issues affecting transmission system security including: legal, regulatory and institutional arrangements; system operating 24

INTRODUCTION

standards and practices; communication and coordination (particularly in integrated networks spanning multiple jurisdictions and control areas); investment in technology and people; and asset management, especially maintenance of easements under transmissions lines. It also explores the potential for using market-based mechanisms to help improve transmission system security at least cost. Discussion endeavours to focus on issues from a policy perspective. Other issues relating to the adequacy of transmission infrastructure to support efficient and reliable trade and competition (investment) and creating incentives for superior transmission network operation and development (regulation) are addressed in the IEA publication Lessons from Liberalised Electricity Markets, which examines key lessons for policy analysts from liberalisation in IEA member countries over the last decade.

25

1 CONCEPTS AND CONTEXT

CONCEPTS AND CONTEXT Key Concepts ........................................................................ Blackouts focus attention on the reliability of electricity systems. Reliability is a very general term. At its simplest it can be defined as ‘keeping the lights on’. However, this relatively simple definition is of only limited value in helping to better understand the multi-faceted nature of reliability within an electricity value chain. The concept of reliability needs to be ‘unbundled’ if it is to be better understood. For instance, the reliable provision and delivery of electricity services can be affected by a range of interrelated factors including: ●

access to adequate and competitively-priced fuel supplies;

●

adequacy of existing generation and transmission capacity to meet current and future demand;

●

efficiency with which existing generation and transmission capacity is used to meet demand;

●

effectiveness of generation and network investment signals and responses to meet current and future electricity demand;

●

effectiveness of system operation;

●

effectiveness of and conformance to reliability standards; and

●

other public policy activities with implications for the security and reliability of electricity systems, such as subsidy assistance for investments in intermittent renewable generating technologies.

Reliability in this context encompasses the ability of the value chain to deliver electricity to all connected users within acceptable standards and in the amounts desired. Reliability possesses two key dimensions: adequacy and security4. Adequacy refers to the ability of the power system to supply the aggregate electric power and energy requirements of the customers within component ratings and voltage limits, taking into account planned and reasonably expected unplanned outages of system components. Security, on the other hand, refers to the ability of a power system to withstand sudden disturbances such as electric short circuits or unanticipated 4. The definitions of reliability, adequacy and system security presented draw on those developed by the North American Electric Reliability Council (NERC) and the International Council on Large Electric Systems (CIGRE).

27

LEARNING FROM THE BLACKOUTS

losses of system components or unusual load conditions together with operating constraints. In this sense, security can be thought of as referring to the operational reliability of an existing power system. This security dimension can also have several facets. For instance, security can refer to the resilience of electricity systems to various forms of external threat, such as cyber or physical attack. It also incorporates the notion of system integrity, which refers to the preservation of interconnected system operation, or the avoidance of uncontrolled separation, in the presence of severe disturbances. Discussions about the reliability of electricity systems tend to focus on issues affecting the capacity of electricity systems to generate and deliver sufficient electricity to meet demand under various conditions. Such discussions focus on the adequacy dimension of reliability. This publication will focus on the key policy issues associated with strengthening the operational reliability dimension of transmission system security in competitive electricity markets. 5

A Brief Outline of the System Security Challenge ............. Maintaining reliable and secure transmission system operation is a challenging exercise that reflects the physical characteristics of electricity. In particular, electrical imbalances at any point within an interconnected transmission network can have immediate and severe repercussions for the quality and deliverability of electricity throughout the whole interconnected network. As a result, supply and demand must be balanced in real time across the whole interconnected network to ensure reliable supply that meets defined voltage and frequency requirements. The challenge is magnified by the dynamic nature of flows on interconnected transmission systems, which follow the path of least resistance determined by the constantly changing interaction between generation and load. Maintaining secure and stable transmission systems is a complex and challenging balancing act. Key to success is the simultaneous balancing of electricity flows to maintain frequency and voltage subject to system stability limits and the thermal operating limits of the transmission infrastructure. Failure to do so can cause substantial damage to electricity infrastructure or lead to catastrophic failures and blackouts. In an alternating current power system, frequency represents the rate at which the direction of current changes. A frequency cycle is completed every time current reverses direction and then returns to its original direction. Power system operators aim to maintain a constant frequency. In North America, 5. See Stoft, S. (2002) and US-Canada Power System Outage Task Force (2004a) for further discussion.

28

1 CONCEPTS AND CONTEXT

frequency is maintained at 60 cycles per second (Hertz - Hz), while in Europe frequency is typically maintained at 50 Hz. Failure to balance supply and demand can cause variations in the frequency on an alternating current power network. When generation exceeds consumption power system frequency will increase, while insufficient generation to meet demand will result in falling frequency. Small deviations in frequency are common, reflecting changes in the aggregate balance between generation and load within an interconnected transmission system as individual generators modify output to meet constantly changing demand. However, large frequency deviations can cause severe damage to generating equipment. Protection systems on this equipment can in turn lead to disconnection of the equipment from the grid. This can worsen the frequency deviation, resulting in more equipment being disconnected, and this can start a cascade of disconnections that leads to a blackout. Conversely, extremely low frequencies can trigger load shedding to avoid system collapse. Voltage can be thought of as representing the amount of electrical force or ‘pressure’ in a transmission system that causes current to flow in a circuit, measured in volts. Like frequency, power system operators aim to maintain voltage at a steady level. High voltage transmission systems tend to operate from around 220 kV, with the bulk transmission system typically operating from between 300 kV and 500 kV. Changes in system conditions, such as increased real or reactive loads, high power transfers, or the loss of generation or transmission facilities can create an imbalance of reactive power supply and demand. This imbalance can lead to voltage instability or voltage collapse, which occurs when voltages progressively decline until stable operating voltages can no longer be maintained. Voltage instability can occur gradually within tens of seconds or minutes. Conversely, excessively high voltages may exceed the insulation capabilities of transmission lines, leading to dangerous electric arcs. Power flows over transmission systems need to be managed within thermal and stability limits. Electrical losses resulting from power flows over transmission systems have the effect of heating transmission lines and other infrastructure used to transport electricity and can expose transmission infrastructure to overheating. Equipment heating rates are also affected by environmental factors such as ambient temperature and wind, and by the use of reactive power. Overheating will cause transmission lines to sag, possibly leading to short circuits resulting from electric arcs or direct contact with trees or other ground-based objects. Persistent overheating can also cause permanent damage to transmission infrastructure such as melting conductors on 29

LEARNING FROM THE BLACKOUTS

transmission lines. Thermal limits are imposed by system operators to set the maximum flow of real power on transmission lines to avoid overheating transmission equipment6. Stability limits are also used to establish maximum real power flows on transmission lines. In general, thermal limits tend to be the binding constraint on power flows over shorter distances, while stability limits tend to be the binding limits for longer distance power flows. Stability limits are often defined in terms of power flow and voltage. Power flow (angle) stability limits are derived from the way power flows in an alternating current transmission system. Essentially, power flows from generation to load in an alternating current transmission line when the voltage at the load end of the line is out of phase with the voltage at the generator end. The greater this phase difference the greater the power flow. As loads take more power, more power flows down the transmission line which causes the phase difference between sending and receiving voltages to increase. However, there is a limit to the amount of power that can flow down a transmission line. Voltage is sinusoidal, which means that when the voltages at either end of a transmission line get out of phase by half a cycle (180o) or by a full cycle (360o) they are identical and no power flows. This implies that the theoretical maximum power flow on an alternating current transmission line occurs when the phase differences between voltages at either end of the line reach 90o. Transfers beyond this limit expose the system to voltage collapse. Lower limits of between 30o and 45o are typically applied, which correspond to a real power flow that sets the power flow stability limit for a transmission line. Power stability limits are also set to avoid the risk of generation and loads within an integrated transmission system loosing synchronisation. Loss of synchronisation with the common system frequency means generators are operating out of step with each other. Loss of synchronisation can damage equipment and in extreme cases lead to the breakdown of an integrated transmission system into separate electrical islands. 6. Power flows on alternating current power systems include both real power and reactive power. Real power refers to electricity that flows from generation to load to power electrical equipment. It is typically measured in kilowatts (kW) or megawatts (MW). Reactive power is that portion of electricity that establishes and sustains the electric and magnetic fields of alternating-current equipment. Reactive power must be supplied to most types of magnetic equipment, such as motors and transformers. It also must supply the reactive losses on transmission facilities. It is typically measured in kilovars (kVAr) or megavars (MVAr). Reactive power consumption tends to depress transmission voltage, while its production (by generators) or injection (from storage devices such as capacitors) tends to support voltage. Reactive power can be transmitted only over relatively short distances during heavy load conditions. If reactive power cannot be supplied promptly and in sufficient quantity, voltages decay, and in extreme cases a ‘voltage collapse’ may result.

30

1 CONCEPTS AND CONTEXT

Increasing power flows also tend to cause the voltage at the load end to decrease. Voltage reductions can be addressed by injecting reactive power, which maintains voltage levels as loads increase. However, there is a limit to this process. Once power flows reach a level that exhausts local reactive power reserves, any further flows will continue to reduce voltages. If voltage falls too low then it can begin to collapse uncontrollably. Voltage stability limits are established to avoid this risk.

Reliability Standards and Operating Practices.................... Experience indicates that effective real-time management of electricity systems can only be achieved through centralised, or centrally coordinated, system operation. System operation is generally undertaken by local transmission system owners, with a degree of coordination between them where integrated regional networks incorporate two or more transmission systems. Coordination has principally focused on the management of flows across interconnects immediately linking adjacent control areas. System operators are also generally responsible for executing emergency procedures to manage extreme events in a manner that minimises the impact on supply while protecting critical electricity infrastructure.

■

Reliability Standards Operational experience has led to the development of various reliability standards and practices to ensure that transmission systems are operated in a stable and secure manner. Given the nature of electricity and the potential for quick and catastrophic system-wide failure, these standards are designed to ensure that appropriate and sufficient resources are available to enable system operators to respond rapidly to manage extreme operating conditions or emergency events. From an operational system security perspective, probably the most important of these reliability protocols is the N-1 standard. A power system can be described as being N-1 secure when it is capable of maintaining normal operations (ie. reliably delivering electricity of a given frequency and voltage subject to technical limits) in the event of a single contingency event, like the unplanned loss of a transmission line, generator or transformer. This standard has been adopted by system operators around the world to inform operational contingency planning, to guide management of system operation and to guide emergency efforts to return systems to a secure and stable operating 31

LEARNING FROM THE BLACKOUTS

condition within a reasonable time following a single contingency event, usually within 15 to 30 minutes. This standard was adopted because all power systems are regularly affected by unpredictable faults and failures, such as lightning strikes or mechanical failure. Since such events are unavoidable, all power systems need to be able to endure them without unduly disrupting supplies. Blackouts can only be avoided in such circumstances where systems are operated with a sufficient security margin. This implies that there must be enough reserve capacity (either generation, demand response or imports) available to accommodate the loss of a generating unit and enough transmission capacity to accommodate changes in power flows resulting from the failure of a transmission line7. In principle, the N-1 standard is applied in a deterministic manner, which means that it is applied in the same way at all points within a system operator’s control area irrespective of the probability or impact of a component failure. Application of this standard to achieve absolute levels of reliability would require a level of resource commitment that is neither practical nor cost-effective. In practice, the N-1 standard is typically applied in a manner that reflects operational resource constraints. As a result, the N-1 standard is typically applied to manage credible contingency events. Credible contingency events in this context typically exclude multiple independent failures within the period normally allowed to return a transmission system to a secure operating condition following an N-1 event. It is also applied in a manner that reflects system operator judgement and experience. For instance, multiple dependent failures are sometimes accommodated in particular cases where system operators consider the risk of their occurrence to be substantial. Probability risk assessment is sometimes used to help identify credible contingencies.

■

Operating Practices Operating practices are designed to ensure that transmission systems are operated within the security bounds established by the N-1 standard. Typically, these practices are built around an iterative process that generally involves contingency assessment and planning, on-going monitoring of system operation and intervention as required to maintain system security. Contingency assessment and planning is the first step and involves analysis and assessment of expected operating conditions based on anticipated power flows within and between system operator control areas. In competitive 7. Krischen, D. and Strbac, G. (2004).

32

1 CONCEPTS AND CONTEXT

electricity markets, operational contingency assessments are typically carried out for each trading interval or operating period, and in some cases for each dispatch interval. Initial assessments are generally carried out day-ahead, based on the expected availability of infrastructure and demand, including power flows associated with bilateral contracts. This initial assessment is then updated to account for proposed day-ahead spot market dispatch bids and offers, to the extent that they can be accommodated within the projected delivery capacity of the transmission system. Contingency assessments are revised regularly in the period leading up to dispatch to incorporate new information that could significantly affect power flows, such as changes in the availability of generation or transmission lines and dispatch patterns resulting from intra-day trading or balancing market outcomes. This information is fed into a computer simulation of a transmission system to identify potential points of network congestion, and to determine the type, location and amount of technical reserves and other resources a system operator may require to prepare for credible N-1 contingencies. Similar analysis is used to decide whether the merit-order dispatch determined on the basis of spot market trading may need to be modified to ensure secure system operation. The outcome of such analysis is referred to as security constrained dispatch and is often used to manage transmission system security in the event of persistent transmission network congestion. Reserve requirements may vary according to the nature and design of a transmission system. Radial transmission systems or other weakly integrated transmission systems, for instance, may be more exposed to the impact of transmission network failures, particularly the failure of interconnectors, which have the potential to immediately isolate portions of an otherwise integrated transmission system. In such circumstances, additional generation and network operating reserves may be required to address a higher risk of electrical islanding compared to a highly meshed transmission system where more transmission paths exist to carry displaced inter-regional power flows. This is a particular issue in integrated transmission systems with relatively weak interconnections, like the Australian National Electricity Market. System operators monitor transmission systems during the operating period to ensure that secure operating conditions are maintained and so that they can respond in a timely and effective manner to emergency events. Operational management of transmission systems relies on real-time and near real-time information on power flows against technical limits at strategic points in transmission networks. Contingency assessment is typically updated throughout an operating period to accommodate actual operating conditions 33

LEARNING FROM THE BLACKOUTS

as they evolve. This practise is particularly important in the context of managing unpredictable changes in power flows resulting from real-time movements in load and unanticipated changes in the availability of generation and transmission network equipment. When an emergency or N-1 contingency event occurs, system operators need to be able to intervene in a timely and effective manner to stabilise a transmission system and then return it to an N-1 secure state within the maximum period permitted by the reliability standards. Reliability rules usually allow system operators between 15 and 30 minutes to return a transmission network to an N-1 secure condition following an N-1 event. The typical control framework adopted by system operators to manage emergency conditions and N-1 events and return systems to a secure operating condition is summarised in Figure 1.

Figure 1

Dy Liacco System Security Model Normal Operating Condition

Restoration Process

Alert Condition

Extreme Event or System Failure

Emergency Condition

Controled Transition

Uncontroled Transition

Source: Dy Liacco, T. E. (1967).

A coordinated approach is required to ensure system security in transmission systems that span multiple system operator control areas, given the potential for cascading failures to rapidly spread and cause system-wide failures. Coordination and communication is critical for successful contingency assessment and planning, for effective system condition monitoring and for effective operator intervention to manage emergency events. Successful coordination in this context is built on effective information sharing. Information systems and protocols have been established to facilitate information exchange. Examples include the Day 34

1 CONCEPTS AND CONTEXT

Ahead Congestion Forecast in Continental Europe or the data provided through the Open Access Same Time Information System in North America. Information is typically shared on expected and actual power flows across interconnectors between neighbouring control areas and on the availability of generators and transmission lines. Coordination has been supported by specific institutional arrangements in some jurisdictions, such as the reliability coordinator framework that operates in North America under the auspices of the North American Electric Reliability Council and affiliated regional reliability councils.

■

Operating Tools System operators have traditionally used a combination of methods to manage transmission system security including various forms of technical or operating reserves and services, redispatch and load shedding. The specific products, or ancillary services, available to help manage system security are usually defined in terms of their function and the time taken to deploy them. Key functions include frequency control, network control and provision for restoration of services following a blackout, usually referred to as black start services. System operators would typically deploy operating reserves or employ redispatch before load shedding. Load shedding is usually treated as a last resort, in order to avoid catastrophic system failures. Products used to manage small routine imbalances and to react immediately to emergency events are deployed automatically in response to particular frequency or voltage triggers, while other services tend to be deployed manually by system operators as required. However, these products provide system operators with only indirect control over power flows. In an alternating current transmission system, electricity flows according to the laws of physics along the path of least resistance. Hence, system operators must manage the mix of inputs – generation and load – to deliver frequency, voltage and power flows that are consistent with secure system operation. System operators normally have access to a range of operating reserves including: ●

Frequency Control Regulation. These reserves are used to manage small movements in frequency resulting from the constantly changing balance between generation and load on an integrated transmission network. They are automatically dispatched.

●

Frequency Control Contingency Reserves. This is essentially the ‘spinning’ reserve, which is provided by power plants with turbines that are 35

LEARNING FROM THE BLACKOUTS

spinning in synchronisation with the common system frequency but are not generating power. Such capacity can provide an immediate and significant injection of power if required. Spinning reserves can typically be ramped up to full production in less than 10 minutes. ●

Fast Response Active Reserves. These are essentially ‘non-spinning’ reserves which can be deployed in a matter of minutes and be ramped up to full production within an hour. Spinning and non-spinning reserves are used to maintain services and to restore the balance between generation and load in the event of a sudden substantial generation or network outage.

●

Slow Response Active Reserves. These reserves are typically employed in response to an unanticipated generation or network failure where sufficient advance notice is provided, or in response to a persistent emergency situation. Such reserves can usually be deployed within 4 to 8 hours.

●

Reactive Power Reserves. These reserves provide reactive power to support voltage stability and power flows. Reactive power diminishes rapidly over relatively short transmission distances and must be provided locally. Reactive power can be provided by generators and by purposespecific equipment such as capacitors.

System operators also procure black start services to facilitate system restoration following a blackout. These services are typically secured through bilateral contracts with generators, though system operators usually have some power to requisition necessary services in an emergency. In principle, demand reductions could act as a substitute for generation reserve. Demand response therefore has the potential to provide many of these services. However, its contribution is likely to be limited to those services where there is relatively more time to respond.

Electricity Markets Help to Strengthen Transmission System Security.............................................. Access to reliable and affordable electricity is a key determinant of economic growth, international competitiveness and community prosperity. Recognising the important role of electricity in modern economies, many governments have pursued electricity market reform in an effort to improve efficiency and economic prosperity. 36

1 CONCEPTS AND CONTEXT

Reliable and effective performance of transmission networks will shape the operation and development of efficient wholesale electricity markets, particularly the emergence of effective regional markets that promote efficient price formation and trade. Potential economic advantages from effective electricity market reform include: ●

efficient development of inter-regional trade which may increase effective competition and reduce the scope for market power abuse;

●

improved capacity utilisation (both generation and networks) and deferral of expensive infrastructure investments; and

●

greater transparency and more efficient price formation.

Substantial efficiency gains have resulted from electricity reforms undertaken in IEA countries. Benefits from electricity reform are more fully discussed in the IEA publication Lessons from Liberalised Electricity Markets. More integrated and efficient electricity markets also have the potential to strengthen system reliability at least cost. Efficient interregional trade enables more effective sharing of reserve capacity, allowing system operators to draw on the reserves and resources of adjacent control areas to ensure reliable and secure system operation. In particular, it can improve management of frequency control and provide access to additional generating capacity to help stabilise the supply-demand balance during emergencies. The potential benefits for reliability are enhanced where greater integration allows for more effective sharing of complementary generation technologies. For instance, greater integration of hydro-based power systems with thermalbased power systems can strengthen security of supply in the hydro-based system, especially during a drought. On the other hand, greater integration of a hydro-based system with a thermal-based system can improve the operational flexibility of the thermal system enabling more effective responses to real-time emergencies. The reliability benefits of such integration were clearly demonstrated during the 2002-03 winter in the largely hydro-based Nordic market, which was able to maintain uninterrupted supplies despite experiencing a 1 in 200 year drought. Effective integration can also facilitate timely restoration following the loss of services by augmenting black-start responses. In some cases it may help to supplement local generation to provide voltage support. In short, greater integration afforded by electricity market reform helps to strengthen the resilience of power systems, which can improve overall system reliability. 37

LEARNING FROM THE BLACKOUTS

Electricity markets also provide greater transparency, which can help to strengthen accountability mechanisms designed to ensure that unbundled electricity sectors make adequate provision for system security. They can also create powerful incentives for more efficient use of transmission systems, consistent with effective management of system operation. Efficient markets also have the potential to help value transmission system security, particularly where greater demand responsiveness can be harnessed.

Electricity Markets are Changing the Way Transmission Systems are Used........................................... Electricity market reform has brought unbundling and independent, decentralised decision-making. As a result, decisions relating to network use and investments affecting network operation and performance that were once made in a centrally coordinated way within vertically integrated utilities are made by a number of independent market participants. Decentralised decision-making has fundamentally changed utilisation of transmission networks. Previously stable and relatively predictable patterns of network use have in many cases been replaced with less predictable usage, greater volatility of flows and greater use of long-distance transportation, reflecting growing inter-regional trade. No existing major transmission systems have been designed for such use. Table 1 summarises several key factors affecting the changing pattern of transmission system use, which have implications for managing transmission system security. A particular feature of reformed electricity markets is the growth in interregional electricity transmission. Figure 2 illustrates the substantial growth in inter-regional electricity transfers within the Nordic market, where interregional trade has risen from around 10.4% of total Nordic load in 1997 to around 16.0% of total load in 2004, an increase of over 50%.

38

1 CONCEPTS AND CONTEXT

Table 1

Emerging Conditions Affecting Transmission System Reliability Pre-reform Conditions

Post-reform Emerging Conditions

Fewer, relatively large resources

Smaller, more numerous resources

Long-term, firm contracts

Contracts shorter in duration More non-firm transactions, fewer long-term firm transactions

Bulk power transactions relatively stable and predictable

Bulk power transactions relatively variable and less predictable

Assessment of system reliability made from stable base (narrower, more predictable range of potential operating states)

Assessment of system reliability made from variable base (wider, less predictable range of potential operating states)

Limited and knowledgable set of utility players

More players making more transactions, some with less interconnected operation experience; increasing with retail access

Unused transmission capacity and high security margins

High transmission utilization and operation closer to security limits

Limited competition, little incentive for reducing reliability investments

Utilities less willing to make investments in transmission reliability that do not increase revenues

Market rules and reliability rules developed together

Market rules undergoing transition, reliability rules developed separately

Limited wheeling

More system throughput

Source: United States-Canada Power System Outage Task Force.

39

LEARNING FROM THE BLACKOUTS

Figure 2

Growth in Inter-regional Power Exchanges in the Nordic Electricity Market: 1997 to 2004 Percentage of Total Load 18%

17,1%

16% 14% 12% 10%

16,0%

15,4% 13,5%

13,7%

2000

2001

11,2% 10,4%

10,1%

1997

1998

8% 6% 4% 2% 1999

2002

2003

2004

Trend Line Source: Nordel Annual Reports 1997-2004.

Inter-regional trade is increasingly driven by the relative cost of electricity across interconnected regions. In general, merit-order dispatch will create electricity flows from generation in low priced regions to load in high priced regions until prices are equalised or congestion on transmission lines prevents transfers. Patterns of network usage tend to reflect regional rather than local supply and demand, resulting in patterns of use that can create transmission congestion even during off-peak times. Transmission systems are being run more efficiently, which means that they are also being run closer to their security limits for longer periods than generally seen before reform. Recent system adequacy assessments conducted in North America by the North American Electric Reliability Council (NERC) and in Continental Europe by the Union for the Coordination of Transmission of Electricity (UCTE) have noted a reduction of spare capacity margins on bulk transmission systems. With inter-regional trade expected to continue to grow at a greater rate than transmission capacity expansion, the prognosis is for margins to continue to tighten. This implies that less spare capacity will be available to support transmission system security in future, with transmission systems facing growing congestion and longer periods of operation at or near their security limits. 40

1 CONCEPTS AND CONTEXT

These trends are reflected in transmission loading relief (TLR) data for the North American Eastern Interconnection, presented in Figure 3. TLRs provide an administrative mechanism for managing potential transmission system security violations on key elements of the bulk transmission system, such as flows that might overload transmission lines or threaten stability limits. Reliability coordinators can use TLRs to reduce such power flows. Requests are normally activated within 30 to 60 minutes of being received. Figure 3 shows a substantial growth in the total number of monthly TLRs between July 1997 and July 2005, consistent with growing congestion and running the transmission system closer to its security limits to accommodate increasing levels of inter-regional trade. Figure 3

Monthly Transmission Loading Relief Requests: 1997 to 2005 Number of Logs 350 300 250 200 150 100 50

Ju l.9 No 7 v M .97 ar Ju .98 l.9 No 8 v M .98 ar Ju .99 l.9 No 9 v M .99 ar Ju .00 l.0 No 0 v M .00 ar Ju .01 l.0 No 1 v. M 01 ar Ju .02 l. No 02 v. M 02 ar Ju .03 l.0 No 3 v M .03 ar Ju .04 l.0 No 4 v M .0 4 ar Ju .05 l.0 5

0

12 Month Rolling Average

Logs

Source: North American Electric Reliability Council.

More volatile and less predictable power flows have also emerged with electricity market reform. Greater volatility reflects, in part, the increasing proportion of flows that are being determined by independent agents responding to differences in inter-regional prices, which in turn drives flows across integrated networks. These trends are facilitated by transparent wholesale markets that reveal regional price variations and facilitate efficient and timely inter-regional responses based on merit-order dispatch of generation. 41

LEARNING FROM THE BLACKOUTS

These trends are revealed in Figure 4, which compares power flows across the interconnection between Norway and West Denmark before and after the introduction of electricity reform in Denmark. This comparison suggests an increase in the magnitude and volatility of inter-regional power flows following the introduction of electricity reform. Figure 4

Power Flows between West Denmark and Norway: 1995 and 2000 MWh 1 200 Denmark West – Norway 900 600 300 0 –300 –600 –900 January 1995

–1 200

January 2000

Source: Energinet.dk.

Volatility of power flows is also reflected in Figure 5, which charts the difference between day-ahead scheduled hourly power flows and actual hourly power flows on interconnections with the region controlled by the New England Independent System Operator for August 2005. In this example, the discrepancies recorded between day-ahead scheduled and actual power flows were regularly quite large, with 419 hours during August 2005, or 56% of total hours, recording discrepancies of over 10%, and 84 hours, or 11% of total hours, recording differences of over 50%. Such considerable differences between day-ahead and actual power flows indicate that inter-regional power flows in competitive electricity markets can exhibit considerable real-time volatility, creating a very dynamic system operating environment with less predictable power flows. Large-scale introduction of less greenhouse gas intensive forms of intermittent generation, such as wind power, have the potential to further promote volatility and reduce predictability of transmission flows. Intermittent 42

1 CONCEPTS AND CONTEXT

Figure 5

Difference between Day-ahead Scheduled and Actual Metered Power Flows on all Interconnectors to the New England Independent System Operator Region: August 2005 Difference (MW) 400

200

0

–200

–400

–600 Source: New England Independent System Operator.

Hours

generation has the potential to drive unscheduled power flows which can create challenges for maintaining transmission system security in real time. It may also create reverse power flows from a distribution system to the transmission system, which can make managing transmission system security considerably more complex in regions with a considerable volume of intermittent generation8. However, the implications of intermittent generation for transmission system security are likely to depend on the volume of such generation and the nature of the wholesale market design. The implications are likely to be relatively low where intermittent generation represents a relatively small proportion of total generation or where considerable excess transmission capacity exists. Similarly, spot markets that permit trade closer to real-time dispatch may have the potential to respond in a more timely and efficient way to manage any imbalances resulting from forecasting errors regarding the availability of intermittent generation than day-ahead markets. More extensive integration of intermittent generation into market dispatch, including exposure to related balancing costs, may strengthen incentives for 8. Elkraft System (2004a).

43

LEARNING FROM THE BLACKOUTS

intermittent generators to more effectively manage their output, which could improve the predictability and manageability of these flows from a transmission system security perspective. Scope may also exist to extend the application of technical generator standards to strengthen the contribution intermittent generators can make to supporting transmission system stability9.

More Dynamic Operating Conditions Magnify Challenges for Managing Transmission System Security ...................... Electricity market reform is changing the way transmission systems are used. Greater regional market integration and inter-regional trade are leading to longer distance, less predictable and more volatile electricity flows. These trends are combining to reduce ‘spare’ transmission capacity while also creating a more dynamic system operating environment. Together, they create new challenges for managing and maintaining transmission system security, with the focus shifting more toward real-time monitoring and management. Greater amounts of reactive power are required to provide the voltage support needed to sustain increasing inter-regional power flows. Reactive power is a locally provided resource. Increasing transmission system use has the potential to rapidly exhaust reactive power resources, placing systems at or near their voltage stability security limits and increasing their exposure to voltagerelated instability phenomena10. However, probably the most significant emerging challenge from a transmission system security management perspective is the growth of unscheduled parallel power flows associated with increasing inter-regional electricity exchanges. Parallel power flows, also referred to as loop flows, reflect the physics of alternating current transmission systems, where electricity flows uncontrollably along the path of least resistance from generation to load. In highly meshed transmission systems serving competitive electricity markets, the physics of power flows can combine with the constantly changing mix of generation and load to create significant and complex unscheduled power flows. Such flows have the potential to substantially weaken transmission system security if they are not fully taken into account by system operators. Figure 6 provides an example of how a power flow from a utility in Wisconsin to the Tennessee Valley Authority would travel across transmission lines in the Eastern Interconnection. A colour contour shows the percentage of the 9. UCTE (2005b). 10. UCTE (2004a).

44

Source: Hauer, J. et al. (2002)

2%

5%

Max % of power transfer carried per transmission line 25 %

Parallel Power Flows on a Transfer from Wisconsin to Tennessee

Figure 6 1 CONCEPTS AND CONTEXT

45

LEARNING FROM THE BLACKOUTS

transfer that would flow on lines carrying at least 2% of the flow. The figure indicates that a single transfer can simultaneously affect flows on many transmission lines in several control areas. Figure 7 provides a similar example tracing the path of power flows associated with a 1000 MW transfer between northern France and Italy. In this example a little over one-third (38%) of power would flow directly from France to Italy, while nearly two-thirds (62%) would flow along different parallel routes, including around 15% flowing through Belgium and the Netherlands. Managing transmission system security in an environment of significant parallel flows is a daunting challenge. Independent decentralised decisionmaking can impose thousands of simultaneous transactions on an integrated transmission system, leading to mutual interference and congestion. Mitigating such congestion can be a technically difficult task, particularly in real time. Effective responses require holistic, real-time information on power flows across an integrated transmission system; information which is not always available to local system operators and hence not always fully reflected in contingency planning and assessment, or in system operating dispatch decisions. The transmission system security challenge is magnified where transmission paths cross multiple control areas. An effective, coordinated response is required to successfully manage transmission system security in these circumstances. To succeed, system operators need effective real-time management procedures, operating experience, holistic modelling for operational contingency assessment and planning purposes and integrated data resources. Present systems and operating practices may not be sufficient to meet the challenge and may leave system operators unaware of the full extent of parallel flows, increasing the risk of flows not being appropriately considered in the context of operational contingency assessment and planning. Such failures would have the potential to significantly degrade system operator real-time situational awareness and management, with the potential to jeopardise transmission system security11. Parallel flow effects have been identified as a contributing factor in both the US-Canadian and Swiss-Italian blackouts of 200312. These events are further discussed among the case studies presented in Chapter 2. Parallel flows make predicting power flows following an N-1 contingency event more difficult, magnifying the challenge of managing emergency situations. Good situational awareness and effective coordination are required to ensure 11. Hauer, J. et al. (2002). 12. See US-Canada Power System Outage Task Force (2004a); the various investigations into the Swiss-Italian event; Bialek (2004).

46

1 CONCEPTS AND CONTEXT

Figure 7

Parallel Flows Associated With a 1 000 MW Transfer from France to Italy

POLAND

UK NETHERLANDS

15 GERMANY

15 BELGIUM

CZECH REP.

22

21 25 FRANCE

12

AUSTRIA

5

6

5

SWITZERLAND

SLOV.

6 38

51 CROATIA ITALY

SPAIN

Generation 1000 MW Consumption 1000 MW Percentage share in transit 22

Source: Haubrich, H-J. and Fritz, W. (1999)

47

LEARNING FROM THE BLACKOUTS

an appropriate response to manage emergency events, particularly within integrated transmission systems spanning multiple control areas. In the emerging more dynamic operating environment, real-time security monitoring and automatic information exchange between system operators is required to effectively manage and coordinate emergency responses. At the same time, unbundling has also reduced transmission system operators’ capacity to manage system balancing through coordinated actions across the value chain, while also magnifying their exposure to systemic risk13. The emergence of regional markets that span multiple control areas has added to the transmission system security challenge by increasing each system operator’s exposure to the operational decisions of other system operators and to potential failures outside their control area. Increased exposure to systemic risk may have implications for the frequency and nature of extreme or emergency events, and how best to manage them. In this more integrated and dynamic operating environment, an event affecting a relatively distant part of an integrated transmission network may have greater potential to interrupt the delivery of electricity throughout an interconnected network and severely disrupt the supply and operation of electricity markets.

A More Dynamic Operating Environment May Also Affect the Nature and Frequency of Outages ................................ Large blackouts are by no means a new phenomenon. Though relatively rare, they are never completely avoidable. Major blackouts have been the catalyst for significant reforms to strengthen transmission system security in the past. Table 2 documents some of the major transmission-related outages recorded in North America between 1965 and 2002. In general, small service disruptions are relatively common events in most power systems, with large events remaining relatively rare. Weather-related events are the most common cause of outages on transmission systems. However, analysis of North American outage data by the Electric Power Research Institute (EPRI)14 suggests that the frequency of larger disturbances may be increasing. Between 1991 and 1995, 41 events were recorded that affected over 50 000 consumers, with over 355 000 consumers affected on

13. Systemic risk in this context refers to the impact of a failure in one part of the value chain on the operational performance of other parts of the value chain. 14. EPRI (2003a).

48

June 1998

Upper Midwest Blackout

49

Source: US-Canada Power System Outage Task Force.

152,000

August 1996 Arizona, California, Colorado, Idaho, Montana, Nebraska, 7.5 Million Nevada, New Mexico, Oregon, South Dakota, Texas, Utah, Washington, and Wyoming (USA); Alberta and British Colombia (Canada); and Baja California (Mexico)

West Coast Blackout

Minnesota, Montana, North Dakota, South Dakota and Wisconsin (USA); Ontario, Manitoba and Saskatchewan (Canada)

2 Million Arizona, California, Colorado, Idaho, Montana, Nebraska, Nevada, New Mexico, Oregon, South Dakota, Texas, Utah, Washington, and Wyoming (USA); Alberta and British Colombia (Canada); and Baja California (Mexico)

July 1996

West Coast Blackout

6,000

20,000

950

Over 28,000

11,850

Over 5 Million 12,350

Pacific North West

December 1982

9 Million

30 Million

Up to 19 hours

A few minutes to 9 hours

A few minutes to several hours.

N.A.

26 hours

13 hours

Customers Load Lost Duration Disconnected (MW)

West Coast Blackout

New York City

July 1977 New York City Blackout

Area Affected

New York, Connecticut, Massachusetts, Rhode Island, parts of Pennsylvania, northeastern New Jersey (USA); and parts of Ontario (Canada)

Date

Great North- November 1965 eastern Blackout

Event

Some Major North American Blackouts: 1965 to 2002

Table 2 1 CONCEPTS AND CONTEXT

LEARNING FROM THE BLACKOUTS

average by these events. By comparison, between 1996 and 2000, 58 events affecting over 50 000 consumers were recorded, with nearly 410 000 consumers affected on average. The total number of events affecting more than 50 000 consumers rose by around 41% during the second half of the 90’s compared to the first half, while around 15% more consumers were effected by these disturbances, on average, between 1996-2000 compared to the period 1991-95. Figure 8 provides a frequency distribution of these events.

Figure 8

Frequency Distribution of North American Outages Affecting More than 50 000 Consumers: 1991-95 and 1996-2000 Number of Occurences 100

1996–2000 Outages • 58 Occurrences over 50 000 Consumers • 409 854 Average Consumers

10

1991-1995 Outages • 41 Occurrences over 50 000 Consumers • 355 204 Average Consumers 1 10 000 Source: EPRI

100 000

1 000 000

10 000 000

Number of Affected Consumers

EPRI’s analysis of this data by quantity of load lost reveals similar trends. Between 1991 and 1995, 66 events occurred that resulted in the disconnection of over 100 MW of load, with an average loss of 798 MW per event. By comparison, between 1996 and 2000, 76 similar events were recorded with an average loss of 1,067 MW per event. The total number of events during the second half of the 90’s exceeded those recorded during the first half of the 90’s by around 15%, while the average loss for the period 1996-2000 exceeded the 1991-95 period by around 34%. Figure 9 provides a frequency distribution of these events.

50

1 CONCEPTS AND CONTEXT

Figure 9

Frequency Distribution of North American Outages Affecting More than 100 MW: 1991-95 and 1996-2000 Number of Occurences 100

1996–2000 Outages • 76 Occurrences over 100 MW • 1 067 Average MW

10 1991-1995 Outages • 66 Occurrences over 100 MW • 798 Average MW

1 100

1 000

Source: EPRI.

10 000

100 000 MW Lost

International Energy Agency analysis of a sample of North American outage data by load lost and duration of outage suggests similar trends. This analysis compares a sample of events that occurred between 1990 and 1994 with a sample of more recent outages that were recorded between 2000 and 2004. The results, presented in Figure 10, may suggest a trend toward larger outages with slightly longer durations. This might be partially explained by growing integration of regional electricity markets and increasing inter-regional electricity flows over the period, which may have helped to spread the impact of disturbances. It may also reflect efforts to mitigate small disruptions, which have the potential to increase the frequency of larger outages due to the complex dynamic relationship between smaller and larger disturbances15. However, care should be exercised in drawing conclusions from this data. Many factors have influenced and possibly biased these results. A more thorough examination of the data to identify and remove any biases would be necessary before drawing more definitive conclusions.

15. For further discussion of this phenomenon see Carreras, B. et al. (2003).

51

LEARNING FROM THE BLACKOUTS

Figure 10

Selected North American Outages: 1990-94 and 2000-04 Duration (Minutes) 100 000

10 000

1990-94

1 000

100 2000-04 10

1 1

10

100

1 000

10 000

100 000

Lost Load (MWs) 1990-94

2000-04

Source: United States Department of Energy Disturbances Database.

Analysis of previous large-scale outages indicates that most occur when transmission systems are operating at or close to their security limits16. When transmission systems are operating at their security limits, an N-1 event immediately creates an unstable operating environment and increases a system’s exposure to multiple dependent contingencies. Multiple dependent contingencies can be characterised as internal failures that occur as a direct result of an earlier N-1 event. Examples include overloads that trip transmission lines and other equipment, inappropriate operation of protection devices that unexpectedly trip equipment, or system operator error. Such dependent events have the potential to rapidly transform a credible N-1 contingency event into a cascading system failure culminating in a large-scale blackout. Electricity reform has led to more efficient use of transmission systems, and greater regional integration of power flows resulting from inter-regional trade. As a result, transmission systems are being run closer to their security limits for longer periods than in the past. Although this does not directly affect a

16. Analysis of previous large-scale events is provided in Knight, U. G. (2001).

52

1 CONCEPTS AND CONTEXT

system operator’s ability to operate a transmission system in an N-1 secure state (given that providing adequate operating reserves to manage credible N-1 events always takes precedence over other uses of transmission resources), it does increase exposure to the multiple dependent contingency phenomenon. Greater integration resulting in longer distance power flows also has the potential to enable disturbances to spread more rapidly through a transmission system, leading to larger-scale blackouts17. Increasing complexity and volatility of power flows is creating a more dynamic real-time operating environment that increases the challenges associated with managing transmission system security, particularly during emergency events. It also has the potential to increase the risk of operator error in response to emergency events, which can result in multiple dependent contingencies and cascading system failures. These risks are likely to be magnified in integrated transmission systems spanning multiple control areas.

Reliability Standards and Operating Practises Must Change to Meet the New Challenges ......................... Electricity market reform is substantially changing the operating environment for maintaining transmission system security. Inter-regional trade is creating longer distance power exchanges, while independent decentralised decisionmaking is leading to less predictable power flows. Growing demand for transmission capacity to accommodate inter-regional trade means that transmission systems are increasingly being run at or near their security limits. These fundamental changes are creating a far more complex and dynamic operating environment from a transmission system security perspective. Despite these fundamental changes, development of regulatory arrangements, reliability standards and system operating practices relating to transmission system security generally have not kept pace with the new real-time system operating challenges resulting from electricity market reform. The major blackouts of 2003 and 2004 raised fundamental questions about the appropriateness of these standards and practices in this new operating environment. Examination of operating standards and practices is being undertaken in North America and in Europe in the wake of these disturbances. But the focus has been on technical matters. Failures of the kind experienced in 2003 and 2004 also raise strategic issues for policymakers.

17. See Hauer, J. et al. (2002) and Carreras, B. et al. (2003) for further discussion of the multiple dependent contingency and cascading failures phenomena.

53

The time has come for policymakers to engage in this debate to ensure the development of integrated and comprehensive responses that will enhance transmission system security in competitive electricity markets.

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

TRANSMISSION SYSTEM SECURITY CASE STUDIES Introduction .......................................................................... Several substantial supply disruptions involving a failure of network services occurred in North America, Europe and Australia during 2003 and 2004. Four of these events are featured in the following case studies: ●

the 14 August 2003 event affecting the north-eastern United States and south-eastern Canada;

●

the 28 September 2003 event affecting southern Switzerland and Italy;

●

the 23 September 2003 event affecting southern Sweden and eastern Denmark; and

●

the 13 August 2004 event affecting the National Electricity Market in eastern Australia.

These events clearly demonstrate the fundamental importance of reliable network services for the efficient and secure operation of electricity markets and at the same time highlight the vulnerability of electricity markets to network failures. They provide a valuable practical context for identifying and understanding the key issues for policymakers.

CASE STUDY 1: The North-eastern United States and South-eastern Canada................................................... The largest supply disruption in North American history struck at around 4.13 p.m. on 14 August 2003. It affected eight states in the Midwest and the Northeast of the United States and the Canadian province of Ontario. This event had two distinct phases. It began with a combination of failures and system operator inaction in Ohio between 12.15 pm and 4.06 pm, which led to the progressive tripping of key 345 kV lines in the Cleveland-Akron area and in northern Ohio. The failure of the Sammis-Star 345 kV line in Ohio at 4.06 pm triggered the regional cascade phase of the event: a seven minute period during which generator and transmission line trips spread from the Cleveland-Akron area in Ohio across most of the Northeast of the United 55

LEARNING FROM THE BLACKOUTS

States and Ontario. The following event summary is drawn from the final report of the US-Canada Power System Outage Task Force (the Task Force)18. A timeline identifying the sequence of key events in the Ohio phase of the blackout is presented in Figure 11. The Ohio phase started at 12.15 pm when erroneous input data rendered the Midwest Independent System Operator’s (MISO’s) state estimator19 ineffective. The state estimator and real time contingency analysis tools were effectively out of service between 12.15 pm and 4.06 pm. Without an effective state estimator and with its normal automatic operation disabled until 2.40 pm, MISO could not perform effective contingency analysis within its reliability area, preventing timely ‘early warning’ assessments of system status and reliability. At 1.31 pm, First Energy’s (FE) Eastlake 5 generation unit tripped and shut down automatically. Eastlake 5 is a major provider of active and reactive power within the FE service area and its loss combined with the unavailability of other substantial local generators significantly increased the potential for transmission line overloads. Following the loss of Eastlake 5, FE’s operators did not perform a contingency analysis to determine whether the loss of further transmission lines or generating capacity would put its system at risk. Subsequent analysis by the Task Force indicated that the loss of Eastlake 5 was a critical step in the sequence of events. It is virtually impossible to manually monitor all events and power system conditions simultaneously. Hence, system operators rely heavily on automatic alarms to provide early warning of events and changing system operating conditions that need to be addressed. Starting around 2.14 pm, FE’s control room lost some of its key system monitoring equipment. First to fail were the audible and on-screen alarm systems and alarm logs. FE’s system operators were unaware of the failure, worsening their capacity to respond to a rapidly degrading situation. FE’s alarm system was not restored until after the blackout.

18. See US-Canada Power System Outage Task Force (2004a). Chapter 5 details the Ohio phase while Chapter 6 details the regional cascade phase of the event. 19. A state estimator is a computer program that models the transmission system. It is used to confirm that the monitored electric power system is operating in a secure state by simulating the system both at the present time and one step ahead, for a particular network topology and loading condition. With the use of a state estimator and its associated contingency analysis software, system operators can review each critical contingency to determine whether each possible future state is within reliability limits.

56

Grid Events

Computer Events

13.00

12.15

14.02

14.32

15.05

15.00

14.41

14.54

15.39

AEP & PJM work TLR

FE EMS doesn’t update

15.42

15.45 AEP calls FE re: multiple line overloads

15.36

FE operator tells IT alarms out

15.46-15.59

FE manning substations

15.48

FE says system could fail

15.46

FE calls MISO re: line outages

15.57

PJM calls FE re: Star-South Canton

15.56

FE IT rebooting H4, H1

16.05 Sammis-Star 345 kV line fails

MISO calls FE re: Star-Juniper

15.35

15.42 on 15 more FE 138 kV lines fail

15.41

15.19

14.32

16.00 Phase 4 Collapse of 138 kV system 15.39-16.08

Star-S Canton 345 kV fails (tree)

First FE 138 kV line fails

AEP again calls FE re: Star-S Canton

FE back-up EMS server fails

15.08

Hanna-Juniper 345 kV fails (tree)

15.32

Phase 3 FE 345 kV line failures 15.05-15.57

FE primary EMS server restarts

Harding-Chamberlin 345 kV fails (tree)

FE primary EMS server fails

AEP calls FE re: Star-S Canton trip

FE loses half its remote consoles

14.20

FE EMS alarms fail

Star-S Canton 345 kV trips & recloses (tree)

14.27

Phase 2 FE’s computer failures 14.14-15.59

Key Elements of the Ohio Phase of the Blackout

14.14

Stuart-Atlanta 345 kV fails (tree)

14.02

14.00

Stuart-Atlanta 345 kV line trip confuses MISO SE

Eastlake 5 trips

13.31

Phase 1 A normal afternoon degrades 12.15-14.14

MISO SE problems begin

12.00

Source: US-Canada Power Outage Task Force (2004a).

Human Events

Figure 11 2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

57

LEARNING FROM THE BLACKOUTS

Shortly thereafter, several remote terminals that relay real-time SCADA data20 to the energy management system (EMS)21 failed, further degrading system operators’ capacity to monitor system performance. Finally, the primary server hosting the EMS’s alarm applications failed, and the back-up system failed shortly thereafter. The combination of these failures disabled Automatic Generation Control between 2.54 pm and 3.08 pm and again between 3.46 pm and 3.59 pm. It also slowed other EMS functions. FE system operators failed to detect the tripping of facilities essential to maintaining system security in their control area. Unaware of the loss of alarms and a limited EMS, they made no alternative arrangements to monitor the system. FE system operators received information suggesting emerging system security problems from several other sources throughout the afternoon including MISO (FE’s reliability co-ordinator), American Electric Power (AEP - a neighboring system operator), PJM (a neighboring reliability coordinator that includes Pennsylvania and other states), FE field personnel and customers. However, FE system operators dismissed this information as either not accurate or not relevant to managing their system. There was no subsequent verification of conditions with MISO and FE did not inform MISO or adjacent system operators when they became aware that system conditions had changed due to unscheduled equipment outages that might affect other control areas. Between 3.05 pm and 3.41 pm, three 345 kV transmission lines into the Cleveland-Akron area tripped due to contacts with overgrown trees. First to trip was the Harding-Chamberlain line at 3.05 pm. This was a significant event which combined with the earlier loss of the Eastlake 5 generation unit, pushed the FE system beyond its N-1 secure operational limits. However, FE system operators were not aware of the failure and took no action to return the system to a reliable and secure condition. MISO’s EMS did not monitor the Harding-Chamberlain line and was therefore unaware of its failure or that its failure would represent a significant contingency event under the circumstances. Loss of this transmission path caused the remaining three southern 345 kV transmission lines into Cleveland to pick up more load and caused more power to flow through the underlying 138 kV network. Two of these remaining 345 kV lines also tripped after coming into contact with trees, leaving only

20. The supervisory control and data acquisition (SCADA) system is a system of remote control and telemetry used to monitor and control the electric system. 21. An energy management system (EMS) is a computer control system used by electric utility dispatchers to monitor the real time performance of various elements of an electric system and to control generation and transmission facilities.

58

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

the Sammis-Star 345 kV transmission line operational. These trips increased loads on the remaining 345 kV and 138 kV network, degrading voltages on the FE system and making it increasingly insecure. From 3.40 pm FE, MISO and neighbouring utilities had begun to realise that the FE system was in jeopardy. PJM and AEP had already begun organising 350 MW of Transmission Loading Relief22 from 3:35 pm to reduce overloading on the Star-South Canton 345 kV line (a line spanning AEP and FE control areas and jointly managed). However, Task Force analysis indicates that this action or other redispatch would not have averted the overload. Task Force analysis suggests that by this time, the only way overloading in Ohio and the subsequent regional cascading could have been averted would have been for FE to initiate at least 1 500 MW of load-shedding in the Cleveland-Akron area. No load-shedding was undertaken by FE system operators. Each 345 kV line failure placed greater stress on FE’s 138 kV network serving the Cleveland-Akron area, overloading lines and reducing voltage. Twelve 138 kV lines tripped between 3.39 pm and 3.59 pm. A further four 138 kV lines and the Sammis-Star 345 kV line tripped between 4.00 pm and 4.08 pm. Declining voltage resulting from these line failures tripped several large industrial users with voltage-sensitive equipment, leading to the disconnection of around 600 MW of load in the Akron area. Failure of the Sammis-Star line at 4.06 pm triggered the uncontrollable regional cascade. At this point it was no longer possible for FE to manage the event with load shedding or any combination of redispatch. Although Sammis-Star was heavily overloaded at the time, the line was tripped by its protective relays which treated the high power flows as a short circuit. Failure of the SammisStar line shut down the 345 kV path into northern Ohio from eastern Ohio. This failure combined with high demand in northern Ohio instantly created substantial, unstable flows as power sought new paths into northern Ohio, overloading lines in adjacent areas. The cascade spread rapidly as lines and generating units were automatically tripped by protective relays to avoid physical damage. Figure 12 summarises the regional cascade phase of the event. Shortly before the collapse, large (but normal) electricity flows were moving across FE’s system from generators in the south (Tennessee and Kentucky) and

22. Transmission Loading Relief (TLR) is an administrative procedure used to manage system security violations on key elements of the bulk transmission system, such as flows that might overload transmission lines or threaten stability limits. Reliability coordinators can use TLRs to reduce such power flows. Requests are normally activated within 30 to 60 minutes of being received. TLRs are typically used to manage relatively small adjustments of between 25 and 50 MWs.

59

LEARNING FROM THE BLACKOUTS

Figure 12

Overview of the Regional Cascade Sequence

1.

1. 4.05:57 pm

6. 4.10:40 pm 2. 4.05:58 pm

7. 4.10:41 pm 3. 4.09:25 pm

8. 4.10:44 pm 4. 4.10:37 pm

9. 4.10:45 pm 5. 4.10:39 pm

10. 4.13:00 pm Source: US-Canada Power Outage Task Force (2004a)

60

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

west (Illinois and Missouri) to load centers in northern Ohio, eastern Michigan and Ontario (Figure 12, Panel 1). Failure of the Sammis-Star line induced unplanned shifts of power across the region, hastened by Zone 3 impedance relays23 (Figure 12, Panel 2). A series of lines in northern Ohio tripped as a result of overloads, causing a series of shifts in power flows and loadings, but the system stabilised after each (Figure 12, Panel 3). After 4.10:36 pm, the power surges resulting from the FE system failures caused lines in neighbouring areas to see overloads that caused impedance relays to operate. The result was a wave of line trips through western Ohio that separated the AEP control area from the FE control area. The line trips progressed northward into Michigan separating western and eastern Michigan, causing a power flow reversal within Michigan toward Cleveland. Many of these line trips were from Zone 3 impedance relay actions that accelerated the speed of the line trips and reduced the potential time in which grid operators might have identified and responded to the growing problem (Figure 12, Panel 4). With paths cut from the west, a massive power surge flowed from PJM into New York and Ontario in a counter-clockwise flow around Lake Erie to serve the load still connected in eastern Michigan and northern Ohio. Relays on transmission lines between PJM and New York saw this massive power surge as faults and tripped those lines. Ontario’s east-west tie line also became overloaded and tripped. The entire northeastern United States and eastern Ontario then became a large electrical island separated from the rest of the Eastern Interconnection. This large area, which had been importing power prior to the cascade, quickly became unstable after 4:10.38 pm as there was insufficient generation online within the island to meet electricity demand. Systems to the south and west of the split, such as PJM and AEP remained intact and were mostly unaffected by the outage. Once the northeast split from the rest of the Eastern Interconnection, the cascade was isolated (Figure 12, Panels 5 to 9). After 4.10:46 pm, the large electrical island in the Northeast had less generation than load, and was unstable with large power surges and swings in frequency and voltage. As a result, many lines and generators across the disturbance area tripped, breaking the area into several electrical islands. 23. Impedance relays are designed to detect currents and voltages outside normal operating parameters, resulting from system faults, and to disconnect transmission lines before they suffer damage. A relay is installed at each end of a transmission line, with each relay monitoring up to 3 zones or lengths of transmission line. Zone 1 is set to monitor 80% of the line next to the relay and to trip with no time delay when a fault is detected. Zone 2 is set to monitor the entire line and slightly beyond, and to trip with a slight delay when a fault is detected, while Zone 3 is set to monitor well beyond the end of the line and act more like a remote relay or back-up circuit breaker. Zone 3 relays should not trip under typical emergency conditions. However, in practice impedance relays may trip in response to ‘apparent’ faults caused by large swings in voltage or current.

61

LEARNING FROM THE BLACKOUTS

Generation and load within these smaller islands was often unbalanced, leading to further tripping of lines and generating units until equilibrium was established in each island or they blacked out. Figure 13 shows the accelerating pace of line and generator trips during the regional cascade phase of the event, and the related loss of generating capacity. Figure 13

Line and Generation Trips, and Lost Load During the Cascade Phase Cumulative Number of Lines/Transformers and Generators

GW

350

60

300

50

250

40

200 30 150 20 100

Lines and Transformers Generating Units

16 :12

16 :11

16 :10

16 :0 9

16 :0 8

0 16 :07

0 16 :0 6

10

16 :0 5

50

Time GWs of Generation Lost

Source: US-Canada Power Outage Task Force (2004a)

At that stage, most of the Northeast was blacked out (the area shown in gray in Figure 12, Panel 10). Some isolated areas of generation and load remained on-line for several minutes, while some of those areas in which a close generation-load balance could be maintained remained operational.

■

Impact and Restoration Once the regional cascade was complete, large portions of the Midwest and Northeast United States and Ontario, Canada had been disconnected. At least 265 power plants and over 500 individual generating units had shut down.

62

Huron

Lake

Erie

Lake

UNITED STATES OF AMERICA

Source: NERC (2004c).

Michigan

Lake

Superior

Lake

Areas affected by the blackout Service maintained in some areas

Some local load interrupted

Lake Ontario

Area Affected by the US-Canadian Blackout

Figure 14

CANADA

Atlantic Ocean

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

63

LEARNING FROM THE BLACKOUTS

Overall, 61,800 MW of load was lost in the states of Ohio, Michigan, Pennsylvania, New York, New Jersey, Connecticut, Vermont, Massachusetts and in the Canadian province of Ontario. Around 50 million people were disconnected initially. Figure 14 shows the area affected by the blackout. In the United States, the economic cost of the disruption has been estimated at between USD 4 billion and USD 10 billion24. In Canada, gross domestic product fell by around 0.7% in August, with 18.9 million working hours lost and manufacturing shipments down by CAD 2.3 billion25. Restoration Process Most services were restored in the United States within two days, with all services fully restored within four days. A regional summary of service restoration in the United States is presented in Table 3. Restoration proceeded in accordance with established restoration and blackstart procedures. Initial capacity shortages were managed through rotating load shedding. Restoration was greatly assisted by the ability to energize transmission from neighboring systems and by some relatively large islands that survived the event. One such island of approximately 5,700 MW around the Niagara area formed the basis for restoration in both New York and Ontario. Another island consisting of most of New England in the United States and the Maritime Provinces in Canada stabilised after the event allowing relatively quick restoration of many local area services. The reliability regions most affected by the blackout - East Central Area Reliability Coordination Agreement (ECAR), the Northeast Power Coordinating Council (NPCC) and the Mid-Atlantic Area Council (MAAC) – have reviewed the restoration process in their reliability regions. Key recommendations from these investigations are summarized in the North American Electric Reliability Council’s recent status report on implementation of the 2003 Blackout recommendations26, and include: ●

Restoration criteria and guides need to be kept up to date.

●

Voice communications between system operators and reliability coordinators could be improved. Conference call protocols could be improved. For instance, “open” conference calls could be used to facilitate

24. ELCON (2004). 25. Statistics Canada (2003a) and (2003b). 26. NERC (2005b).

64

Southern Lower Michigan and small areas near Flint, Alma, Saginaw, and Lansing Michigan

Consumers Power (ECAR)

2 500

Southwestern Connecticut and a portion of Western Massachusetts and Vermont

New York State

ISO New England (NPCC)

New York ISO (NPCC)

65

11 202

Source: US Energy Information Administration, Electricity Power Monthly December 2003 [DOE/EIA-0226 (2003/12)].

5 boroughs of NYC and Westchester County

4 500

PJM Interconnection, LLC Northern New Jersey Erie, Pennsylvania area (MAAC)

Consolidated Edison Co of New York (NPCC)

Not Available

Niagara Mohawk (NPCC) New York- Buffalo to Albany; Ontario, Canada to Pennsylvania

22 934

7 000

First Energy Corporation Northeast, Ohio (ECAR)

1 007

11 000

Southeastern Michigan including Detroit

Detroit Edison (ECAR)

Capacity Loss (MW) 18 500

Area Affected

Midwest Independent Geographic areas for MISO reliability coordination System Operator (ECAR) footprint: Michigan and Ohio

Utility/Power Pool (NERC Region)

Service Restoration in the United States

Table 3

8/15/03, 9:03 pm

Approximately 8/15/03, 6:00 am

8/14/03, 11:48 pm

8/18/03, 12:03 am

8/16/03, 3:45 am (restoration ended) 8/17/03, 7:00 pm (incident ended)

8/16/03, 8:27 pm

8/16/03, 1:03 pm

8/16/03, 7:00 am

Approximately 8/17/03, 5:00 pm

29

14

8

80

76

52

45

39

74

Restoration Date and Time Period (Hours)

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

LEARNING FROM THE BLACKOUTS

communication between reliability coordinators, and, more locally, between transmission operators and balancing authorities. Backup communications procedures should be regularly tested to verify their availability and effectiveness. Consideration could be given to more effective means of managing incoming phone calls. ●

Access to real-time system information could be improved to help strengthen system operator situational awareness during restoration and to support more effective management of alarm systems.

●

Contingency analysis should be improved during restoration.

●

Training exercises should be established to improve system operator capacity to undertake efficient, coordinated restoration. Drills could focus on synchronizing procedures (with generators and in the event that load “islands” emerge), and stabilizing procedures (to ensure that generation and demand can be matched to preserve “island” operation in emergency situations).

●

Adequate facilities need to be made available to accommodate additional staff, and adequate fuel needs to be stored for emergency power supply.

●

Load management during restoration should be improved. Public appeals to reduce demand could be employed and more effective forms of rotating load shedding could be developed and implemented to support more efficient restoration. Exchanges between control areas could also be improved to support more timely restoration of services.

All NERC regions are in the process of reviewing their blackstart and system restoration plans and procedures, and making necessary revisions. In Ontario27, a provincial state of emergency was declared, followed by suspension of the wholesale electricity market at 4.20 pm on 14 August. With the market suspended, Ontario’s Independent Electricity Market Operator (IMO – now known as the Independent Electricity System Operator) began issuing manual dispatch instructions to coordinate the restoration of the system. IMO’s first priority was to assess the extent of the disturbance and its implications for implementing the Ontario Power System Restoration Plan. Five electrical islands had survived in Ontario and these provided the basis for rebuilding the power system. Five key restoration paths were identified with restoration undertaken on each path simultaneously and independently, by building on surviving electrical islands and gradually adding generation and load. Figure 15 provides an overview of the restoration process in Ontario. 27. IMO (2003) and (2004) for further information about the restoration process in Ontario.

66

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

Figure 15

Restoration Strategy in Ontario

CANADA Ontario

Quebec

4

4 3 3

Lake

2

Huron Lake Ontario 1A

1B

New York

Michigan

Lake Erie UNITED STATES OF AMERICA Primary restoration paths Revised restoration paths Secondary restoration paths

Source: IMO (2004) 67

LEARNING FROM THE BLACKOUTS

Ontario’s basic minimum power system was reestablished at 5.20 am on August 15, a little over 13 hours after the blackout. Restoration of the IMO controlled grid was completed at 6.26 pm on August 15 (with the exception of the OntarioMichigan interconnections), 26 hours and 16 minutes after the blackout. The Ontario-Michigan interconnection was re-established on August 17 at 7.46 pm. Although grid supply had been restored, some generators could not return to service for several days, aggravating an already tight supply situation in Ontario. From this point, priority was given to restoring generating capacity, while rotational load shedding was implemented to help manage the initial shortage28. Initial load restoration over this period is shown in Figure 16. Figure 16

MW

Initial Load Restoration in Ontario: 14-15 August 2003

25 000 23 000 21 000 19 000 17 000 15 000 13 000 11 000 9 000 7 000 5 000 3 000 Th u. Au Th g. 1 u. 4 Au :00 g Th .1 u. 4 Au :03 Th g. 1 u. 4 Au :06 Th g. 14 u. Au :09 Th g. 14 u. Au :12 Th g. 1 u. 4 Au :15 Th g. 1 u. 4 Au :18 g . Fri . A 14:2 ug 1 Fri . 15 .A : u 00 Fri g. 1 . A 5:0 ug 3 Fri . 15 .A : ug 06 . 1 Fri . A 5:09 ug Fri . 15 .A : u 12 Fri g. 1 . A 5:1 ug 5 Fri . 15 .A :1 ug 8 .1 5: 21

1 000

Source: IMO (2004).

On August 17, the government of Ontario issued an urgent appeal to all consumers to reduce their use of electricity by 50% for the duration of the

28. Rotational load shedding was used to restrict loads served by distributors to approximately 75% of normal levels until all loads were restored at 10.40 pm on 15 August. See IMO (2004) for further details.

68

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

provincial emergency. Figure 17 illustrates Ontario’s actual and forecast load for August 14 to 22 (when the emergency was lifted). The demand response curve shows a significant customer response. Figure 17

Load Restoration in Ontario: 14-22 August 2003 Demand (MW) 25 000

20 000

15 000

10 000

Au g. 1 Au 4 0 g. 3 1 1 : Au 4 0 00 g. 3 1 Au 15 0 3:0 g. 3 0 1 1 Au 5 0 :00 g. 3 1 Au 16 0 3:0 g. 3 0 1 1 Au 6 0 :00 g. 3 1 Au 17 0 3:0 g. 3 0 1 1 Au 7 0 :00 g. 3 1 Au 18 0 3:0 g. 3 0 1 1 Au 8 0 :00 g. 3 1 Au 19 0 3:0 g. 3 0 1 1 Au 9 0 :00 g. 3 1 Au 20 3:0 g. 03 0 2 1 Au 0 0 :00 g. 3 1 Au 21 0 3:0 g. 3 0 2 1 Au 1 0 :00 g. 3 1 Au 22 3:0 g. 03 0 22 1: 03 00 13 :0 0

5 000

Normal Actual

Estimated Demand Response Forecast

Source: IMO (2004).

The provincial emergency lasted for 9 days and was terminated by the Ontario government at 8.00 pm, 22 August. The Ontario Power Market resumed normal operations from midnight, 22 August.

■

Key Findings of Investigations and Inquiries Comprehensive investigations of the nature and causes of this event were undertaken by the US-Canada Power System Outage Task Force (the Task Force) and by the North American Electric Reliability Council (NERC)29. 29. See US-Canada Power System Outage Task Force (2004a) and NERC (2004c) for further details.

69

LEARNING FROM THE BLACKOUTS

The Final Report of the Task Force identified four groups of causes for the Ohio phase of the blackout: ●

Inadequate System Understanding. The Task Force concluded that FE and ECAR (FE’s reliability coordinator) failed to understand, assess and effectively manage the vulnerabilities of the FE system, particularly in relation to the Cleveland-Akron area. FE did not undertake rigorous longterm planning studies of its system, nor did it conduct appropriate multiple contingency analysis for its control area. FE’s voltage analysis and operational voltage criteria were inadequate and did not reflect actual voltage stability conditions or needs. ECAR did not independently review or correct FE’s inappropriate operational practises. Some NERC planning and operational standards were sufficiently ambiguous to enable FE to adopt inadequate operational practises.

●

Inadequate Situational Awareness. FE was unaware of the deteriorating condition of its system and failed to ensure the security of its transmission network following significant contingency events. FE did not have the capacity to run regular contingency analysis. It also lacked procedures to monitor the status of its key system monitoring equipment or to test the functionality of this equipment following repairs. Additional monitoring tools were not available to facilitate effective system operation once the primary alarm systems had failed. Communications between computer support staff and operators was ineffective.

●

Inadequate Tree Trimming. FE failed to undertake effective vegetation management of easements associated with its 345 kV network. Contact with trees was the principal cause of three 345 kV and one 138 kV transmission line failure.

●

Inadequate Diagnostic Support from Reliability Coordinators. The reliability coordinators responsible for overseeing secure regional operations failed to provide effective real-time diagnostic support. The absence of real-time information on the operational status of transmission flowgates precluded MISO from detecting N-1 security violations in FE’s system. The situation was exacerbated on the day by an absence of realtime data from the Stuart-Atlanta transmission line. A lack of real-time information prevented MISO from diagnosing FE’s system problems in a timely manner, degrading its capacity to provide effective diagnostic assistance or direction. MISO also lacked an effective means of identifying the location or significance of transmission line trips reported by its EMS. PJM and MISO also lacked procedures to coordinate a response to an N-1 violation observed in one another’s area.

70

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

The investigation also concluded that protection devices acted to accelerate the cascade sequence and to increase the magnitude of the disruption. Impedance relays on transmission lines and protective devices on generators are designed to detect high currents and low voltages resulting from system faults and to disconnect equipment when such faults are registered to protect them from being damaged. However, multiple contingency events can send transient oscillations through interconnected alternating current networks, which can be misinterpreted as faults by protection devices, leading to erroneous tripping of equipment. During the event, protective devices, especially Zone 3 impedance relays on transmission lines, activated so quickly that system operators did not have an opportunity to initiate any emergency actions. The Task Force report notes that most transmission lines affected at the time could have carried the likely resulting overloads for at least 15 minutes and possibly up to an hour without sustaining significant damage, which might have provided sufficient time for effective emergency intervention to limit the impact of the event. In addition, the North American Electric Reliability Council’s initial investigations identified several violations of its operating policies and planning standards, which in its view directly contributed to causing the outage, including those listed below. ●

Following the outage of the Chamberlin-Harding 345 kV line, FE operating personnel did not take the necessary action to return the system to a safe operating state as required by NERC Policy 2, Section A, Standard 1.

●

FE operations personnel did not adequately communicate FE’s emergency operating conditions to neighboring systems as required by NERC Policy 5, Section A.

●

FE’s state estimation and contingency analysis tools were not used to assess system conditions, violating NERC Operating Policy 5, Section C, Requirement 3, and Policy 4, Section A, Requirement 5.

●

MISO did not notify other reliability coordinators of potential system problems as required by NERC Policy 9, Section C, Requirement 2.

●

MISO was using non-real-time data to support real-time operations, in violation of NERC Policy 9, Appendix D, Section A, Criteria 5.2.

●

PJM and MISO, as reliability coordinators, lacked procedures or guidelines between their respective organisations regarding the coordination of actions to address an operating security limit violation observed by one of them in the other’s area due to a contingency near their common boundary, as required by Policy 9, Appendix C. 71

LEARNING FROM THE BLACKOUTS

●

FE’s operational monitoring equipment was not adequate to alert FE’s operators about important deviations in operating conditions and the need for corrective action as required by NERC Policy 4, Section A, Requirement 530.

Investigations also found that restoration of services in the United States and Canada had generally been undertaken in a timely and effective manner given the magnitude of the disturbance. The Ontario IMO concluded that effective collaboration throughout the supply chain had helped to eliminate restoration errors and minimise the outage period. However the IMO’s report also noted opportunities to improve future restoration processes particularly in relation to the management of emergency load shedding and operational communications. Investigations on the US side are continuing31.

■

Key Recommendations Recommendations proposed by the Task Force and by NERC are presented in Annex 1, along with a summary of the implementation status. These investigations proposed over 60 recommendations. Key issues raised in the recommendations included the need to: ●

improve governance, regulation and enforcement, in particular by clarifying the regulatory standards, clarifying the responsibilities and accountabilities of parties subject to those standards, and by strengthening accountability through mandatory application and enforcement of those standards;

●

improve system operator training, particularly in relation to managing emergency events and avoiding future emergency events through more effective management of near emergency conditions;

●

ensure system operation is supported by effective diagnostic and management tools and information, facilitate real-time multiple contingency analysis, system monitoring and operational management;

●

clarify and strengthen standards and operating practises relating to vegetation management; and

●

improve coordination, communication and data exchange between system operators and reliability coordinators at a local and regional level.

30. See US-Canada Power System Outage Task Force (2004a) and NERC (2004a) for details. 31. See IMO (2004) for further details.

72

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

CASE STUDY 2: Switzerland and Italy ................................. Italy’s worst supply disruption in over 50 years struck on Sunday, 28 September 2003. Parts of southern Switzerland were also affected. Cascading failure of the interconnectors serving Italy led to the separation of the Italian grid from the Union for the Co-ordination of Transmission of Electricity (UCTE) system. Separation created highly unstable operating conditions throughout the Italian system, causing widespread generator and line trips. Emergency responses were unable to arrest the situation and the Italian system collapsed around 3.30 am. The following event summary is largely drawn from the final report of the UCTE Italian Blackout Investigation Committee32. The event started at around 3.01 am on 28 September 2003 with the failure of the Swiss 380 kV line Mettlen-Lavorgo (also known as the “Lukmanier” line). The line was relatively heavily loaded just prior to its failure, with loading levels at around 86% of maximum rated capacity. High loading levels result in overheating of conductors, which causes transmission lines to sag, increasing the potential for a short circuit caused by an electric arc between a line and a grounded object, such as a tree (ie. a flashover), or possibly by direct contact with a grounded object. The Mettlen-Lavorgo line failed as a result of a flashover with a tree. ETRANS (the Swiss high voltage transmission system co-ordinator) made several attempts to automatically reclose the line without success. A manual attempt at 3.08 am also failed. Reclosure after a line trip of this kind is standard operating practice, allowing lines to be reconnected where they remain physically intact. However, on this occasion reclosure attempts failed because of high power flows into Italy at the time, which had created a high phase angle33 difference in voltages at either end of the line. This phase angle difference exceeded the stability limits set for the line and resulted in its protection devices cutting in to prevent reclosure. At 3.11 am, a phone conversation took place between the ETRANS coordination centre in Laufenburg and the Gestore della Rete di Trasmissione Nazional (GRTN – the Italian transmission system operator) control centre in Rome. ETRANS asked GRTN to activate countermeasures within the Italian system, in order to 32. UCTE (2004a). 33. Phase angles measure the difference in voltage caused by the voltage at the load end of the line being out of phase with the voltage at the generator end. This difference determines the rate at which power flows along a line from generation to load in an AC environment. The greater the phase angle difference the greater the power flow. However, power flow only increases until the phase difference between two voltages reaches 90°, after which the power transfer becomes unstable possibly leading to voltage collapse and damage to infrastructure. This is a theoretical maximum. Power lines are never run near this limit. In practise, lower limits of between 30° and 45° are typically set and these form a stability limit for power flows on a transmission line.

73

LEARNING FROM THE BLACKOUTS

help relieve the overloads in Switzerland and return the system to a secure state. In essence, the request was to reduce Italian imports by 300 MW, because Italy was at that time importing around 300 MW more than the scheduled power transfers, which amounted to 6 400 MW on the northern border. The reduction of Italian imports by about 300 MW took effect at 3.21 am, 10 minutes after the phone call and returned Italy close to the agreed schedule. However, the import reduction together with internal countermeasures taken within the Swiss system was insufficient to relieve the overload. After the loss of the Mettlen-Lavorgo line, loads on neighboring lines increased. In particular, the Swiss 380 kV Sils-Soazza line (also called the “San Bernardino” line) was operating at 110% of its normal maximum rated capacity following the loss of the Mettlen-Lavorgo line. According to UCTE operating standards, an overload of this magnitude can be maintained in emergency circumstances, but only for a short period. Under the operating standards, the system operator had no more than 15 minutes to eliminate the overload. At 3.25:21 am, around 24 minutes after the loss of the MettlenLavorgo line, the Sils-Soazza line also tripped as a result of a flashover with a tree. The UCTE Investigating Committee concluded that this flashover was probably caused by the sag in the line, due to overheating of the conductors. A significant part of the additional power flow resulting from failure of the Sils-Soazza line was redistributed to an internal 220 kV line in Switzerland (the Mettlen-Airolo line), which became highly overloaded and was tripped by its protection devices almost immediately. Other Swiss 220 kV lines in the area subsequently tripped due to overloads, separating the southern part of Switzerland from the Swiss network. Loss of the Mettlen-Lavorgo and Sils-Soazza lines also created substantial overloads on the other transmission lines in the area. The remaining lines from Riddes and Robbia to Italy then tripped, leading to an overload on the interconnectors with France. This overload caused a significant and rapid decrease in voltage at the French border. Low voltages and high currents caused protection devices to trip the French 380 kV Albertville-La Coche-Praz line. At this point the Italian system lost synchronisation with the UCTE network and all remaining interconnectors were disconnected almost simultaneously by their automatic protection devices. The Italian system was isolated from the UCTE network about 12 seconds after the loss of the Sils-Soazza line, at around 3.25:34 am. A weak 220/132 kV interconnection remained with Slovenia till it tripped at around 3.26:30 am, after which disconnection was complete. Disconnection largely occurred along the Italian borders with 74

Piemonte

ITALY

2

3

4

Emilia-Romagna

11. 380 Magliano-Piossasco 380 Magliano-Vado Ligure 380 Casanova-Magliano Liguria Mediterranean 380 Vignole-Vado Ligure Sea

9. 380 Villarodin-Venaus 380 Villarodin-Praz

Lombardia

8. 380 Lavorgo-Musignano

5

6

Veneto

TrentinoAlto Adige

13. 380 Soazza-Bulciago 220 Magadino-Soazza

2. 380 Sils-Soazza

SWITZERLAND

1. 380 Mettlen-Lavorgo

7

8

9 10 11 12

Adriatic Sea

13

7. 380 Divaca-Redipuglia

Friuli-Venezia Guilia

CROATIA

SLOVENIA

5. 220 Lienz-Soverzene

AUSTRIA

75

Source: UCTE (2004a)

03.01:00 03.01:30 03.02:00 03.02:30 03.03:00 03.03:30 03.04:00 03.04:30 03.05:00 03.05:30 03.06:00 03.06:30 03.07:00 03.07:30 03.08:00 03.08:30 03.09:00 03.09:30 03.10:00 03.10:30 03.11:00 03.11:30 03.12:00 03.12:30 03.13:00 03.13:30 03.14:00 03.14:30 03.15:00 03.15:30 03.16:00 03.16:30 03.17:00 03.17:30 03.18:00 03.18:30 03.19:00 03.19:30 03.20:00 03.20:30 03.21:00 03.21:30 03.22:00 03.22:30 03.23:00 03.23:30 03.24:00 03.24:30 03.25:00 03.25:30 03.26:00 03.26:30 03.27:00 03.27:30 03.28:00 03.28:30 03.29:00 03.29:30 03.30:00 03.30:30 03.31:00 03.31:30 03.32:00 03.32:30 03.33:00 03.33:30 03.34:00 03.34:30 03.35:00 03.35:30 03.36:00 03.36:30 03.37:00 03.37:30 03.38:00 03.38:30 03.39:00 03.39:30

1

Valle d’Aosta

6. 380 Albertville-Rondissone 1+2

FRANCE

12. 220 Bavona-Robiei 4. 220 Avise-Riddes-Valpelline

10. 220 Robiei-Innertkirchen

3. 220 Mettlen-Airolo

Overview of the Italian Interconnector Separation Sequence

Figure 18 2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

76

Source: UCTE (2004a)

FRANCE

SWITZERLAND

Line of separation from the European grid

ITALY

SLOV.

CROATIA

AUSTRIA

Final Line of Separation of the Italian Transmission System from the UCTE Transmission Network

Figure 19

LEARNING FROM THE BLACKOUTS

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

France, Switzerland and Austria. Pockets within France and Switzerland were also disconnected. On the Slovenian border, disconnection occurred within Italy. An overview of the interconnector separation sequence and the final line of separation of the Italian transmission system from the continental European transmission network are presented in Figure 18 and Figure 19. During the 12 seconds of very high overloads, instability phenomena started in the affected area of the system. This led to very low voltage levels in northern Italy that tripped several Italian generators. Separation from the UCTE network created a large generation deficit of nearly 6 650 MW, which also caused a fast frequency drop throughout the Italian system. GRTN’s under-frequency load-shedding plan automatically activated. The plan was designed to cover between 50% and 60% of total Italian demand, and was especially calibrated for response during peak periods. The plan involved stepped shedding of pump storage plants, large industrials and other high voltage and medium voltage users. However, implementation was ‘hard wired’, which meant that it was not possible to quickly modify implementation to account for the specific characteristics of load at the time of the event. Once the pump storage units had been shed, the remaining load available to be shed at around 3.30 am accounted for less than 11,000 MW. Much of the remaining load was residential and largely excluded from the defense plan due to binding contractual and legal conditions. Although the plan operated as programmed, its effectiveness was probably reduced by the relative lack of available load to shed during the early hours on a Sunday morning. Primary frequency control combined with automatic shedding of the pumped storage power plants and some industrial demand helped to slow the rate of decline. However, it was insufficient to prevent the ultimate collapse of the islanded Italian system. Figure 20 provides a summary of the key events during the transitory period until system collapse. Additional generating units tripped for various reasons that typically reflected under-frequency relay operation or the activation of gas turbine protection devices. Despite additional load shedding, the frequency continued to decline and the system collapsed 2 minutes and 30 seconds after the separation when the frequency within the Italian system reached the threshold of 47.5 Hz at around 3.30 am. Following the separation, the significant fall in load on the UCTE system resulting from the loss of demand in Italy led to a sharp increase in frequency 42. IEA 2005c. 43. EFET, 2004.

77

LEARNING FROM THE BLACKOUTS

Figure 20

Key Events during the Transitory Period Hz

MW

50.500

1 000

50.000

0

49.500

-1 000

49.000

-2 000

48.500

-3 000

48.000 -4 000

47.500

-5 000

47.000 46.500

-6 000

46.000

-7 000 -8 000

45.500 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 03: 25: 25: 25: 25: 25: 25: 25: 25: 26: 26: 26: 26: 26: 26: 26: 26: 26: 26: 26: 26: 27: 27: 27: 27: 27: 27: 27: 27: 27: 27: 27: 27: 28: 20 25 30 35 40 45 50 55 00 05 10 15 20 25 30 35 40 45 50 55 00 05 10 15 20 25 30 35 40 45 50 55 00

Total Imbalance Hz 50.500

Soazza-Sils Outage

Frequency Separation from European grid

50.000 49.500

48.000 47.500

Critical threshold

Loss of all power plants operating at distribution grid ~3 400 MW Loss of generation units (> 50 MW before the critical threshold ~4 100 MW)

49.000 48.500

Time

Automatic pumping shedding ~3 200 MW

47,50 Definitive blackout

Load shedding (EAC) equal to ~600 MW each (–0,1 Hz) frequency loss: total ~7 700 MW

47.000 46.500 46.000

03 :24 03 :00 :24 03 :08 :24 03 :16 :24 03 :24 :24 03 :32 :24 03 :40 :24 03 :48 :24 03 :56 :25 03 :04 :25 03 :12 :25 03 :20 :25 03 :28 :25 03 :36 :25 03 :44 :25 03 :52 :26 03 :00 :26 03 :08 :26 03 :16 :26 03 :24 :26 03 :32 :26 03 :40 :26 03 :48 :26 03 :56 :27 03 :04 :27 03 :12 :27 03 :20 :27 03 :28 :27 03 :36 :27 03 :44 :27 03 :52 :28 :00

45.500

Time 3:25 Source: UCTE (2004a).

78

~2,5 minutes

3:28

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

across the UCTE network, creating a potential danger for system security across continental Europe. Some generating units were tripped by overfrequency and under-voltage relays. Although loading on lines from France to Germany to Belgium increased significantly, system operators successfully implemented emergency responses that quarantined the effects of the outage and prevented a cascading failure across the UCTE network.

■

Impact and Restoration Immediately after the event, all of Italy (with the exception of Sardinia), a portion of southern Switzerland up to the outskirts of Geneva, and a small area within France near the Italian border was disconnected. Over 340 power plants operating at the time in Italy had been shut down, representing a loss of over 20 500 MW of domestic generation. Around 6 600 MW of electricity imports to Italy had also been lost. Over 27 000 MW of load was lost at the time of the failure. Around 55 million people were initially disconnected. The economic cost of the disruption has been estimated at around USD 139 million34. 31 thermal units initiated the sequence to switch to in-house operating mode prior to system collapse. However, only 8 of these plants completed the sequence, allowing them to remain in operation after the collapse and to provide immediate support for restoration activities. Two small electrical islands continued to operate following the general collapse: one south of Rome and the other in the south western region of Calabria. These electrical islands helped support restoration of services in the southern and central regions. Restoration Process In Switzerland, services were largely restored within 1.5 hours. The San Bernardino region was the exception, where lengthy interruptions continued until the afternoon. In Italy, restoration proceeded according to GRTN’s established plan. The plan was based on implementing several restoration paths in parallel to restore auxiliary services to shut down power plants and to reconnect thermal plants that either succeeded in maintaining operation within an electrical island or by switching to in-house operating mode. Implementation proceeded once operators had a clear understanding of the nature of the outage and was managed with caution to avoid further service failures. Load was progressively added to the islands that emerged from each restoration path. Strategic level implementation was planned, coordinated and supervised centrally. However, 34. OXERA (2005).

79

LEARNING FROM THE BLACKOUTS

many initial activities were undertaken locally and autonomously, with central coordination taking over as the islands developed and were ultimately reintegrated. A key objective was to restore services in the shortest possible time, especially to metropolitan areas. In practice, timely and successful implementation of the restoration plan was influenced by the limited availability of base-load thermal units which remained operational after the collapse; the availability of hydro and gas-fired generators to provide black-start services, particularly voltage and frequency control as load was restored; and the reliability of telecommunications systems to support coordinated restoration activities. Although some difficulties were encountered, the UCTE Investigation Committee considers that given the severity of the outage the restoration process was timely and successfully performed. However, the duration of the emergency may have been reduced if more generating units had managed to switch successfully to in-house operating mode or if more of the black-start capable plants had been operational. The main 380 kV network was re-energised 13.5 hours after the initial collapse. Fifty percent of load was reconnected after 6.5 hours, 70% was reconnected after 10 hours and 99% was reconnected after 15 hours. Services were completely restored to all consumers 18 hours and 12 minutes after the outage. The UCTE Investigation Committee estimates that around 177 GWh of energy was not supplied as a result of the outage. Twenty-six restoration paths were used, with service restoration largely progressing from north to south. Figure 21 provides a diagrammatic overview of the restoration process. Approximately three hours after the outage, services to the northwestern region of Italy had been almost completely restored and reconnected with the French network. Hydroelectric generation was available to progressively resupply load in this area, while around 750 MW of thermal capacity had completed the initial start-up sequence in the Lombardy region. The Lombardy region was also being supplied through two energised transmission backbones that had been synchronised with the UCTE system through interconnections with Switzerland. Services had also been restored to parts of northern Florence through a connection with Lombardy, and to eastern and northern Venice through interconnections with Slovenia. By 8.00 am, the northern network had been re-energised and was operating synchronously with the UCTE system. No significant progress had been made on restoring services in the central region. Restoration progress was delayed 80

Source: UCTE (2004a)

Existing line (380 kV)

Panel 1: Restoration by 8.00 am, 28 September 2003 Planned line (380 kV)

Existing power plant

Panel 2: Restoration by 12.00 pm, 28 September 2003

Planned power plant

Panel 3: Restoration by 5.00 pm, 28 September 2003

Restoration Progress Following the Italian Blackout

Figure 21

0

Km

200

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

81

LEARNING FROM THE BLACKOUTS

in the southern region by telecommunications failures and remote control problems with some sub-stations on the preferred restoration paths. Panel 1 of Figure 21 shows progress by 8.00 am. Between 8.00 am and 12.00 pm restoration continued more slowly than expected due to technical difficulties with switching operations at substations, and telecommunications outages which continued to complicate restoration. In particular, loss of the telecommunications network around Naples and Palermo led to a loss of all SCADA data for these areas between 8.50 am and 12.03 pm and between 8.50 am and 1.17 pm respectively. This complicated coordination of activities and re-supply in the central and southern regions. Progressive depletion of the hydro reservoirs, which had not been fully replenished as a result of the outage the night before, emerged as a significant issue with the potential to disrupt the restoration effort. GRTN asked distributors to interrupt industrial loads and implement the first level of rotating load shedding between 11.00 am and 6.00 pm in the northern and central-northern regions. However, the response was not effective. By 12.00 pm load in the northern region was mostly restored. In the southern region the two electrical islands south of Rome and southwest of Calabria were reintegrated and expanded. Panel 2 of Figure 21 shows the progress of restoration by 12.00 pm. From noon, the focus of the restoration shifted toward re-supplying central and southern Italy and Sicily as soon as possible. The relative weakness of the transmission backbone from northern to central-southern Italy combined with the lack of available generation in the central-southern regions had induced high power flows from north to south, creating a phase angle risk. Restoration of the 500 MW interconnection with Greece helped strengthen system security and allowed some additional domestic capacity to be released from maintaining system security to serve load. Services were restored to metropolitan Rome at 1.17 pm. Panel 3 of Figure 21 shows that services had been largely restored to Italy by 5.00 pm. The interconnection with the UCTE network had also been fully restored. Progress during the initial phases of restoration was slowed by the relatively small number of thermal plants that had managed to switch to in-house operating mode, and by failures and difficulties in starting the black-start power plants. A lack of information on the primary causes of the outage also slowed implementation, while problems with voice and data communications hindered restoration. Implementation and coordination difficulties were typically overcome by employing alternative restoration routes and by issuing instructions over the phone. UCTE reserves were also used to facilitate black82

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

start activities and progressive restoration of load once sufficient interconnection had been restored. Restoration of services in Sicily proved more difficult than expected. Several attempts to restore services independently without interconnection failed. Ultimately, restoration was undertaken using the interconnection with Calabria. However, power exchange on the interconnector was effectively limited to 200 MW. Power exchange limitations combined with the need to direct scarce generation capacity to meet the evening peak slowed the pace of restoration. Emergency conditions were terminated with the restoration of power supplies to Sicily at 9.40 pm.

■

Key Findings of Investigations and Inquiries Several official investigations of the event have been undertaken, most notably by the Swiss Federal Office of Energy, through the joint report by the Italian and French regulators and by the UCTE. Table 4 summarises UCTE’s analysis of the key underlying causes of the event, the relative impact of each cause on the event and the actions initiated in response to the event. The influence of the underlying causes on system security leading to the outage is summarised in Figure 22. The UCTE Investigating Committee Final Report notes that the integrated system was operating in an N-1 secure state, consistent with UCTE operating procedures, prior to the failure of the Mettlen-Lavorgo line. The inability of the system operator to reclose this line after its initial failure due to the phase angle exceeding stability limits was a direct and decisive underlying cause of the event. 10 minutes were lost in unsuccessful attempts to reclose the line, time which proved crucial given that the resulting overload on the Sils-Soazza line could only be sustained for around 15 minutes. Subsequent responses by system operators in Switzerland and Italy were considered slow, given the nature of the emergency, and the measures adopted were insufficient to return the system to an N-1 secure state within the 15 minute margin to reduce the overload on the Sils-Soazza line. The UCTE Investigation Committee Final Report also noted the differing accounts of a conversation between the Italian and Swiss system operators, and suggested that ineffective communication and information exchange contributed to an ineffective response. Therefore, although the countermeasures were available to return the system to a secure state, operational, technical and organisational factors prevented a timely and appropriate response. 83

LEARNING FROM THE BLACKOUTS

Table 4

Primary Causes of the Swiss-Italian Blackout Origin of Primary Cause

Action

Decisive Unsuccessful re-closing of the Lukmanier line because the phase angle difference was too high

Large phase angle due to power flows and network topology

Study settings of concerned protection devices. Reassess possible consequences for net transfer capacity to Italy. Coordination of emergency procedures.

Decisive

Human factor

Operator training for emergency procedures. Reassess acceptable overload margins. Study realtime monitoring of transmission line capacities

General tendency towards grid use close to its limits

Further studies necessary on how to integrate stability issues in UCTE security & reliability policy.

Operational practices

Perform technical audit if necessary, improve tree cutting practices

Primary Cause

Lacking a sense of urgency regarding the San Bernardino line overload and call for inadequate countermeasures in Italy

Impact on Event

Angle instability Not an underlying and voltage collapse cause of the events in Italy but was the reason that island operation of Italy after its disconnection failed. Right-of-way maintenance practices

Possible

Source: UCTE (2004a).

UCTE analysis also indicates that angle and voltage instability occurred in the moments just prior to complete disconnection of the Italian system, resulting in very low voltages in the northern portion of the Italian system at the time of disconnection from the UCTE network. Successful migration to island 84

Collapsed network

Source: UCTE (2004a).

and

Italy disconnected by almost simultaneous line tripping

Endangered network state

2 Lacking a sense of urgency regarding the San Bernardino line overload and call for inadequate countermeasures in Italy

1 Unsuccessful reclosing of the Lukmanier line because the phase angle difference was too high

Network in (n–2) state with excessive overload of remaining lines

Disturbed network state

2nd tree flashover line tripping

and

1st tree flashover line tripping

and

Network in (n–1) state with short-term (15’) allowable overload

Safe network state

Pre-incident network in (n–1) secure state

Network State Overview & Roots Causes

Separation of Italy from the UCTE main grid

Island operation fails due to unit tripping

Event or automatic action

Primary cause

Angle instability and Voltage collapse in Italy

3

Tripping of many power units

and

Impact of Primary Causes on System Security during the Swiss-Italian Blackout

Figure 22 2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

85

LEARNING FROM THE BLACKOUTS

operation was very dependent on keeping generating capacity in service, and in the Italian case a necessary condition for achieving this was that voltage and frequency conditions at the moment of disconnection were close to normal. This condition was not met and 21 of 50 large thermal generators were lost during the 2.5 minutes between disconnection and ultimate system failure. According to the UCTE Investigation Committee, the observed instability at the time of disconnection was a primary cause for the failure of the Italian system to operate in an islanded mode. The UCTE Final Report also suggests that line flashovers that led to key transmission line failures triggering the event may have been caused by insufficient maintenance of easements under transmission lines. This issue was beyond the scope of the UCTE investigation. Hence, a definitive view was not presented. The Swiss Inspectorate for Heavy Current Installations conducted an investigation of this issue and concluded that the network owners involved had undertaken vegetation management in accordance with relevant regulations and standard practise35.

■

UCTE Investigating Committee Recommendations The UCTE Investigating Committee Final Report proposes several recommendations for action at a regional UCTE and national transmission system operator (TSO) level. It also raises several issues in relation to the interface with regulatory arrangements, principally regarding factors affecting system adequacy, especially investments in the transmission system, and the need to strengthen the powers and independence of TSOs to ensure system security. Recommendations relating to the UCTE-regional level include: ●

Operational Procedures. For interconnections between UCTE control blocks, confirm, set up or update where necessary the emergency procedures between the involved TSOs. They should be made mandatory and integrated into joint operator training programs. Their performance should be evaluated at regular intervals.

●

Application of the N-1 Criterion. UCTE’s rules in relation to security assessment should seek to: • harmonise criteria for compliance with the N-1 principle; • determine criteria for the period allowed to return the system to an N-1 secure state after a contingency;

35. Swiss Federal Inspectorate for Heavy Current Installations (2003).

86

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

• include issues such as phase angle and voltage stability in the standard short-term contingency analysis; and • define clear guidelines for sharing of tasks to be performed, taking into account the boundaries of each control area. ●

Day Ahead Congestion Forecasts (DACF). Intensify the ongoing work on DACF in UCTE. In particular increase the frequency of DACF calculations, increase the number of areas involved and provide quality indicators for the data and computed results in order to assess the performance of the forecasts.

●

Real-time Data Exchange. Extend the existing real-time data exchange among neighbouring TSOs. Data should be consistent so that state estimators can produce more accurate results for a wider area.

●

Technical Specifications for Generating Equipment. Determine on a UCTE level a set of minimum requirements for generation equipment, defense plans and restoration plans, as a basis for harmonization to be implemented throughout the respective national grid codes and regulations.

●

Emergency Control and Co-ordination of Load Frequency Control. Further work at the UCTE level is needed to agree on the implementation of appropriate load-frequency control strategies should an accidental split of the synchronous area occur.

●

Improving Wide Area Measurement System (WAMS) Management. As a support tool for dynamic analysis and monitoring of the UCTE system, accelerate the ongoing WAMS installation program.

Recommendations relating to TSO’s at the national level include: ●

Enforcement of Technical Specifications for Generating Equipment. National Grid Codes (or equivalent regulation) should enforce a set of minimum requirements, to be harmonised at the UCTE level, with respect to the tolerance of generation units to frequency and voltage disturbances.

●

Defense and Restoration Plans. National regulations should, insofar as they are not yet implemented, provide for: binding defense plans with frequency coordination between load shedding, if any, and generator trip settings; and binding restoration plans with units sufficiently capable of switching to in-house operation and black-start capability. Due consideration should be given to joint simulation, training and evaluation of these plans with all involved parties. 87

LEARNING FROM THE BLACKOUTS

■

●

Vegetation Management. Tree trimming practices should be evaluated and the operational results should be audited with respect to the potential line sag under maximum rated overload conditions.

●

Distribution Level Measures. The blocking of On Load Tap Changers (OLTC) of transformers in case of severe voltage drop should be accepted practice.

Recommendations from the Joint Inquiry by the Italian and French Regulators Commission de Regulation de l’Energie (CRE – the French Regulator) and Autorita per l’Energia Elettrica e il Gas (AEEG – the Italian Regulator) released a joint report on the event in April 200436. The Report concluded that Swiss operational contingency planning and resources for 28 September 2003 were inadequate and that Swiss system operators lacked sufficient countermeasures to manage a foreseeable event. The Report stated that this lack of preparation contravened UCTE operating procedures and put the whole UCTE system at risk. It also found that inappropriate operational responses exacerbated the emergency and were responsible for the failure of the Sils-Soazza line. Key recommendations included the following. ●

The potential for diverging interpretation and application of UCTE operating rules should be eliminated by providing further detail to remove ambiguity.

●

Compliance with UCTE rules should be made legally binding.

●

The application of UCTE rules should be monitored and enforced by independent bodies formally responsible for verifying TSO compliance. CRE and AEEG suggest that the industry regulators be the responsible body in this context.

●

TSO coordination of operational planning and real-time operation of integrated networks should be reinforced.

●

A legal and regulatory framework coherent with EU legislation should be implemented and enforced in all European countries operating within the integrated UCTE system to ensure secure network operation and supply in Continental Europe. Switzerland in particular should adopt a compatible framework as a matter of priority.

36. CRE and AEEG (2004).

88

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

■

Recommendations from the Inquiry of the Swiss Federal Office of Energy The Swiss Federal Office of Energy released its report on the event in November 200337. The report concludes that the underlying causes of this incident lay in “the unresolved conflict between the trading interests of the involved countries and companies, and the technical requirements of the existing trans-national electricity system”.38 Deviations between planned transit volumes and actual volumes represent a threat to system security and should be minimised to the greatest extent possible. Current operational requirements are lagging behind economic realities and need to be addressed. In particular, real-time corrective measures that needed to be undertaken interactively by the system operators in Switzerland, France and Italy could have been implemented more rapidly and coordinated more efficiently. Priority needs to be given to the development and implementation of binding regulation of international electricity trading. Key recommendations included the following. ●

Provisions contained in EU Directive 1228/2003 governing network access for cross-border trade need to be implemented in a manner that does not compromise system security. Procedures need to be applied fairly and transparently.

●

Swiss authorities should have the right to jointly determine capacity allocations for export to Italy with Italian and French authorities.

●

Swiss transmission owners should voluntarily align themselves with the organisational requirements contained in the EU Directives. In particular, they should create an independent TSO as soon as possible.

●

Roles and responsibilities of market participants, regulators and market institutions need to be clarified, with binding requirements introduced for: • network-related security of supply; • allocation of export volumes to Italy; • minimum national reserve capacity requirements; and • financial compensation for eliminating network congestion.

●

Switzerland needs to implement comprehensive electricity market legislation consistent with the EU Directives and to establish a strong electricity sector regulatory authority to ensure: – non-discriminatory network access and allocation of transmission capacity;

37. SFOE (2003). 38. SFOE (2003), p21.

89

LEARNING FROM THE BLACKOUTS

• adequate network investment; and • adequate system security. ●

Technical regulatory requirements in relation to high voltage transmission line loadings should be reviewed.

●

Governmental arrangements for handling power failures should be reviewed. Communication arrangements between network operators and government organisations responsible for energy crisis management should also be reviewed.

CASE STUDY 3: Sweden and Denmark ................................ The Nordic transmission system experienced its worst power failure in 20 years at around 12.37 pm on Tuesday, 23 September 2003. A combination of mechanical faults in southern Sweden occurred in quick succession at around 12.30 pm and 12.35 pm, creating system conditions that were beyond the capacity of normal operational reserves. As a consequence, supplies to southern Sweden and eastern Denmark, including the Danish capital, Copenhagen, were disrupted. The following event summary is largely drawn from the reports prepared by the transmission system operators responsible for the Swedish and eastern Danish networks39. Prior to the disturbance, operating conditions were stable and well within the tolerances set in the operational planning and grid security assessment. Contingency planning had taken into account the unavailability of several facilities due to annual or scheduled maintenance. These included nuclear generating units in southern Sweden and several transmission lines, including two 400 kV transmission lines connecting central and southern Sweden, the DC interconnectors between Sweden and continental Europe and the Kontek DC interconnector between Zealand and Germany. Eastern Denmark is closely integrated with Sweden through 400 kV and 132 kV AC submarine cables. Total consumption in eastern Denmark and southern Sweden before the event was around 4 850 MW. Around 400 MW of Swedish demand was being met by exports from eastern Denmark. At 12.30 pm, Unit 3 at the Oskarshamn nuclear power plant shut down due to mechanical problems with the valves controlling its feedwater circuits. 1 175 MW of generation was lost as a result of the shut-down, and Nordic system frequency began to fall as a result of the production shortfall. This was a standard N-1 39. See Elkraft System (2003); Svenska Kraftnat (2003) and (2004). Elfrakt System is now part of Energinet.DK.

90

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

contingency event managed through operational reserves. Spinning generating reserves from Norway, northern Sweden and Finland activated automatically to restore the supply-demand balance and to stabilise system frequency. Although power flows increased on the western part of the Swedish transmission system serving loads in southern Sweden, the flows were within pre-determined security limits. After a normal frequency transient the system returned to stable operating conditions in less than a minute without any loss of services. Under the Nordic system security standards, operators have up to 15 minutes to return the system to an N-1 secure state, where it would be able to absorb another similar disturbance without threatening reliable system operations. At 12.35 pm, a double busbar failure occurred at a 400 kV substation on the west coast of Sweden. This represented a serious system failure consistent with an N-2 event. The fault was caused by mechanical failure of an isolator in one of the busbars, which collapsed in a manner that ignited an electric arc between that busbar and the adjacent busbar. Both busbars were automatically disconnected by protection devices, which immediately tripped the circuit-breakers for all incoming lines to both busbars. This was an extraordinary occurrence. As a result of this failure, four 400 kV transmission lines were disconnected. Two of these lines provided a key link between central and southern Sweden, while the other two lines each connected a 900 MW nuclear unit at the Ringhals power station to the transmission network. As a result, the transmission path along the west cost of Sweden was lost, as was 1 750 MW of production with the disconnection of the two Ringhals units. However, Ringhals Unit 4 managed to switch to in-house operating mode and was available to support the subsequent restoration. The sudden loss of generating and transmission capacity triggered large power oscillations, low voltages and a drop in system frequency, initiating automatic under-frequency load shedding. Power flows increased further on the remaining transmission links between central and southern Sweden. Automatic responses from generators in northern Sweden, Norway and Finland to compensate for the loss of the Ringhals units exacerbated these power flows. By this stage, no major generators were left connected to the transmission system in southern Sweden, which substantially reduced the level of reactive power support available for voltage control. During the 90 seconds following the busbar failures the power oscillations began to fade and the system seemed to stabilise. Load levels began to recover, placing further stress on the remaining 400 kV transmission links between central and southern Sweden. As a result, voltage levels on these lines dropped to critical levels. The situation developed into a voltage collapse 91

LEARNING FROM THE BLACKOUTS

in a section of the transmission network southwest of Stockholm. Distance relays on the transmission system in central and southern Sweden registered this event as a distant short circuit, and automatically tripped these transmission lines, severing all remaining transmission connections between northern and southern Sweden. Distance relays for the Oresund interconnectors between Sweden and eastern Denmark did not react as the voltage at the Swedish end of the connection remained relatively high. An electrical island containing southern Sweden and eastern Denmark formed momentarily. However, the large generation deficit led to a collapse of frequency and voltage over a 1 to 2 second period, triggering generator and network protection devices. The islanded system collapsed at 12.37 pm. Eastern Denmark was automatically disconnected from southern Sweden at the time of the voltage collapse by zero-voltage relay protection devices on the Oresund link. The voltage collapse caused damage to the two largest power stations in eastern Denmark. Damage to the largest of these precluded it from supporting restoration activities after the blackout.

■

Impact and Restoration Services in southern Sweden and eastern Denmark were disrupted as a result of the outage. In Sweden, all supplies south of a geographic line between the cities of Norrkoping in the east and Varberg in the west were interrupted, with the exception of some small islands which were being supported by local small hydro power stations that survived the collapse. Figure 23 shows the area affected by the outage. Around 6 350 MW of load was lost at the time of the outage, around 4 500 MW in Sweden and 1 850 MW in eastern Denmark. Four million people were initially disconnected and the economic cost of the disruption has been estimated at around USD 310 million40. In southern Sweden, services were cut to 1.6 million people. Provincial centers were worst affected, along with regional airports and rail services. In eastern Denmark, services were cut to around 2.4 million people. The capital city of Copenhagen was the worst affected area, along with the international airport and rail services. Restoration Process Restoration of most services in southern Sweden and eastern Denmark was successfully achieved by around 7.00 pm, with full service restoration achieved 40. OXERA (2005).

92

Source: Larsson, S. (2004)

Bergum

Stavanger

DENMARK

HVDC

SWEDEN

HVDC

Ringhals

Hasle

GERMANY

Kristiansand

NORWAY

HVDC

POLAND

Oskarshamn

Stockholm

Area Affected by the Swedish-Danish Blackout, 23 September 2003

Figure 23

RUSSIA

LITHUANIA

LATVIA

ESTONIA

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

93

LEARNING FROM THE BLACKOUTS

in eastern Denmark around 10.00 pm. Some technical difficulties were encountered that slowed restoration to some extent, particularly in eastern Denmark. In southern Sweden, restoration was facilitated through the northern part of the network, which had remained intact after the separation. Hydropower from Norway, northern Sweden and Finland was used to re-energise the network and to supply load as it was gradually restored in southern Sweden. Transmission lines and substations were re-energised in a manner that allowed the network to be rebuilt from north to south. Restoration proceeded along the eastern and central 400 kV transmission paths from north to south. Technical difficulties with a faulty disconnector at the Horred substation (near the Ringhals power station) slowed network restoration on the western transmission path. Despite these technical difficulties, the main 400 kV network was successfully re-energised down to the southernmost substations within an hour after the initial collapse. Initial voltages were volatile due to a lack of reactive power before generation was reconnected to the southern network. A serious loss of remote control to a key substation delayed efforts to stabilise voltage and slowed restoration of services. Regional and local network services were restored through feeder transformers connected to the 400 kV network. Initially some restrictions were imposed on restoration of load. However, these were quickly removed as networks returned to service. One of the nuclear units at Ringhals successfully switched to in-house operation during the event and was immediately available to support restoration of load once the network had been restored. Svenska Kraftnat estimates that around 10 GWh of energy was not supplied as a result of the outage. Figure 24 charts the return of load in southern Sweden during the restoration period. In eastern Denmark, preparations to restore voltage to the main transmission network began within minutes of the outage. All loads were disconnected so that the supply-demand balance could be properly controlled during the restoration. To achieve the quickest possible return of services, the network was prepared for voltage restoration from either the black-start capable generators at the Kyndby power station or through the Oresund interconnection with Sweden. Elkraft System’s communications strategy for emergency events was also immediately activated to inform key stakeholders and the public about the event and progress on the restoration. Restoration of voltage through the interconnection with Sweden was preferred as it would normally be expected to provide the fastest and most reliable solution. It was also the most practical route for timely restoration at the time. Damage to regulating equipment, possibly caused by the voltage collapse 94

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

Figure 24

MW (Demand)

Swedish Load Restoration Following the Outage

7 000 Total non-supplied demand 10 GWh

6 000 5 000 4 000 3 000 2 000 1 000

08 -0 9 09 -1 0 10 -1 1 11 -1 2 12 -1 3 13 -1 4 14 -1 5 15 -1 6 16 -1 7 17 -1 8 18 -1 9 19 -2 0 20 -2 1 21 -2 2 22 -2 3

0

Source: Svenska Kraftnat

Time

associated with the outage, meant that the diesel plant at Kyndby was unable to affect black-start restoration of voltage to the eastern Danish transmission system. Svenska Kraftnat had re-energised a 400 kV transmission link to the Swedish end of the Oresund interconnector by 1.30 pm. Voltage had been restored to the transmission system in eastern Denmark through the Oresund link by around 1.45 pm, around 70 minutes after the outage. Initial voltages fluctuated due to a lack of reactive power before generation could be restored to the eastern Danish network. About 200 MW was imported from Sweden to stabilise voltage and provide for initial reconnection of some priority loads. Restoration of further load followed the reconnection of eastern Danish generators. Load was resupplied in step with the reconnection of major generating facilities, based on a pre-determined 10-part reconnection plan agreed between Elkraft System and local distributors. Restoration of load may have been achieved more quickly had some of east Denmark’s large generators successfully switched to in-house operation. Four of ten succeeded temporarily but subsequently failed when the final voltage collapse occurred. Protection systems on generators are usually set to disconnect in response to large frequency or voltage variances. But on this occasion, voltage and frequency 95

LEARNING FROM THE BLACKOUTS

conditions developed in a manner that led to the large power stations disconnecting at a relatively late stage in the event, making switching to inhouse operation mode more difficult than usual. Elkraft System estimates that around 8 GWh of energy was not supplied as a result of the outage. Figure 25 charts the return of load in eastern Denmark during the restoration period using load from the previous day as a benchmark for comparison. Figure 25

Danish Load Restoration Following the Outage MW (Demand)

2 000 Total non-supplied demand 8 GWh

1 500

1 000

500

08 :0 0 09 :0 0 10 :0 0 11 :0 0 12 :0 0 13 :0 0 14 :0 0 15 :0 0 16 :0 0 17 :0 0 18 :0 0 19 :0 0 20 :0 0 21 :0 0 22 :0 0 23 :0 0

0

22 Sep.

23 Sep.

Time

Source: Elkraft System.

■

Key Findings of Investigations and Inquiries Official investigations of this event were undertaken by Svenska Kraftnat and Elkraft System, the transmission system operators responsible for the Swedish and eastern Danish networks respectively. These investigations were largely undertaken as a joint exercise. Both reports note that the system was operating in an N-1 secure condition prior to the first fault and that the operational reserves were deployed appropriately to return the system to a stable operating condition. Svenska Kraftnat notes that the underlying cause of the outage was that a second major fault occurred within the 15 minute period allowed under Nordel 96

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

operating practices for a system operator to return the system to an N-1 secure condition. This second fault shut down two major generators and severely reduced transmission capacity in southern and central Sweden, and occurred only 5 minutes after the first N-1 contingency event. The subsequent disconnection of high voltage transmission lines severely weakened the transmission grid in southern Sweden to an extent that it became impossible to transmit sufficient electricity to meet load. Elkraft System notes that neither increased production in eastern Denmark nor imports from the continent would have been capable of preventing the subsequent voltage collapse. The combination of these events constituted at least an N-3 contingency event, leading to conditions that were beyond the operational limits of the system and beyond prepared countermeasures. Elkraft System concluded that given the present design of the power system and the current security criteria for system operation and infrastructure protection, it was not possible to prevent the blackout once the second failure had occurred. Several other key conclusions can be drawn from these investigations. In particular, large disturbances can stem from a sequence of interrelated faults that would be manageable if they appeared alone. Rapid restoration is also extremely important to minimise the adverse impact of outages on modern societies that are highly dependent on electricity supplies. Similarly, the community has a right to be fully informed about outages and progress toward restoration. The provision of timely information to key stakeholders should be a key performance objective for system operators and other entities involved in managing system restoration after an outage.

■

Key Recommendations Key recommendations from the investigations included the following. ●

System Operation Practices. Examine the Planning and Operational Reliability Standards applied within Nordel with a view to assessing whether current technical standards and operational practices are sufficient to deliver reliable electricity services consistent with the efficient operation of electricity markets and community expectations. Modify technical specifications and the Nordic Grid Operation Agreement where appropriate to accommodate new operational conditions.

●

Emergency Procedures. Consideration should be given to including automatic under-voltage load shedding as a means to manage potential voltage collapse situations, particularly where there is a shortage of 97

LEARNING FROM THE BLACKOUTS

generating capacity. Review protection strategies for generation and network infrastructure to ensure that an appropriate balance is maintained between protecting the infrastructure and maintaining services during emergency situations. Review load-shedding arrangements to ensure that consumer disconnection is appropriately prioritised. ●

Restoration Procedures. Develop procedures to reduce the time taken to prepare the system for black-start, possibly including establishing relays to automatically disconnect the underlying network in the event of zero voltage. In relation to eastern Denmark, consideration should be given to ways of strengthening restoration, including the potential for using generators operating in in-house mode to provide black-start services and for establishing plants dedicated to voltage restoration. Review reconnection plans to ensure that consumer reconnection is appropriately prioritised.

●

Emergency Operation of Generation. Mandatory technical requirements are to be enforced, particularly in relation to managing the transition to inhouse operation following external network disturbances.

●

Tools and Protection Devices. Examine the potential for more advanced measurement and control systems in which information on system conditions from several control areas can be integrated. Such systems are not yet fully developed and their implementation would require an integrated evaluation of the design and system operation strategy for the Nordic synchronous area in order to avoid unplanned disconnections. Consider developing more advanced protection systems, possibly including the potential to detect voltage collapse in the entire system as part of a more integrated strategy for system operation. Investigate the potential to improve remote control functionalities under outage and transient conditions.

●

Communication. Communication strategies should be reviewed and adjusted in the light of this experience, to strengthen timely flow of information to distributors, consumers, authorities and the media as appropriate.

●

Physical Design and Maintenance of the Transmission System. The second fault focuses general attention on the design, inspection and preventative maintenance of the transmission system, particularly at vulnerable points with substantial consequences for reliable system operation. Specific issues include: • examination of the potential for restructuring switching gear at substations to eliminate the risk of flashovers between main busbars; • mandatory inspections of disconnectors and scheduled replacement of critical parts; and

98

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

• review of the methodology and resources applied to manage outsourced maintenance. ●

Investment. Consider the potential for upgrading transmission lines to help improve system reliability. Proceed to reinforce transmission capacity to southern Sweden. New generation being constructed in southern Sweden will help to improve reliability within the region.

CASE STUDY 4: Australian National Electricity Market...... On Friday, 13 August 2004, a current transformer at the Bayswater Power Station in New South Wales developed an internal fault causing it to later explode. This failure triggered a major power disturbance in the Australian National Electricity Market. Five large generating units and a medium generating unit were lost in New South Wales within 25 seconds. Although emergency responses led to the disruption of power supplies in the states of New South Wales, Victoria, Queensland and South Australia, a complete disruption of services was avoided. The following event summary is drawn from the reports prepared by the National Electricity Market Management Company (NEMMCO), which is the independent market and system operator responsible for the Australian National Electricity Market41. At 9.41 pm, a current transformer in the number 1 bay at the Bayswater Power Station switchyard (New South Wales) failed and later exploded. The switchyard has three 330 kV circuit breakers connecting the number 1 generating unit of the Bayswater Power Station and a transmission line (34 Line) to the Liddell Power Station. The current transformer fault was detected by protection devices, which activated 34 Line circuit breakers at the Liddell and Bayswater power stations, isolating the fault from the power system. The great majority of faults that trip transmission lines are transient in nature with a relatively short duration. As a result, transmission line protection systems usually incorporate auto-re-close facilities so that the transmission network can be fully restored as soon as possible after such events. Fifteen seconds after the initial fault, the auto-re-close function operated as designed to automatically re-close the main bus circuit breaker on 34 Line. However, the fault on the current transformer remained, and resulted in the fault effectively being reapplied to the power system when 34 Line’s 41. NEMMCO (2004); (2005a) and (2005b).

99

LEARNING FROM THE BLACKOUTS

protection system automatically re-opened the circuit breaker. But on this occasion generator differential protection on Units 1, 2 and 3 at the Bayswater Power Station also tripped, resulting in the immediate loss of 1 971 MW. Negative phase sequence protection also activated the automatic shutdown sequence for Unit 2 at the Eraring Power Station, ultimately resulting in the loss of a further 424 MW. Sixteen seconds after the initial fault, the auto-re-close function on 34 Line operated again, this time to attempt an auto-re-close of the coupler circuit breaker. However, the current transformer fault was still present and the subsequent auto-re-close procedure effectively applied the fault to the power system for a third time. Twelve seconds later automatic voltage regulator protection tripped Unit 6 at the Vales Point Power Station, resulting in the loss of a further 542 MW. One second later Unit 2 at the Eraring Power Station completed its shutdown sequence and the unit was automatically disconnected from the power system. Around 36 seconds after the initial fault, under-frequency protection activated automatically to trip the Redbank Power Station, resulting in a further loss of 150 MW.

■

Impact and Restoration Within 40 seconds of the original fault, six generating units had been disconnected in New South Wales, resulting in the loss of around 3 100 MW, representing around 14% of total supply to the National Electricity Market. The sudden loss of generation caused a substantial frequency disturbance on the power system. Frequency had fallen to near 49.0 Hz around 6.5 seconds after the initial power plant trips. Emergency under-frequency load shedding in New South Wales, Queensland, Victoria and South Australia combined with frequency control ancillary services (FCAS) procured by NEMMCO from the FCAS market stabilised the situation. However, the subsequent tripping of the Vales Point Power Station resulted in frequency again falling below 49.0 Hz. Available FCAS reserves had been exhausted by that stage, resulting in further emergency under-frequency load shedding in New South Wales, Victoria and Queensland to stabilise frequency. Table 5 shows the distribution of emergency under-frequency load shedding across the National Electricity Market. Over 1 500 MW of load was disconnected as a result of automatic under-frequency load shedding, representing nearly 7% of total demand. Queensland bore the brunt of the impact, losing around 9.5% of regional demand. Victoria also lost significant load, representing 100

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

around 8% of regional demand. However, the impact in Queensland was worsened by the nature of the pre-determined load shedding schedule, which led to the early disconnection of around 250 000 residential consumers. Load shedding arrangements in the other affected states tended to favour the early disconnection of industrial loads and hence minimised the impact on residential consumers. Table 5

Automatic Under-frequency Load Shedding by Region Region

Demand (MW)

Load shed (MW)

As % of demand

New South Wales

9 208

462

5.0

Victoria

5 998

488

8.1

Queensland

5 726

542

9.5

South Australia

1 669

18

1.1

Snowy

28

0

0.0

NEM Total

22 629

1 535

6.8

Source: NEMMCO.

The remaining 1 600 MW power deficit was met through a combination of increased generation, low frequency load relief (via FCAS) and other customer disconnections due largely to related voltage disturbances. In New South Wales these latter load losses totalled a further 342 MW, bringing the total lost load in New South Wales to around 800 MW or 8.7% of regional demand. Immediately after the loss of generation in New South Wales all interconnector transfers into the State increased substantially as reserves were called upon to stabilise the system. Generation reserves from Victoria, Queensland and the Snowy region all contributed to restore the balance. Table 6 shows interregional flows just prior to the incident and the change in inter-regional flows following the disturbance. Power transfers on the QNI interconnector from Queensland to New South Wales exceeded operational security limits by up to 25% following the disturbance, but were reduced below the security limit after around 15 minutes, well within the 30 minute maximum period established in the

101

LEARNING FROM THE BLACKOUTS

Table 6

Inter-regional Power Flows Following the Disturbance Interconnector

Initial (MW) Maximum After 10 Minutes (MW) (MW)

Approximate Line Limits (MW)

QNI to NSW (AC)

213

1 188

1 040

905

Directlink to NSW

0

0

0

N.A.

Snowy to NSW

-15

1 313

1 313

3 000

VIC to Snowy

-100

905

660

760

SA to VIC (AC)

-460

-241

-460

-460

Murraylink to VIC

-80

-3

-3

N.A.

Source: NEMMCO.

National Electricity Code for reducing transmission line overloads. Similarly, power transfers from Victoria to the Snowy region exceeded security limits by up to 20%, but were reduced below the security limit within 4 minutes. NEMMCO had purchased sufficient FCAS to ensure that frequency levels consistent with operating standards could be maintained in the event of a single credible contingency (ie. an N-1 event). This disturbance, however, represented a multiple contingency event. Under current operational practises, NEMMCO does not purchase sufficient FCAS in advance to cater for multiple contingencies given the low probability of their occurrence and the likely considerable additional cost. However, given the emergency, NEMMCO called on all enabled service providers to supply the maximum possible amount of FCAS. Table 7 summarises FCAS service providers’ performance in response to the emergency during the 5-minute dispatch interval commencing at 9.40 pm. Table 7 shows that services delivered were well above the services enabled for the period. Fast response (6 second) raise FCAS provided was over twice the level enabled for the period, while 1 minute raise FCAS provided was over 7 times the levels enabled and 5 minute raise FCAS provided was nearly 2.5 times the levels enabled. Automatic generator governor responses to the frequency drop accounted for a large proportion of the FCAS response. The FCAS response together with the low frequency load relief supported the under-frequency load shedding in containing, stabilising and restoring power system frequency. 102

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

Table 7

FCAS Service Delivery During the Event Fast (6 Second) Raise Slow (60 Second) (MW) Raise (MW)

Delayed (5 Minute) Raise (MW)

Delivered

Enabled

Delivered

Enabled

Delivered

Enabled

NSW

268

69-

531

67

225

90

VIC

69

95

183

80

181

118

QLD

146

75

641

92

206

119

SA

124

36

195

15

27

30

Snowy

5

0

294

0

281

120

Total

612

275

1 844

2 540

920

477

Source: NEMMCO.

Media and community interest focused on Queensland given the impact of load shedding on residential consumers in that state. NEMMCO acted quickly to provide information to key stakeholders and the media. Under the National Electricity Market Emergency Protocol, NEMMCO was also required to inform representatives of member governments. There were some delays in contacting certain officials but NEMMCO notes that these delays did not affect management of the disturbance or delay service restoration. Restoration Process A secure operating state was achieved when frequency was stabilised within the normal operating range at around 9.47 pm, some 6 minutes after the disturbance. At 9.56 pm, all affected loads were restored in South Australia, around 15 minutes after the outage. 50 MW of load was restored in Queensland at 10.02 pm. Two small ‘sensitive loads’ were also restored in New South Wales around this time. Two loads were restored in Victoria at 10.04 pm. Another 50 MW of load was restored in Queensland at 10.05 pm, followed by a further 100 MW at 10.28 pm. At 10.35 pm NEMMCO issued instructions to restore all remaining affected load in Queensland at a rate of 100 MW every 5 minutes. By 10.20 pm, around 40 minutes after the outage, services to all affected load in New South Wales and Victoria had been restored, while most affected services were restored in Queensland by 11.00 pm, around 80 minutes after 103

LEARNING FROM THE BLACKOUTS

the outage. NEMMCO estimates that a little over 1 GWh of energy was not supplied as a result of the disturbance. All but one of the affected generating units in New South Wales returned to service within 24 hours, with the remaining unit returning to service around 44 hours after the incident.

■

Key Findings of Investigations and Inquiries NEMMCO undertook the official investigation of this event drawing on information provided by market participants and transmission network service providers. The primary cause of the event was the failure of a current transformer at the Bayswater Power Station switchyard. Other recent current transformer failures led the transmission owner to install monitoring devices on all similar equipment to provide an early warning of a potential failure. However, the 13 August fault was qualitatively different to the faults previously experienced. The monitoring device was not designed to register this kind of fault, and therefore did not provide any warning of the approaching failure. NEMMCO’s analysis also suggests that a combination of technical failures may have magnified the impact and duration of the event. In particular, protection devices installed at some of the disconnected generators did not appear to operate as intended. For instance, the Bayswater units 1-3 were tripped by differential protection relays, which are intended to detect faults within generating equipment. However, these relays activated even though the current transformer fault was well outside the differential protection relay zone for these units. Similarly, the Eraring Unit 2 was tripped by its negative phase sequence protection systems, which are normally only meant to activate as a last resort response to external uncleared faults that may damage the generating equipment. On this occasion, the time delay function did not operate correctly and the unit was shut down as soon as the fault was detected. Problems with the power system stabilisers at the Vales Point Power Station may also have contributed to faulty operation of its automatic excitation regulator, leading to its shut-down. The Vales Point trip, which initiated the second dip in system frequency, was important in this context given NEMMCO’s conclusion that this second frequency fall significantly delayed eventual recovery of frequency and restoration of services. NEMMCO concluded that the emergency automatic under-frequency load shedding response performed according to design and in conjunction with FCAS operated effectively to stabilise power system frequency within the limits and timeframes established by operating standards, avoiding a catastrophic 104

2 TRANSMISSION SYSTEM SECURITY CASE STUDIES

system collapse. However, some automatic under-frequency load shedding, which should have occurred, did not. Also, the impact of the load shedding may have been more equitably distributed among the regions. This latter issue was the cause of some concern, particularly in Queensland, which bore the brunt of the disturbance. Potential to rebalance load shedding is being investigated, but as NEMMCO notes, it may not be possible to ensure equal sharing between jurisdictions within an integrated regional market for every possible contingency, especially in the case of multiple contingency events. Several other key conclusions can be drawn from NEMMCO’s investigation. Successful management of such a major failure provides a practical example of the benefits for system security resulting from regional integration. Integration helped to strengthen the resilience of the power system in response to emergency conditions. In this case, the system operator had the authority and capability to draw on inter-regional resources in real-time to efficiently stabilise and restore secure operating conditions, which helped to minimise the impact of the event. It demonstrates the value of having a system operator with sufficient authority and scope to prepare operational contingency plans and execute them in real time on a whole-of-market basis in response to emergency situations. The event also revealed potential system security issues associated with load flows on interconnectors. The National Electricity Market (NEM) is built around a largely radial (ie. point to point) bulk transmission system. As a result, key transmission line failures, particularly at points of interconnection between NEM regions, have the potential to immediately isolate regions and magnify the impact of any emergency event within an isolated region. NEMMCO’s analysis suggests that a further generator failure could have tripped the QNI interconnector leading to effective separation of Queensland from the NEM, and potentially exacerbating the impact of the event in Queensland. Temporary overloading of transmission lines occurred as a result of the emergency response, and is to be expected following a multiple contingency disturbance of the kind experienced. The event has raised questions about how to strengthen the resilience of interconnectors and other key transmission lines so that they can support emergency responses and restoration activities to the greatest extent possible.

■

Key Recommendations Recommendations from NEMMCO’s investigation included the following. ●

Operational Standards. National Electricity Code provisions relating to automatic re-closure, generating unit technical standards and incident investigation should be urgently reviewed, as follows: 105

LEARNING FROM THE BLACKOUTS

• The role of automatic re-closure in relation to power system security and the obligations on network service providers and generators should be reviewed to ensure that generating units are required to withstand faults reapplied by re-closure in the network. • Generators required to provide information to NEMMCO to support its investigation of a system disturbance should be required to report to NEMMCO within a reasonable timeframe, 20 business days for example. ●

Emergency Procedures. The under-frequency load shedding arrangements across the National Electricity Market should be reviewed. (NEMMCO has initiated a review of the under-frequency load shedding arrangements.)

●

Generator Protection Relays. Macquarie Generation and Delta Electricity should complete their investigations into the reasons why their respective generating units tripped, and provide NEMMCO with: • details of the conclusions in their reports; • assurances that their generating units can withstand disturbances following credible contingency events that are cleared in primary protection time and with subsequent auto-reclosure onto a fault; and • Redbank Project Pty Ltd. should amend its under-frequency tripping settings as soon as practicable and provide NEMMCO with an assurance that their generating unit can withstand frequency disturbances within the limits of the frequency operating standards.

●

Communication. NEMMCO should review and improve the arrangements for responsible officer (ie. key government contact point) communications following major system incidents.

106

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY Electricity market reform has fundamentally changed the environment for maintaining reliable and secure power supplies. Growing inter-regional trade has placed new demands on transmission systems, creating a more integrated and dynamic network environment with new real-time challenges for reliable and secure transmission system operation. These operational challenges are intensified as transmission capacity is absorbed. Although it would be impossible to cost-effectively eliminate the risk of future blackouts, the case studies demonstrate that opportunities exist to improve the resilience of transmission systems and the flexibility of system operation in response to disturbances and emergency events. Management of system security needs to be transformed to maintain reliable electricity services in this changed operating environment. Clearly defined responsibilities and authority to act are required. Improved operating practices, with greater emphasis on system-wide preparation, and coordination to support flexible, integrated real-time system management are also needed. Effective real-time system operation requires accurate and timely information and state-of-the-art technology to facilitate effective contingency planning, system monitoring, flow management and co-ordinated emergency response. The case studies highlight these issues and demonstrate the potentially substantial cost of failure to adequately address them. These challenges raise fundamental issues for policymakers. Scope exists to clarify responsibilities and accountabilities, and to improve enforcement in this context. Rules and practices governing transmission system security can also be enhanced. New and existing technology could be more fully employed to enhance effective system operation. Asset and vegetation management could be strengthened, while co-ordination, particularly within integrated transmission systems spanning multiple control areas, could be improved. Inadequate responses to these issues have the potential to encourage overly conservative management of transmission capacity at the expense of efficient inter-regional trade, or to potentially leave interconnected networks unduly exposed to the risk of further substantial power failures.

107

LEARNING FROM THE BLACKOUTS

Strengthening Transmission System Security Incentives ............................................................... Effective incentives are required to ensure transmission system security in competitive electricity markets. Incentives must reflect the key characteristics that affect transmission system security. In particular, maintaining system security requires a coordinated effort through the value chain and across integrated electricity systems spanning multiple control areas. A breakdown of coordination can lead to imbalances that can push the transmission system beyond its reliability limits, magnifying the risk of disturbances and rapidly cascading system failures. Incentive structures must also have regard for the implications arising from system operators being responsible for making operational decisions about system security but not being directly exposed to the consequences of those decisions42. Transmission system security also exhibits certain public good characteristics. Although transmission system security is of value to all system users, it is not readily possible to define an exclusive property right for it. As a result, system security is a good that can be consumed by any network user irrespective of their willingness to pay, which creates incentives to use the good and rely on others to pay for it – the ‘free rider’ problem. This could lead to insufficient provision for transmission system security where responsibilities and accountabilities are poorly defined43. Successful incentive structures are built on sound governance principles. In general, sound governance aims to establish clear responsibilities that are aligned with the role and function of each party so that those parties best able to manage a risk or function at least cost have the authority, means and incentive to act, and are held accountable for their actions. All parties whose actions may significantly affect transmission system security should be involved including governments, regulators, system operators, transmission owners, electricity users and generators. Appropriate and inclusive governance arrangements create the foundation for developing a mutually reinforcing web of incentives and accountabilities for maintaining transmission system security. Before electricity liberalisation, industry structure provided the key foundation for effective governance. Each system operator was part of a vertically integrated electricity utility. Within its local control area, each operator had complete authority to co-ordinate and control electricity production and flows to 42. Kirby, B. and Hirst, E. (2002). 43. IEA (2003).

108

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

meet demand while maintaining transmission system security. Flows between control areas were relatively minor and closely controlled. Excess network and generation capacity also assisted system operators in maintaining system security. In this environment, responsibility and accountability for maintaining transmission system security could be clearly and readily allocated to local utilities as their functions and the scope of their activities provided them with the means to effectively co-ordinate the value chain to deliver system security. Reliability standards and regulatory arrangements were built upon the governance foundation created by the vertically integrated industry structure. Standards were typically developed and approved by industry bodies, or developed by industry bodies and approved by technical regulators. Application of standards was often voluntary, and enforcement was in some cases largely left to the industry and peer or public pressure. Regulatory requirements were often defined in very general terms such as the obligation to supply or an obligation to provide a reliable service. In a vertically integrated environment the costs associated with system operator actions to deliver system security were almost always hidden and averaged, with cost plus regulatory regimes typically permitting all such costs to be opaquely passed through to end users. Co-ordination across control areas was less problematic in this less competitive environment, which was more conducive to inter-regional cooperation. Hence, the need for external, independent regulation, clear and legally enforceable standards and greater transparency to strengthen incentives for maintaining transmission system security was less apparent. Electricity market reform has fundamentally altered the underlying drivers for sound governance and weakened previous arrangements for maintaining effective transmission system security. In particular, unbundling and independent, decentralised decision-making in response to competitive commercial incentives has greatly reduced system operators’ capacity to co-ordinate and control production and electricity flows throughout the value chain. At the same time, system management practices are receiving greater scrutiny from market participants whose commercial interests are now directly affected by any interventions to manage reliability and system security. These commercial interests, arising as a result of electricity reform, raise questions about transparency, objectivity and legal liability associated with system operation. Independent, decentralised decision-making has created a much more dynamic real-time operating environment, while the concurrent growth of electricity trade across integrated control areas has exposed local system operators to a 109

LEARNING FROM THE BLACKOUTS

degree of systemic risk not previously experienced. Together these new patterns of network use have greatly complicated the task of maintaining transmission system security, especially in real-time and across integrated control areas. This would imply a greater need for co-ordination and cooperation to successfully manage system security in more dynamic and integrated electricity markets. However, the combination of these factors may have in effect weakened incentives for co-ordination and information exchange between system operators, particularly where they remain within vertically integrated utilities that are competing in regional wholesale markets. Functions previously concentrated within vertically integrated utilities are now spread among system operators, generators and users, while the emergence of more integrated regional electricity markets has seen a horizontal spreading of functional responsibility for transmission system security across multiple control areas. Governance arrangements need to accommodate these fundamental structural and market changes. Opportunities exist to improve the governance framework in ways that will help to strengthen transmission system security in a manner that is consistent with the emergence of efficient electricity markets, and which provides scope for flexible evolution of standards and practices to accommodate new technologies and market developments. An effective framework should: ●

clarify individual and shared responsibilities for transmission system security;

●

align accountabilities with the new functional responsibilities resulting from unbundling;

●

ensure the boundaries of authority to act are specified for each party, and that parties have sufficient authority to undertake their responsibilities within those boundaries;

●

provide strong incentives for effective co-ordination and information exchange, within the value chain and across systems spanning multiple control areas, reflecting the shared nature of responsibility for aspects of transmission system security;

●

create transparency and objectivity, given the potential commercial implications of system operators’ actions in competitive electricity markets;

●

strengthen coverage, accountability and enforcement, where necessary, to help reinforce incentives for providing appropriate levels of transmission system security, and to build the credibility of the regime;

110

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

●

be applied consistently across an intergrated transmission system; and

●

balance market requirements for access to transmission capacity with transmission system security requirements.

Addressing these challenges raises policy issues in relation to the legal, regulatory and structural framework. It also raises questions about the nature, development and application of security standards and operating practices employed to maintain transmission system security in competitive electricity markets.

Legal, Regulatory and Structural Framework...................... Effective legal and regulatory regimes provide an essential foundation for establishing an effective governance and incentive structure for ensuring transmission system security in competitive electricity markets. Ideally, the legal and regulatory regime should clearly identify the responsibilities and accountabilities of the various parties affecting transmission system security and provide a means to codify those responsibilities and accountabilities. Legal and regulatory regimes should also clearly delineate the nature and scope of authority each stakeholder has to discharge their responsibility. They also need to provide an effective mechanism for enforcement, especially where other incentives for appropriate action may be relatively weak. Legal and regulatory arrangements could be expected to emerge as the principal means for establishing effective governance and incentive structures for transmission system security in competitive markets, replacing previous substantial reliance on the vertically integrated local utility structure. However, there may be a transitional ‘vacuum’ while these arrangements are being developed and implemented. During this period there is a risk that responsibilities and accountabilities may remain poorly defined or default to system operators who may no longer have sufficient authority to effectively manage them. The case studies highlight several examples of such uncertainty and poorly defined responsibilities. For instance, concerns about this uncertainty were raised by the UCTE in the context of its investigation of the Swiss-Italian blackout. UCTE notes that uncertainty regarding the role and authority of transmission system operators (TSOs) to act, particularly in emergency situations, needs to be removed. Legal and regulatory clarity is required to define the boundaries of TSO responsibilities and accountabilities so that they can be empowered to effectively manage system security44. 44. UCTE (2004a).

111

LEARNING FROM THE BLACKOUTS

Another dimension of this uncertainty was raised in the Swedish-Danish and Australian case studies. Both cases suggest that poorly defined obligations on generators to ensure that plants could withstand credible contingencies may have contributed to their failure to meet specified performance requirements45. These examples highlight the importance of clarifying roles, responsibilities and accountabilities within a legal and regulatory framework enabling effective enforcement, and the need to remove or minimise any related uncertainties as quickly as possible. Several legal and regulatory approaches have been employed in competitive electricity markets to create accountability and to provide a basis for enforcement. Legal instruments, including legally enforceable codes that specify responsibilities and accountabilities in relation to transmission system security, have been adopted in the United Kingdom and by jurisdictions participating in the Australian National Electricity Market. In the United Kingdom, code requirements have been augmented with specific transmission system operator licensing agreements incorporating a more precise interpretation of the requirements for reliability and system security46. In the United States, the recently enacted Electricity Modernisation Act 2005 provides the legal basis for the creation of an independent regulatory framework to define, monitor and enforce transmission system security standards47. A mandatory system of reliability standards has also applied in Ontario since market opening in 2002. Under this arrangement, the Independent Electricity System Operator (IESO) has statutory authority to maintain the reliability of the IESO-controlled grid in accordance with NERC reliability standards. The electricity market rules reference the Northeast Power Coordinating Council (NPCC - the regional reliability coordinator) reliability requirements and require market participants to meet these requirements through compliance with the market rules. The IESO assesses and enforces compliance with these reliability provisions and can restrict participation or impose financial penalties on market participants. The NPCC monitors compliance by the IESO and can sanction the IESO. The regulator, the Ontario Energy Board, acts as an appeal body to these compliance and enforcement decisions. A System Operator Agreement has been implemented by the Nordel transmission system operators48. This agreement addresses issues of shared 45. 46. 47. 48.

NEMMCO (2005a) and Elkraft System (2003). See National Grid Transco (2004) and (2005) for details. United States Congress (2005). Nordel (2004a).

112

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

responsibility in relation to managing reliability and system security within the Nordic electricity market. It is designed, among other things, to facilitate effective co-operation and information exchange on matters affecting joint system security. It incorporates agreed standards for system security, agreed principles for managing operational disturbances and agreed procedures for procuring operating reserves through a market-based process. Although the agreement is considered binding by member transmission system operators, it does not appear to be legally enforceable. Nordic governments separately prescribe certain system security requirements and emergency responses, which are enforceable through national legislation. Contractual mechanisms have also been employed to ensure accountability and compliance. UCTE has recently implemented a multilateral agreement to make the system security and reliability standards contained in the UCTE Operation Handbook legally binding for all member transmission system operators in continental Europe. The multilateral agreement entered into force on 1 July 200549. A key objective of the multilateral mechanism is to help clarify system operator responsibilities and to provide a legal framework for undertaking system operation functions. The mechanism also incorporates sufficient flexibility to permit ongoing evolution of the underlying standards without invalidating the agreement. In North America, the Western Systems Co-ordinating Council has employed a Reliability Criteria Agreement with independent power producers and transmission service providers to enforce reliability requirements in restructured electricity markets. This agreement is similar to a bilateral contract that specifies compliance requirements for system operators in relation to operating reserves, disturbance control, control performance and operating transfer capability. Penalties for violation range from a letter to the chief executive through to fines of USD 10 000 or USD 10/MWh, whichever is higher, and include an increasing contingency reserve penalty for a three-month period following any violation of the disturbance control criteria. Although coverage is voluntary, initial results suggested that there had been an improvement in compliance with reliability criteria following its introduction50. NERC has also sought to enhance its compliance audit program following the 2003 blackout and to increase peer and public pressure for compliance by posting the names of violators on its website. The recently enacted legislation will substantially strengthen the effectiveness of the compliance auditing program by providing an effective means of ensuring compliance and enforcement51. 49. UCTE (2005a). 50. Kirby, B. and Hirst, E. (2002). 51. NERC (2005b).

113

LEARNING FROM THE BLACKOUTS

Several challenges associated with clarifying responsibility, compliance and enforcement of transmission system security may arise. Opportunities clearly exist to move toward more comprehensive coverage and enforcement, and efforts are being made to do so. One significant challenge relates to defining responsibilities and accountabilities. For instance, precise clarification of roles, responsibilities and accountabilities may prove a considerable practical challenge in this context, especially where the boundaries of responsibility may be overlapping and somewhat ambiguous. This could occur, for instance, at the interface between control areas in the context of managing parallel flows, or at the interface between an independent system operator and transmission owners in the context of managing network control ancillary services. The European Union has sought to address this issue in its draft security of supply directive. Article 4 of the draft directive envisages that transmission system operators would be largely responsible for ensuring transmission system security, subject to performance standards endorsed by a member state or an appropriate national regulator52. However, assigning responsibility and accountability for ‘operational reliability’ largely to TSO’s may prove problematic in practice. As discussed previously, unbundling and decentralised decision-making has created a new operational environment where system security is essentially a responsibility shared across the value chain and among all TSO’s within an integrated transmission system. Inappropriate allocation of responsibility in this context has the potential to distort system operator behaviour, with the potential to produce overly conservative management of transmission systems at the expense of efficient interregional trade, or to encourage actions that may weaken overall system security. Member states will face the challenge of precisely specifying these responsibilities and accountabilities. Management of shared responsibilities that cannot be readily allocated to individual stakeholders will require the development of innovative legal and regulatory approaches. Prior to the recently enacted Energy Policy Act in the United States, NERC had begun to address this challenge through the development of a functional model. The model identifies 17 basic functions associated with operating a transmission system, and with maintaining power system reliability. It aims to provide a means for identifying key functions and assigning responsibilities for performing each function, irrespective of the institutional framework applied. It also identifies the 52. European Parliament (2005).

114

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

key tasks and minimum capabilities each responsible entity should have to undertake each function, and the interaction between functions both in terms of the inputs needed from other functions to undertake a particular function and in terms of the other functions that are directly affected by the performance of each function. Responsibilities can be combined under this model, however, they cannot be split between different responsible entities. A single responsible entity would remain accountable for undertaking all tasks within a functional responsibility, including any tasks assigned to other parties. In this way accountabilities would remain clearly and precisely allocated between all responsible entities. Key features of the functional model are presented in Annex 2. The model has received cautious support from the industry and is undergoing further development and evaluation. It clearly illustrates the complex interdependence between functions, reflecting the shared nature of responsibility for key elements of transmission system security and reliable delivery of electricity services in general. It also highlights the challenge associated with defining functional responsibilities and assigning accountabilities to create incentives for effective, coordinated management of all aspects of reliability in reformed electricity markets. Meeting this challenge requires a sophisticated response. The NERC proposal represents a positive a step in this direction. Care also needs to be exercised in relation to defining obligations in the context of establishing performance contracts for transmission system security. For instance, during the early phase of the merchant power plant development in North America, some system operators entered into ancillary services contracts with independent power producers (IPPs) that failed to explicitly account for certain services like reactive power for voltage control. IPPs that signed these contracts had no guarantee of payment for these services, and reportedly balked at providing them. Similar lack of incentives to provide ancillary services may exist in competitive markets where these services can only be provided at the expense of producing active power for sale on wholesale markets.53 Common definitions of responsibility for system security may also prove to be a challenge across integrated networks spanning multiple jurisdictions and control areas. A recent Nordel report to Nordic energy ministers noted that regulations defining system security responsibilities vary across the Nordic market and recommended that a common definition be developed with legislation and regulations across the Nordic market harmonised accordingly54. A similar argument can be made in favour of implementing 53. Kirby, B. and Hirst, E. (2002); Bialek, J. (2004). 54. Nordel (2005a).

115

LEARNING FROM THE BLACKOUTS

consistent rules for system security across control areas within integrated regional transmission systems55. Mechanisms designed to uphold legal rights, enforce accountabilities, resolve disputes and determine compensation for damages need to be credible – in other words, they need to be consistent, objective and transparent. The challenge here, given the likely technical nature of any dispute, is to create a credible and timely process that produces fair and binding rulings. Standard legal procedures would achieve binding rulings, but may not have access to sufficient technical expertise to enable effective adjudication in a timely manner. Conciliation or arbitration processes could be established to address technical expertise and timeliness, but may not have sufficient legal standing to deliver binding rulings. Questions may also be raised over the independence of conciliation or arbitration processes, depending on how they are constituted. Issues surrounding the determination of compensation in the event of damages may also prove problematic given the shared nature of the responsibility for many aspects of transmission system security. UCTE is in the process of developing a conciliation process in the context of its multilateral agreement56. Exposure to legal liability has the potential to strengthen incentives on system operators to effectively manage transmission system security. However, caution needs to be exercised. The shared nature of responsibility for certain aspects of transmission system security may make application of legal liability problematic. For instance, individual responsibility and liability for the events described in the case studies would be impossible to determine objectively. Full exposure is intuitively appealing as it has the potential to maximise system operator incentives to effectively manage transmission system security. However, it may also encourage behavior that is not conducive to efficient system operation or to effectively managing system security. For example, a system operator that is fully exposed to liability for system security failures would be encouraged to adopt very conservative operating practices that may be inconsistent with promoting efficient trade and market development. Full exposure to legal liability may also discourage transparency and information exchange, which could serve to reduce transmission system security and the quality of co-ordinated responses to manage disturbances in integrated transmission systems spanning multiple control areas. It may also discourage efficient use of operational tools to manage emergency events. Uncertainty regarding legal liability was cited to help explain the reluctance 55. Alvarado, F. and Oren, S. (2002). 56. Vandenberghe, F. (2004).

116

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

of system operators to employ load shedding to help manage the 14 August 2003 blackout in North America. The US-Canada Power System Outage Task Force Final Report indicates that effective application of 1 500 MW load shedding in Ohio could have prevented the disturbance from cascading and ultimately disconnecting over 60 000 MW57. Conversely, a complete liability exemption could substantially weaken incentives for system operators to effectively manage transmission system security, and possibly even encourage less-than-prudent operational behaviour. A balance needs to be struck to ensure that exposure to legal liability provides appropriate incentives for transparency, information exchange and prudent management of system security. Limited liability arrangements established on an ex ante basis may provide a practical approach in this context.

■

Regulatory Arrangements Effective regulatory arrangements are essential for monitoring and enforcing compliance with transmission system security standards in restructured electricity markets. They provide a practical institutional framework for reinforcing responsibilities and accountability, and for applying regulatory discretion and judgment to resolve remaining uncertainty. In an unbundled environment, effective regulatory arrangements have the potential to assure competing market participants that system operators’ decisions in relation to transmission system security are objective, nondiscriminatory and comply with security standards. This is important given that: system operation is a natural monopoly, system operators are largely separated from the financial consequences of their system security decisions, and these decisions can have substantial commercial consequences for competing market participants. Management of transmission system security has implications for electricity trade and market efficiency. Ideally, regulatory arrangements should be independent, with regulatory processes characterized by transparency, objectivity and consistency. Such arrangements have the potential to ensure effective monitoring, compliance and enforcement of transmission system security standards, while building credibility and confidence in the regulatory framework among stakeholders. Prior to reform, regulatory structures were created to support the development, implementation and enforcement of reliability standards. In North America, 57. United States-Canada Power System Outage Task Force (2004a).

117

LEARNING FROM THE BLACKOUTS

for example, the North American Electric Reliability Council and related regional reliability councils were formed following the 1965 Great North Eastern Blackout to provide industry self-regulation including the development, implementation and enforcement of reliability standards. Other jurisdictions pursued the formation of technical regulatory bodies to oversee and approve industry-developed reliability standards. However, an institutional framework based largely on industry self-regulation may no longer be considered credible by market participants in competitive electricity markets. Concerns have been raised about the independence and objectivity of such arrangements. Conflicts of interest could compromise the development, application and enforcement of system security rules, or lead to inertia as competing interests are unable to resolve rule-making issues in a timely or effective manner. The US-Canada Power System Outage Task Force final report suggests that vested interests may have compromised the organisational and financial independence of regional reliability councils to a degree that may undermine their effectiveness. Box 1 provides an example elaborating on these concerns. The Task Force made several recommendations to address these deficiencies, including the development of a stronger and more independent institutional framework and funding mechanisms approved by an independent regulator58. Box 1 . East Central Area Reliability Co-ordination Agreement (ECAR) and Regulatory Independence ECAR was established in 1967 as a regional reliability council, to “augment the reliability of the members’ electricity supply systems through co-ordination of the planning and operation of the members’ generation and transmission facilities.” a ECAR’s membership includes 29 major electricity suppliers serving more than 36 million people. ECAR’s annual budget for 2003 was USD 5.15 million, including USD 1.775 million paid directly to NERC.b These costs are funded by its members in a formula that reflects megawatts generated, megawatt load served and miles of high voltage lines. AEP, ECAR’s largest member, pays about 15% of total ECAR expenses. FirstEnergy pays approximately 8 to 10%.c Utilities “whose generation and transmission have an impact on the reliability of the interconnected electric systems” of the region are full ECAR members, while small utilities, independent power producers, and 58. United States-Canada Power System Outage Task Force (2004a).

118

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

marketers can be associate members.d Its Executive Board has 22 seats, one for each full member utility or major supplier (including every control area operator in ECAR). Associate members do not have voting rights, either on the board or on the technical committees that do all the work and policy-setting for the ECAR region. All of the policy and technical decisions for ECAR, including all interpretations of NERC guidelines, policies and standards within ECAR, are developed by committees (called “panels”), staffed by representatives from the ECAR member companies. Work allocation and leadership within ECAR are provided by the board, the Coordination Review Committee and the Market Interface Committee. ECAR has a staff of 18 full-time employees, headquartered in Akron, Ohio. The staff provides engineering analysis and support to the various committees and working groups. Ohio Edison, a FirstEnergy subsidiary, administers salary, benefits and accounting services for ECAR. ECAR employees automatically become part of Ohio Edison’s (FirstEnergy’s) 401(k) retirement plan; they receive FE stock as a matching share to employee 401(k) investments and can purchase FE stock as well. Neither ECAR staff nor board members are required to divest stock holdings in ECAR member companies.e Despite the close link between FirstEnergy’s financial health and the interest of ECAR’s staff and management, the investigation team found no evidence to suggest that ECAR staff favour FirstEnergy’s interests relative to other members. ECAR decisions appear to be dominated by the member control areas, which have consistently allowed the continuation of past practices within each control area to meet NERC requirements, rather than insisting on more stringent, consistent requirements for such matters as operating voltage criteria or planning studies. ECAR member representatives also staff the reliability council’s audit program, measuring individual control area compliance against local standards and interpretations. It is difficult for an entity dominated by its members to find that the members’ standards and practices are inadequate. But it should also be recognized that NERC’s broadly worded and ambiguous standards have enabled and facilitated the lax interpretation of reliability requirements within ECAR over the years. a. ECAR “Executive Manager’s Remarks,” http://www.ecar.org. b. Interview with Brantley Eldridge, ECAR Executive Manager, March 10, 2004. c. Interview with Brantley Eldridge, ECAR Executive Manager, March 3, 2004. d. ECAR “Executive Manager’s Remarks,” http://www.ecar.org. e. Interview with Brantley Eldridge, ECAR Executive Manager, March 3, 2004. Source: US-Canada Power System Outage Task Force (2004a).

119

LEARNING FROM THE BLACKOUTS

NERC reports that considerable progress has been made to improve organisational independence since the 2003 blackout, with all regional reliability councils currently assessed as having open and inclusive membership and fair and balanced governance arrangements59. Notwithstanding recent progress, the ECAR example illustrates the potential difficulties with industry selfregulation of transmission system security in competitive electricity markets. One option to address the issue of independence would be to allocate regulatory responsibility for transmission system security functions to existing industry regulators, as proposed in the joint report of the Italian and French regulators into the Swiss-Italian blackout. Not only would this option deliver institutional independence, it could also facilitate a more integrated and objective regulatory approach to related issues of transmission system security and electricity trade, allowing issues of non-discriminatory system operation to be supervised in an appropriate regulatory context. However, economic regulators may not possess sufficient technical competence to effectively verify and enforce compliance with transmission system security requirements. In addition, they may not be capable of effectively supervising processes to develop system security rules. This goes to the heart of the issue of regulatory credibility and raises something of a conundrum: an independent body may lack the technical expertise required to effectively regulate system security whereas a technically competent body may become conflicted and exposed to capture by vested interests. The balance between independence and competence needs to be carefully managed. These issues were addressed in the Australian National Electricity Market by establishing a separate regulatory body – the National Electricity Code Administrator (NECA)60. NECA was legally empowered to regulate the system operator to ensure that it complied with transmission system security and other standards specified in the National Electricity Code. NECA was advised on technical matters by the Reliability Panel, which drew its membership from industry experts and transmission owners. The Reliability Panel advised NECA about reliability standards and made recommendations for amendments as appropriate. The Electricity Modernisation Act 2005 recently enacted in the United States proposes a similar approach. It provides the legal basis for the creation of an independent technical regulator, the Electric Reliability Organisation (ERO), which will be supervised by the Federal Energy Regulatory Commission (FERC). 59. NERC (2005b). 60. These roles and functions were transferred to the Australian Energy Market Commission, effective from 1 July 2005.

120

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

The ERO will have authority to make binding reliability rules. It will also have legal authority to monitor and enforce compliance among control areas within the existing NERC organisational boundary. It will remain an industry body able to draw on the expertise of industry participants. However, its independence will be assured by FERC, which has authority to supervise its activities and to approve its funding arrangements. Its governing board will be representative of all stakeholders. FERC has recently released a Notice of Proposed Rulemaking detailing its proposals for establishing the ERO and related institutional arrangements61. The new legislation requires FERC to issue final rules for implementing the new reliability arrangements by February 2006. Another emerging issue relates to the consistency of regulatory monitoring and enforcement across integrated transmission systems spanning multiple control areas and jurisdictions. A minimum goal should be to establish consistent reliability rules across integrated electricity systems. Approaches adopted by NERC, UCTE, Nordel and in the Australian NEM have sought to establish consistent rules across each region, however, interpretation, application and enforcement can vary considerably across jurisdictions within these regions. There may be a case for developing a single independent regulatory structure to facilitate transparent, objective, and consistent monitoring and enforcement across integrated transmission systems. A single regulatory body has recently been established to manage an integrated transmission system spanning several regulatory jurisdictions in Australia. The Australian Energy Regulator (AER) assumed responsibility for economic regulation of the Australian National Electricity Market and related transmission systems in June 2004. It is planned to expand the AER’s regulatory functions over the next two years to include economic regulation of electricity distribution, and gas transmission and distribution in all states and territories except Western Australia. Once this transfer of regulatory authority has been completed, the AER will have effectively replaced at least 13 jurisdictional regulators at the federal, state and territory level, which were previously responsible for regulating electricity and gas networks in 8 jurisdictions. Creation of a single regulator structure among multiple jurisdictions raises considerable challenges, particularly in relation to cross-jurisdictional empowerment of the body and oversight arrangements. The challenges can be magnified in an international context, particularly where regulatory arrangements differ fundamentally between countries within an integrated transmission system. For instance, regulatory oversight of transmission systems

61. See FERC (2005) for details.

121

LEARNING FROM THE BLACKOUTS

is a federal responsibility in the United States and a provincial responsibility in Canada, which could complicate negotiations around the creation of a single regulatory structure. Efforts to co-ordinate related regulatory activities become particularly important in the absence of such a structure. Detailed and timely information will be required to support effective regulatory arrangements. Regulators need to be empowered to collect sufficient information to enable them to effectively monitor and verify compliance, and to pursue enforcement cases where necessary. Information gathering powers should be prescribed in a manner that will minimise the compliance burden and provide appropriate protection of confidentiality.

■

Structural Arrangements Effective structural arrangements can greatly reinforce the incentives created for transmission system security by the legal and regulatory framework. Structural arrangements should aim to clearly combine roles and functions in a manner that strengthens commercial incentives for efficient reliability and markets outcomes. It should also seek to minimise the potential conflicts of interest that may dilute incentives for effective management of transmission system security. Structural arrangements possess vertical and horizontal dimensions. Vertical issues refer to how functions and responsibilities for transmission system security are allocated within the electricity value chain, while horizontal issues refer to how functions and responsibilities are allocated across integrated transmission systems. Given the nature of transmission system security, structural issues in this context largely relate to the allocation of system operation functions. In several markets, system operation functions remain within partially or fully vertically integrated utilities. Combination of system operation, which is essentially a natural monopoly function, with contestable functions such as generation raises immediate concerns about the independence and credibility of system operation. As noted earlier, system operation is of fundamental importance to efficient market operation and effective provision of reliability. Confidence in electricity markets is closely linked to the credibility of its institutions, and particularly to the perceived and real independence and objectivity of its system operators. The significance of this issue has been recognised. In most cases where system operation functions remain within a vertically integrated structure, they have been organisationally, financially and, in some cases, legally separated from 122

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

contestable functions. External regulation has been established to monitor the activities of system operators to ensure that they act in a nondiscriminatory manner. However, concerns are beginning to emerge about the effectiveness of these arrangements. For instance, a recent analysis of transmission loading relief (TLR)62 in the United States suggests that system operators within some vertically integrated utilities may have used the TLR mechanism to curtail scheduled inter-regional power exchanges from generators competing with their incumbent parent utility. Such actions may be very difficult to distinguish from legitimate use of the TLR mechanism to manage congestion; raising the question of whether a regulator would be well placed to identify and respond in a timely manner to any such abuse should it arise. At the very least such concerns can raise the perception that system operators within vertically integrated utilities may be conflicted and may use reliability as a means of masking anti-competitive behavior63. Independence and objectivity could be strengthened by ownership unbundling of system operating functions from contestable components of the value chain. Ownership unbundling would help to reinforce incentives for transparent and non-discriminatory behaviour that are more closely aligned to efficient management of system security and market-based dispatch. It would also remove any related conflict of interest that could influence operator interventions and possibly adversely affect transmission system security. Ownership unbundling of system operation functions could help to enhance the real – and perceived – credibility of system operation in competitive electricity markets. Transmission system security could be strengthened through greater aggregation of system operating functions across integrated transmission systems spanning multiple control areas. North America and continental Europe, in particular, have large integrated transmission systems with many control areas, which presents substantial challenges for effective co-ordination and management of transmission system security, particularly in response to disturbances and emergency situations. This challenge is probably greatest in North America where transmission system security within the area covered by NERC is managed by around 140 separate control areas within 3 integrated transmission systems. Figure 26 illustrates the extent of the challenge. 62. The transmission loading relief mechanism was introduced by NERC for the North American Eastern Interconnection in the late 90s, essentially to provide a means of managing potential overloading of key transmission lines. The TLR mechanism incorporates 6 main levels reflecting increasing risk to the transmission system. TLR levels 3-5 involve actual curtailment of transfers, while TLR level 6 signifies an emergency action. 63. Moss, D. (2004).

123

124

SPPC

DOPD

AVA

CFE

IID

CISO

NEVP

GCPD

CHPD

PSEI

SRP

TEPC

PACE

DEAA

GRMA

AZPS

WALC

IPCO

MPCO

ESBI

PNM

WACM

PSCO

WAUW

EPE

SPC

SPS

SECI

WFEC

ERCO

DEMC

MEC

GRDA

CSWS

EDE

INDN MPS KCPL

OKGE

GRE NSP SMP

OPPD

OTP

LES KACY

WR

WPEK

NPPD

WAUE

MHEB

Source: US-Canada Power System Outage Task Force (2004a).

as of September 2003

LDWP

SMUD

PACW

PGE

BPAT

TPWR

SCL

BCHA

CLEC

SPA

CILC

LAFA

DENL

CWLD

LEPA

AECI

WPS

LAGN

EES

BCA

SME

DEEM

FE

IMO

HE

AEC

SOCO

RC

SCEG DESG

TEC

FPC

GVL

SERU TAL

SETH

SEHA

JEA

SC

YAD

FMPP

SEC

DLCO

CPLW DUK

OVEC AEBN EKPC

TVA DEMG DEMT

LGEE

CIN

AEP DPL AEWC DELO DEWO

IPL

MECS

BREC DEMK

SIGE

DEVI

AEWI

NIPS

DEAM

AEGL

EEI

AMRN

IP

SIPC

CE

ALTE WEC

MGE

UPPC

DELI MPW AELC

CWLP

ALTW

DPC

MP

NSB

HST

CPLE

FPL

VAP

PJM

NYISO

HQT

ISONE

MAR

WECC

MAPP

ERCOT

FRCC

MAAC

SPP

MAIN

ECAR

NPCC

SERC

NERC reliability regions

Major AC transmission line Dynamically controlled generation Alternating current (AC) to direct current (DC) converter Back-to-back AC to DC and DC to AC converter

NERC Regions and Control Areas

Figure 26

LEARNING FROM THE BLACKOUTS

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

The fractured responsibility for transmission system security resulting from this distribution of functions has led to the creation of a complex institutional framework to help manage transmission system security across each integrated transmission system involving reliability co-ordinators, regional reliability councils and, ultimately, NERC. These issues have been recognised and several different approaches are emerging in competitive electricity markets to address them. In the United States, a system of regional transmission organisations is being implemented to help co-ordinate the activities of individual control areas within larger regional markets. In the Nordic region, independent system operation is undertaken by transmission system owners in each country with co-ordination achieved through co-operative agreements that address operational standards and emergency procedures. In the United Kingdom, system operation is undertaken by a single independent transmission owner, and market operation is independent of the transmission owner. In Australia, an integrated market and system operator has been established to enable effective whole-of-market system operation for the interconnected transmission network. The market and system operator is independent of the transmission network owners. In the case studies, institutional complexity created a degree of uncertainty and co-ordination problems, particularly in the North American and Swiss-Italian cases, which may have delayed timely and appropriate system operator intervention. By contrast, the Australian case study illustrates the benefits of an integrated, independent system operator (ISO). In this case, NEMMCO was able to efficiently mobilise resources from several inter-connected regions to successfully manage a major system emergency that could easily have led to system-wide collapse. Although some load shedding occurred, system collapse was avoided and services were fully restored within 1.5 hours. This suggests that institutional arrangements involving a single ISO with clear responsibility for an entire integrated network can provide an effective means of managing emergency events in real time to minimise impacts and deliver least-cost outcomes. Greater consolidation of system operation functions across integrated transmission systems has the potential to reduce transaction costs, increase consistency of application of reliability rules and improve real-time coordination and action to manage transmission system security. However, these benefits may accrue at the expense of more flexible, local responses to manage system security. More localised control areas may provide greater flexibility to accommodate local differences in market rules, structures, generation and fuel mix and public policy.64 64. Vandenberghe, F. (2004); and Kirby, B and Hirst, E. (2002).

125

LEARNING FROM THE BLACKOUTS

Excessive aggregation has also been criticised for potentially exposing integrated transmission systems to greater risk of system operator error, which could have a more immediate impact on power flows over a much larger area compared to system operation based on more localised control areas. There may be an intuitive appeal to this view. However, any such benefits associated with highly localised system operation could be outweighed by the additional co-ordination costs and inherent inertia such arrangements suffer in responding to emergency situations that need to be addressed in real time. Rather than reducing the potential for error, multiple system operators acting in relative isolation with little knowledge of real-time system conditions beyond their control areas may be more likely to misjudge real-time conditions and, hence, to make erroneous interventions that may threaten system security within and beyond their control areas. Integrated alternating current transmission networks are particularly susceptible to the failure of any integrated component, the effect of which can cascade through an interconnected network almost instantaneously. Hence, the impact of an error is not necessarily lessened by localised system operation compared to a more consolidated model. Further, effective co-ordination may prove difficult or even impossible under certain real-time conditions in an environment where multiple system operators need to co-ordinate their actions to identify and manage an emergency event. A single independent system operator with a holistic, real-time perspective of an integrated transmission system and with sufficient resources to manage credible emergency events is in a strong position to effectively respond to such events in a manner that minimises their overall impact. A more compelling concern may be the practical challenges of further aggregation. It has been claimed that it would not be practically possible for a single system operator to effectively manage and coordinate the many thousands of simultaneous transactions undertaken at any time in a large integrated transmission system like the Eastern Interconnection in North America or the UCTE system in continental Europe. Effective system operation would require more localised control, even if under the banner of a single system operator, which would imply the need for ongoing coordination of system operator activities65. Furthermore, a single integrated ISO spanning multiple jurisdictions may not be acceptable from a public policy perspective. It may also raise legal, technical and organisational challenges in a cross-jurisdictional context, not

65. Alvarado, F. and Oren, S. (2002).

126

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

the least of which may involve convincing incumbent transmission system operators to give up some control over their transmission assets66. Under an ISO model, separation of the system operation functions from transmission ownership may also raise challenges associated with clarifying the respective technical roles and responsibilities for co-ordinating system operation and network management. It also raises the challenge of ensuring that the ISO acts in a manner that does not devalue transmission assets or in any other way cause undue financial loss for the transmission owner. Notwithstanding these concerns, appropriate aggregation of system operation functions can strengthen co-ordination of system security, leading to more efficient responses to disturbances and emergency events in real time. The potential benefits are too great to be ignored. A key issue revolves around defining an appropriate level of aggregation that balances the potential benefits and costs.

Improving Transmission System Security Standards.......... Official investigations of the events described in the case studies focused considerable attention on the rules and practices governing transmission system security. These reviews suggest that opportunities exist to reform operating rules and practices so that they can provide a more effective means of ensuring transmission system security in the more dynamic, real-time and integrated operating environment created by electricity market reform. Transmission system security standards have typically changed little since the introduction of electricity market reform. In practice, great reliance is placed on the N-1 standard. This methodology has the advantage of being relatively simple to compute, implement and monitor. It is typically applied in a deterministic manner67 based on the notion of managing a single credible contingency and returning the transmission system to a stable and secure condition within a minimum timeframe, usually somewhere between 15 and 30 minutes. It provided an effective and practical framework for managing system security in the operational conditions existing prior to electricity reform.

66. Bialek, J. (2004). 67. Deterministic application in this context refers to the criterion being applied in a manner that delivers the same level of protection to all points in an integrated system (or control area) at the same time, irrespective of the likelihood of failure or the potential impact failure would have on the overall reliability of an integrated transmission system.

127

LEARNING FROM THE BLACKOUTS

However, the case studies reveal that transmission systems have become more dynamic with the advent of electricity market reform. Transmission systems are driven more efficiently and closer to security limits for longer periods than in the past. Although this may not necessarily increase the risk of an N-1 event, it may increase the potential impact of such events. It may also increase the probability and potential impact of multiple dependent contingencies, particularly as they are likely to occur within the timeframes usually given to return the transmission system to a secure operational state. In view of these concerns, questions are being raised about the relevance of the current standards and the extent to which they provide an appropriate basis for transmission system security in competitive electricity markets spanning larger integrated networks. Several suggestions have been advanced to improve the effectiveness of transmission system security standards in the wake of the 2003 blackouts. An immediate response was to pursue ways to strengthen the legal basis for enforcement. Proposals to strengthen the legal framework have been discussed previously. Such proposals aim to extend coverage to all relevant parties and to make standards legally binding. Mandatory application of transmission system security standards would greatly strengthen compliance with existing standards. However, improved compliance with existing reliability standards would not address the fundamental effectiveness of existing standards in reformed electricity markets. Another option would be to extend the existing standard to increase the absolute level of security delivered. This could be achieved, for example, by moving from an N-1 standard to an N-2 standard. Although an N-2 standard would generally increase the level of transmission system security within an integrated transmission system, it may do so at great cost. Figure 27 presents a very simple example to illustrate this point. Deterministic application of a more stringent security criterion to all transmission lines within an integrated transmission system or control area, regardless of the risk of failure or potential consequences for transmission system security, would represent a poorly targeted response to strengthen system security with the potential to substantially increase total costs. Deterministic approaches are also by nature static and do not readily permit real-time incorporation of dynamic factors affecting transmission system security, such as angle stability, voltage and frequency conditions68.

68. Bialek, J. (2004).

128

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Figure 27

Cost Comparison of N-1 and N-2 Security: Simple Example

Assumptions ●

In this example, two regions (A and B) form an integrated electricity market and are interconnected by three transmission lines with a maximum rated capacity of 300 MW each.

●

A load of 1 200 MW is connected in Region B.

●

The system marginal price of electricity in Region A is USD 20/MWh, while in Region B it is USD 50/MWh.

N-1 Secure Operation Application of an N-1 security criterion in this example would mean guaranteeing power transfers between Region A and Region B if one of the 300 MW transmission lines was lost. As a result, the maximum power transfer on each of the three transmission lines would be 200 MW. A 600 MW

200 MW 200 MW

B 600 MW

200 MW 1200 MW

N-2 Secure Operation

Application of an N-2 security criterion would double this requirement, leading to a maximum total transfer capability on each line of 100 MW. A 300 MW

100 MW 100 MW

B 900 MW

100 MW Cost Comparison ●

1200 MW

The hourly cost of adopting an N-2 security standard would be the difference between the system marginal price in each region multiplied by the reduction in transfer capacity. In this example that would be USD 9,000 (ie. USD 30 × 300 MW), representing an increase in electricity costs of around 20%.

Source: Krischen, D. and Strbac, G. (2004).

129

LEARNING FROM THE BLACKOUTS

An alternative, more radical option could involve modifying the way the N-1 criterion is interpreted and applied, to provide a more flexible and adaptable approach for managing transmission system security that better reflects more dynamic and uncertain real-time system operating conditions. Current deterministic approaches could be enhanced with probabilistic methodologies. Probabilistic techniques could be used to identify and assess factors that may affect transmission system security and to calculate the risk of failure and impact of failure using indices such as probability, frequency, duration and severity. For instance, probability risk assessment (PRA) has been successfully applied in several industries with complex engineering systems that are exposed to low probability, high consequence failures, such as the nuclear power sector. PRA involves calculating a measure of the probability of undesired events on the transmission system and a measure of the severity or impact of those events. Considerable progress has been made toward the development of PRA tools to support contingency planning and real-time system operation. For example, the Electric Power Research Institute (EPRI) as part of its Power Delivery Reliability Initiative has developed several PRA related methods and tools for the electricity industry69. Probabilistic methods may facilitate the development of more effective risk management strategies, including more pro-active targeting of contingency resources to meet transmission system security standards at least cost. According to EPRI, PRA offers several potential advantages for real-time system operation including: ●

helping to identify component failures that would most jeopardise transmission system reliability and components most affected by the failure of other system components;

●

providing a more effective means of simulating and evaluating transmission system behaviour under various conditions to identify weak points and bottlenecks;

●

supporting the development of efficient response strategies to address transmission system security vulnerabilities; and

●

enabling the effective operational reliability of the transmission system to be measured objectively at any point in time.

Ideally PRA should consider every possible scenario. However, in large integrated transmission systems this would require a prohibitively large number of 69. The following discussion on PRA is drawn from several EPRI publications including EPRI (2001) and EPRI (2003c).

130

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

calculations. EPRI’s solution was to develop a hybrid methodology that involves undertaking a reliability assessment of a representative set of simulations, resulting in an estimation of reliability. As a result, the EPRI methodology combines probabilistic and deterministic features to develop a practical approach. Achievements to date have included the development of a series of integrated software packages. Key among these is the Physical Operation Margin (POM) program and the Probabilistic Reliability Index (PRI) program. The POM program can simulate a large number of contingencies quickly and determine their likely impacts, including identifying those most critical to system security. Its outputs are a list of potential contingencies including voltage violations, thermal overloads, voltage instability and loss of load. POM results are then fed into the PRI program, which computes probabilities for various contingencies. It then multiplies the probability of critical contingencies by the potential impacts to determine reliability indices for each potential event. Box 2 provides further information on EPRI’s PRA methodology.

Box 2 . Overview of EPRI’s Probability Risk Assessment (PRA) Methodology EPRI’s PRA methodology consists of a range of software packages that provide probabilistic results that allow system operators to incorporate the probability of a contingency event into their operational contingency analysis and preparation. Key software components include the Physical Operation Margin (POM) program and the Probabilistic Reliability Index (PRI) program. Physical Operation Margin Program The POM software program essentially simulates power transfers while monitoring technical constraints. The software can simultaneously monitor voltage stability, voltage constraints and thermal constraints while calculating voltage stability margins. POM provides flexibility in monitoring these constraints. For example, thermal constraints can be monitored based on total power flow (MVA) or current loading (MVA/pu). Voltage constraints may be monitored at load buses, or different voltage limits may be monitored at different buses. The program automatically generates N-1 and N-2 contingency lists and performs simultaneous contingency and transfer studies by identifying critical contingencies for each transfer level and by determining the location type

131

LEARNING FROM THE BLACKOUTS

and severity of potential system security violations. The program is available in two versions to support contingency planning and real-time operation. Probabilistic Reliability Index Program Output from POM and other probabilistic data become inputs for the PRI program. The PRI program calculates the likelihood of a simulated outage and its physical impact or severity. Each contingency is deduced from a unique combination of the availability and unavailability of pieces of equipment, with each unique situation assigned an occurrence probability. For every simulated situation (i), the PRI calculates the sum of its impact (Ii) and its occurrence probability (pi) according to the formula: PRI = ΣpiIi In this equation, “probability” represents the likelihood of experiencing the impact (the probability of the initiating events – contingencies that could lead to violations of operating security limits – that cause the impact). “Impact” represents the calculated severity, which includes deviations from: ●

acceptable thermal loading of transmission lines and transformers;

●

acceptable bus voltage levels;

●

voltage stability; and

●

dynamic stability (in some systems).

The impact may be measured by the distinct number of buses or lines that experience voltage violations or overloads. The results can also be presented as an aggregate reliability index which is the sum of the product of the occurrence probability multiplied by its severity. For example, a voltage index could be calculated as the sum of the product of the probability and amount of the voltage deviation below a certain level, summed over all outage situations and summed over all buses with violations. Similar reliability indices can be defined using this methodology for thermal overloads, voltage stability and dynamic stability. However, it is generally not possible to evaluate all possible contingencies in real-time using this methodology due to the huge number of potential contingency events associated with large integrated transmission systems. In practice, a representative set of simulations would be analysed to provide an approximation of system reliability. Source: EPRI.

132

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

EPRI’s methodology has undergone extensive testing and yielded some promising results, particularly in the context of improving contingency planning. Box 3 provides a summary of recent experience from utilities involved with testing the methodology.

Box 3 . Applying PRA Methodologies: Three Case Studies Consolidated Edison Company of New York Consolidated Edison Company of New York (Con Edison) is a regulated utility serving 3.1 million customers over a 660 square mile area that includes most of New York City. Con Edison undertakes internal reliability assessments that incorporate probabilistic methodologies. Programs includes: transmission system analysis using contingency analysis programs, physical operating margin and reliability evaluation programs; probabilistic representation of component failures; planning for multiple contingencies; and prioritising outages and maintenance work. Con Edison has developed a network modelling program to evaluate the risk of cascading failures and widespread outages, which the company uses to target its maintenance expenditures. Con Edison has also developed a detailed probabilistic reliability model for planning, and extended this model to provide its system operators with a reliability monitor for outage management. The company’s PRA methodology provides a coherent and quantitative structure enabling its system planners and operators to make accurate reliability decisions. The methodology also enables the company to integrate its maintenance and operations functions. The company is planning to complete a detailed model of its entire bulk transmission system including all substations in time for a comprehensive planning process in 2008. Con Edison intends to continue to use PRA methods to support planning analysis, to analyse how to deploy and augment its assets and to provide reliability monitoring and management tools for its system operators. Entergy Corporation Entergy Corporation is an integrated utility, serving around 2.4 million customers and 22 000 MW of load in Louisiana, Mississippi, Arkansas and Texas. Its 500 kV transmission system includes 15 500 miles of

133

LEARNING FROM THE BLACKOUTS

transmission lines, 1 450 substations and 14 interfaces with neighbouring control areas with over 75 points of interconnection. From a system operating perspective, Entergy faces dynamic and less predictable system conditions, especially loop flows. To address some of its system planning issues, Entergy created a transmission outage database to use in its load flow analysis process. It used EPRI’s POM and PRI software to evaluate potential reliability projects. Entergy found that the POM software provided fast and robust simulations, while the PRI program was easy to learn and use. Entergy plans to continue to use the PRA methodology to improve understanding of its transmission system and to justify and develop future projects to improve transmission system reliability. American Electric Power American Electric Power (AEP) is one of the largest electric utilities in the United States with nearly 5 million customers linked to its 11 state transmission system. It is the largest power producer in the United States. It operates over 42 000 MW of generation in the United States and abroad. AEP is using PRA as part of a 4-phase project to strengthen its system planning and management. Phase 1 is underway and involves developing knowledge and experience in using the PRA tools. Phase 2 will involve application of the tools to transmission system operations. Phase 3 will involve extending application to system planning functions. Phase 4 will involve an evaluation of outcomes, including an assessment of PRA compared to other probabilistic methodologies. Source: EPRI.

Work is progressing to adapt these tools to support real-time system operation. Development of a visualisation tool (the Community Activity Room – see Box 4 for a brief description) is progressing, as is work to combine the visualisation tool with the other PRA software to create a visual online probabilistic risk monitor that would have the potential to provide a visual representation of transmission system security in real time. The development and successful application of real-time probability assessment tools would greatly assist the incorporation of probability methodologies to enhance transmission system security standards. They 134

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

would also have the potential to support greater co-ordination of contingency planning and real-time management of transmission system security, possibly leading to an eventual convergence of these activities. Application of PRA would also be supported by technologies that provide system operators with greater direct control over power flows, such as flexible alternating current technologies (see Box 4 for further description of FACTS). Scope may exist to extend the application of probability risk assessment to incorporate some measure of the cost of potential impacts. Contingency analysis using PRA could be further refined to incorporate an estimate of the value of lost load70. This approach may provide a means to encourage management of transmission system security to be more responsive to the value markets might place on reliable delivery of electricity. Processes to develop and review transmission system security standards need to be transparent, objective and inclusive, given that transmission system security is a shared responsibility and that management of system security has implications for operation of markets and the commercial interests of market participants. Active stakeholder participation can help strengthen the credibility of these processes and ensure that the outcomes are consistent with the level of transmission system security market participants require. For instance, from a user’s perspective, reliability has multiple dimensions including the cost of poor power quality, the cost of mitigation measures, and willingness to pay. These issues can not be effectively addressed without participation of electricity end users71. The need for transparency and greater stakeholder participation is beginning to be recognised. For example, NERC noted in a recent report the efforts made by all reliability regions to develop more open and inclusive memberships and to adopt fair and balanced governance arrangements72. Maintaining effective transparency, objectivity and inclusiveness will remain an ongoing challenge. Lack of technical expertise may represent a barrier to effective participation for some stakeholders, particularly end users. Education and, possibly, funded representation or regulatory representation should be considered to address such deficiencies if they emerge.

70. Value of lost load (VoLL) can be defined as the monetary value consumers place on a marginal unit of electricity consumed. Defining VoLL is problematic given the relative weakness of demand-side participation in electricity markets and the quite different valuations different consumers are likely to place on electricity consumption at the margin. As a result, it is generally used to refer to the highest monetary valuation among all potential consumers for the marginal unit of electricity consumed. Hence, it can be used to define the maximum price consumers would be willing to pay in exchange for not having to reduce electricity consumption at the margin. 71. Burns, R., Potter, S. and Witkind-Davis, V. (2004). 72. NERC (2005b).

135

LEARNING FROM THE BLACKOUTS

Co-ordination, Communication and Information Exchange .......................................................... System operating practices are key determinants of actual transmission system security. They translate the incentives created by the governance regime and security standards into practical activities that provide the means for delivering transmission system security. In general terms, system operating practices in relation to transmission system security have changed little with the introduction of electricity reform. The fundamental basis for system operation remains the local control area. Local system operators remain largely responsible for operational contingency planning and secure system operation within control areas, and can act in an operational context without a great deal of co-ordination or consultation. Co-ordination among system operators within larger integrated transmission systems has been enhanced to a degree to take account of greater interregional electricity flows. Various mechanisms have been developed to improve day-ahead exchange of information on projected trade flows, such as the Day-Ahead Congestion Forecast system in continental Europe, the Open Access Same Time Information System in North America and the Nordic Operational Information System (NOIS) system in Nordel. The NOIS system also provides the platform for the Nordic balancing market. In some cases an institutional framework has been placed over the top of control areas to improve co-ordination and information exchange in relation to reliability and power flows, such as the NERC framework in North America or the System Operation Agreement in the Nordic region. However, co-ordination and communication continues to be handled largely on an exception basis in the context of managing transmission system security, particularly in real-time. The more dynamic operating environment created by unbundling, decentralised decision-making and greater interregional trade needs to be appropriately reflected in operating practices if they are to continue to ensure effective transmission system security in competitive electricity markets. In particular, operating practices need to be flexible and adaptable to permit effective real-time response to system emergencies and disturbances. Contingency planning and emergency responses need to be undertaken from a whole-of-system perspective, reflecting the shared nature of responsibility and action required to maintain transmission system security, particularly in integrated transmission systems spanning multiple control areas. Effective coordination and communication is vital for successful system operation to maintain transmission system security. 136

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

■

Co-ordination Co-ordination in this context has vertical and horizontal dimensions. Vertical dimensions refer to the co-ordination between system operators and other parties in the value chain including generators, distribution network operators and users. Horizontal dimensions largely refer to co-ordination between system operators across integrated transmission systems that span multiple control areas. The greatest co-ordination challenges are likely to occur in integrated transmission systems with multiple control areas. The US-Canadian and Swiss-Italian case studies show that effective management of an emergency event may require a response co-ordinated in real-time, rather than a collection of individual responses. In the Swiss-Italian case study, the Swiss transmission system co-ordinator, ETRANS, could not effectively manage the emergency situation with the resources solely under its control. Support was required from Italy. However, from GRTN’s perspective there was no problem requiring its immediate response as its portion of the integrated network was operating in an N-1 secure condition at the time of the original Mettlen-Lavorgo line trip73. UCTE’s analysis drew attention to the same issue, and went further to suggest that effective operational contingency planning would also be problematic in an integrated transmission system spanning multiple control areas given that no single system operator has the capacity to assess or manage the related systemic risks74. Given the speed with which emergency situations can materialise and spread across integrated transmission systems, it is not surprising that effective emergency responses involving co-ordinated manual interventions in separate control areas are virtually impossible to organise in real time. The decentralised operational model is inadequate and too slow to respond to real-time emergencies spanning multiple control areas75. The need for realtime flexibility to respond in this more dynamic operational environment can magnify the co-ordination challenge. Co-ordination that facilitates integrated real-time operation and emergency responses across multiple control areas would greatly improve contingency preparation and management of system security. Interaction between control areas needs to move beyond day-ahead information exchange and exception based co-ordination closer to real time. Co-ordinated real-time security assessment and control is required. 73. CRE and AEEG (2004). 74. UCTE (2004a). 75. Bialek, J. (2004).

137

LEARNING FROM THE BLACKOUTS

Nordel has an effective multilateral agreement that incorporates protocols for coordinated actions to address aspects of transmission system security where responsibility is shared76. UCTE has recently implemented a multilateral agreement to enforce its operating standards across continental Europe. However, potential may exist to further develop multilateral approaches where system operators within an integrated region jointly prepare contingency plans with agreed protocols for co-ordinated action in the event of an emergency situation. Appropriate operating practices need to be supported by real-time data exchange and communication, to facilitate effective real-time and holistic security assessment and system control. An effective ongoing commitment to real-time data exchange and communication will also be required. However, implementation of more integrated and coordinated system operation is likely to raise organisational, legal, technical and local challenges, particularly across transmission systems spanning several jurisdictions77. Tension associated with the competing objectives of maximising capacity for trade and maximising contingency reserves to ensure transmission system security may need to be addressed in this context. Confidentiality of information exchanges may also need to be addressed. Managing the impact of system disturbances across integrated systems spanning multiple jurisdictions and control areas may also need to be addressed. Co-ordinated emergency responses that lead to load shedding in adjacent or distant control areas or jurisdictions may undermine public support for co-ordinated action and lead to practices that may not be consistent with an effective, holistic emergency response78. Tensions were revealed in the case studies, in North America and Australia in particular, in relation to the use and impact of emergency load shedding. In North America, concerns were raised that load shedding was delayed or not considered because of the potential legal liabilities. In Australia, concerns were raised about the distributional impact, with claims that one jurisdiction bore an undue share of the loadshedding burden79. Details in relation to these and other operational practices need to be carefully considered and developed in consultation with all key stakeholders, especially governments. Greater transparency in the development and implementation of emergency procedures will promote better understanding 76. 77. 78. 79.

Hagman Energy AB (2005). Bialek, J. (2004). See Kirby, B. and Hirst, E. (2002) for further discussion of this issue. NEMMCO (2005a) and (2005b).

138

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

among stakeholders about the potential impacts and can strengthen support for appropriate actions taken to protect system security in an emergency situation. For instance, experience in Denmark and Australia suggests that there is considerable value in agreeing load shedding priorities in advance and making those priorities transparent so that potentially affected parties are aware and can respond80.

■

Communication and Information Exchange Effective communication and information exchange provides an essential platform for improving system operator situational awareness and for developing more effective operational planning and co-ordination of transmission system security, particularly in integrated transmission systems spanning multiple control areas. Public access to accurate and timely information would also permit independent assessment of transmission system security, creating the transparency required to support more effective governance and regulation. It may also support the development of efficient and innovative market-based responses, particularly to emerging transmission congestion. The North American and Swiss-Italian case studies in particular highlight the importance of effective and timely communication and information exchange, with investigations finding that poor communication and information exchange contributed to a lack of situational awareness that undermined effective emergency responses81. Opportunities exist to improve the quality of information available to support operational contingency planning and transmission system security management, particularly during emergencies. Accurate, real-time information is required to support effective contingency preparation and emergency responses. The North American event in particular shows that electricity flows can become very volatile following an N-1 contingency event, highlighting the importance of accurate, real-time information. Information also needs to provide a holistic picture of an integrated transmission system, to support effective co-ordinated management of transmission system security. Greater harmonisation of data standards could improve the quality of 80. Elkraft System (2004a) and NEMMCO (2005a). In the Australian case study, reaction to load shedding in Queensland, which bore the brunt of the disruption, may also have reflected the very negative response of the media and public to disconnections that largely affected residential consumers (around 250 000 of them), whereas in other states the loadshedding protocols agreed to with state governments disconnected large industrial loads as a first priority. This suggests that even where load-shedding priorities are agreed in advance, support may evaporate in the face of public criticism, which highlights the importance of careful examination of any load-shedding priorities to ensure they minimise the impact of disruptions to the greatest extent possible. 81. CRE and AEEG (2004); and US-Canada Power System Outage Task Force (2004a).

139

LEARNING FROM THE BLACKOUTS

information, and promote more effective information exchange between system operators within an integrated transmission system. Technology also offers potential to improve the accuracy and timeliness of data gathering and analysis. This potential is discussed in the investment section below. Efforts to improve information exchange are not without their challenges. For instance, some utilities may resist greater information sharing for commercial or confidentiality reasons. Such issues can be addressed to a large degree through the governance framework by minimising any commercial conflicts of interest resulting from ill-defined regulatory or structural arrangements. The need for more accurate and timely information has been recognised and various efforts are progressing to improve the quality of information and exchange of information between system operators. For example, Nordel is refining its NOIS information exchange platform to improve the accuracy and real-time flow of information between control areas. These enhancements have the potential to improve real-time situational awareness in relation to power flows, bottlenecks, reserves and lines out for maintenance in neighbouring control areas. Such developments will provide more effective information exchange between control areas, helping to improve co-ordination of operational planning and system operation in real-time. In particular, these enhancements have the potential to support more effective management of operational reserves on a holistic Nordic basis82. Efforts are also being made to improve communication between system operators. For example, UCTE noted that ETRANS and GRTN took immediate steps to improve their communication arrangements following the SwissItalian event in the context of a wider process to improve co-operation and coordination83. NERC also noted recent improvements to hardware and procedures to support more effective communication between reliability coordinators in North America. NERC is also upgrading its Reliability Coordinator Information System, which is an on-line, real-time messaging system that provides a key means of sharing emergency alerts between reliability co-ordinators and many control areas84. Other stakeholders including governments, generators, distribution system operators, end users and the community also expect to be kept well informed about events that affect the reliability of electricity services, particularly disruptions, and efforts to restore services following an outage.

82. Nordel (2004b) and (2005a). 83. UCTE (2004a). 84. NERC (2005b).

140

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Experience from the case studies indicates the importance of effective communication with key stakeholders and the community, especially during the restoration period in the wake of a major disruption. In each case, system operators faced tremendous demand for timely relevant information on the nature of the disturbance and efforts to restore services. Various strategies were employed to keep stakeholders informed. For example, in Ontario, the IMO activated its Crisis Management Support Team (CMST) within minutes of the disturbance to co-ordinate emergency management, including providing a central point for information dissemination. Throughout the emergency, CMST communicated with key industry, end user and government stakeholders by conference call and through the IMO’s secure website. CMST held 30 conference calls with key stakeholders during the 8-day restoration period. Information was provided to the general public through media releases, interviews and through regular updates on the IMO website. The IMO supported the government’s emergency communications by providing real-time system status updates and by responding to technical questions85. Elkraft implemented its information preparedness plan in response to the 2003 Swedish-Danish blackout. The plan required Elkraft to inform the authorities as soon as possible and to then ensure that major media outlets with the widest possible public reach were informed about the cause, extent and likely duration of the disturbance. Consistent with the plan, control room operators informed the Copenhagen Police a few minutes after the outage. The police then informed the media. DK Radio News was first to be advised officially of the event given its large regional penetration in Zealand, where the outage had its greatest impact. DK Radio News had a journalist stationed at Elkraft System during the course of the restoration. Other media were then informed with priority given to radio, television and other electronic media. Elkraft System also gave priority to television interviews. A press conference was held the day after the event to explain the cause of the power failure and course of the restoration. Elkraft System stationed an engineer at the Copenhagen Police Headquarters to facilitate communication with government authorities. Elkraft’s communication strategy was well received and succeeded in reducing pressure on civil emergency call centers86. Market messaging systems were also used to keep market participants informed. During the Australian event, NEMMCO employed SMS messaging to improve communication with government officials. 85. IMO (2004). 86. Elkraft System (2003).

141

LEARNING FROM THE BLACKOUTS

Event investigations reinforced the need for effective communication with governments. They also noted the potential to enhance emergency communications; through, for example, advance preparation of emergency communication strategies and improving the resilience of control centers and official websites to power failures.

Investing in Transmission Capacity...................................... One of the more immediate reactions following the spate of large-scale blackouts in 2003 was to suggest that the transmission infrastructure was inadequate and that major new investment in transmission capacity was needed to improve transmission system reliability. Although investment in transmission capacity may help to improve transmission system security to a degree over a period of time, it is not likely to be a critical factor in the context of managing the ‘operational’ reliability of an existing transmission system. Transmission investment issues are of greater importance in the context of discussions about the adequacy of existing capacity to meet growing demand for transmission services. Such issues are critical for the efficient development of electricity trade and for maximising the benefits from electricity reform. These issues are discussed in the IEA publication Lessons from Liberalised Electricity Markets. Transmission system security, however, is not simply a function of available transmission line capacity. Under standard operational practices, system operators are bound to operate any transmission system irrespective of its capacity within its security limits. Hence an N-1 security criterion, for example, can be applied effectively to a transmission system whether it has ‘excess’ transfer capacity or not. In cases where capacity is relatively scarce, system security requirements will take precedence to ensure that sufficient transmission capacity is set aside to accommodate credible N-1 contingencies, and transfer capacity will be reduced. This can have significant implications for efficient inter-regional trade, but is less important from a transmission system security perspective. To the extent that investment in new transmission line capacity creates ‘spare’ capacity, it may help to increase effective transmission system security. Applying an N-2 criterion would have a similar effect. However, both of these approaches could be very expensive and the outcomes may not be certain. Investments in integrated transmission systems should be considered from a holistic perspective. Individual expansions or additions may simply serve to 142

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

move points of congestion elsewhere in the transmission system, delivering no net improvement in transmission system security. Alternatively, ‘spare’ capacity may facilitate more interregional trade, especially where investment alleviates transmission congestion between regions with chronic wholesale price differences. As power flows increase, ‘spare’ transmission capacity will be absorbed until flows reach security limits. At that point, system security would be much the same as it was before the expansion, with the investment providing only a temporary improvement in overall transmission system security. It is even possible that such investment may serve to marginally increase the probability of a system disturbance due to the higher exposure to network instability resulting from longer distance power flows87. However, there are other investments that have the potential to significantly improve transmission system security. Investments in upgrading and improving system operating tools would have the potential to enhance system operators’ capacity to effectively monitor, understand and more flexibly control transmission systems in real time. Investments to strengthen the competence and expertise of system operators and other professionals directly involved in the task of maintaining system security could also serve to improve real-time responses, particularly during emergency situations. Investments that enhance the likelihood of system components operating as designed, especially during emergency situations, may also help to improve transmission system security.

Investing in Technology ....................................................... Transmission system security cannot be achieved without effective system management and operating tools. Effective system operating tools allow system operators to assess and prepare for potential contingencies, provide effective monitoring and situational awareness, and enable timely intervention in real time to manage potential emergency situations or to return power systems to a secure operating state after the loss of key equipment. The potential for developing and deploying technologies to improve operating tools to enhance transmission system security is vast, covering the whole range of activities from operational contingency planning through to security monitoring and network control. Technology has the potential to improve the accuracy, quality and timeliness of information. It can also support the development of more accurate and dynamic system modeling, which in turn can support more flexible and adaptable contingency preparation and promote greater real-time situational awareness. 87. For further discussion of these issues see Kirschen, D. and Stribac, G. (2004); and Bialek, J. (2004).

143

144

Widely used when overhead is not practical, mostly in urban areas and underwater. Research is ongoing to reduce costs. Just entering commercial testing. More experience is needed to lower total life cycle costs.

Underground cables transmit power with very low electromagnetic fields in areas where overhead lines are impractical or unpopular. Costs are 5 to 10 times that of overhead lines, and electrical characteristics limit AC lines to about 25 miles.

New transmission conductors with composite cores, as opposed to steel cores, are both lighter and have greater current-carrying capacity, allowing more power to flow in existing right-of-ways.

New computer-optimised transmission line tower designs allow for more power to flow in existing right-of-ways.

Nearly all AC high voltage power transmission is performed using three phases. The use of six or even twelve phases allows for greater power transfer in a particular right-of-way with reduced electromagnetic fields due to greater phase cancellation.

Underground cables

Advanced composite conductors

More compact transmission line configurations

Six or twelve phase transmission line configurations

Sources: US DOE (2002); EPRI (2003b); EPRI (2003c); Gellings, C. (2004a); Hauer, J. et al. (2002); and IEA (2004b).

Demonstration project underway with cables up to 400 ft. Self-contained current limiters are close to commercial availability.

Superconducting ceramic cables can carry much more current than standard wires of the same size, with extremely low resistance, allowing more power to flow in existing right-of-ways. But the required refrigeration results in higher initial and ongoing costs.

High-temperature superconducting cables

Demonstration lines have been built. Key challenge is cost and complexity of integrating with existing three-phase systems.

Commercially available, with increasing use.

Development status

Description

Technology

Examples of Technologies with the Potential to Enhance Transmission System Security

Table 8

LEARNING FROM THE BLACKOUTS

Converter costs are decreasing making DC an increasingly attractive alternative. Most merchant transmission lines utilise HVDC. Demonstrations are underway for many advanced storage technologies. The economics is problematic in most cases.

Higher voltage lines can carry more power than lower voltage lines. Higher voltages are possible, but require much larger right-of-ways, increase the need for reactive power reserves and generate stronger electromagnetic fields.

Ultra high voltage lines

High-voltage DC (HVDC) HVDC provides an economic and controllable alternative to AC for long-distance power transmission. DC can also be used to link asynchronous systems and for long-distance transmission under ground or water. Conversion costs from AC to DC and then back to AC have limited their usage.

Energy storage devices

Sources: US DOE (2002); EPRI (2003b); EPRI (2003c); Gellings, C. (2004a); Hauer, J. et al. (2002); and IEA (2004b).

Energy storage devices permit use of lower cost, off-peak energy during higher-cost peak-consumption periods. Some specialised energy storage devices can be used to improve power system control. Technologies include pumped hydro, compressed air, superconducting magnetic energy storage (SMES), flywheels and batteries.

Voltage levels of 1 000 kV are currently used in Japan. Electromagnetic fields, right-of-way and technical concerns may limit use in other jurisdictions.

Modular equipment designs provide greater transmission system flexibility, allowing the grid to quickly adapt to changing usage. They could also facilitate emergency deployment from a “strategic reserve” of critical devices, such as transformers.

Modular equipment

Many standards already exist, but further work is needed.

Description

Technology

Development status

Examples of Technologies with the Potential to Enhance Transmission System Security (Continued)

Table 8 3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

145

146

The operation of many power system devices, such as transmission lines, cables and transformers is limited by the device’s thermal characteristics. The high operating voltages of these devices make direct temperature measurement difficult. Lack of direct measurements requires conservative operation, resulting in less power transmission capacity and reduced operational flexibility. Newer dynamic sensors have the potential to increase transmission system capacity and operational flexibility.

In some situations, the capability of the transmission system is limited by region-wide dynamic constraints. Direct system voltage and flow sensors can be used to rapidly measure the system operating conditions, allowing for enhanced system control.

More accurate and dynamic modeling of transmission systems, employing methods such as probabilistic risk assessment, offers the potential for more flexible, real-time operational contingency planning and management of transmission system security.

Enhanced power device monitoring

Direct system state sensors

Enhanced system modeling

Sources: US DOE (2002); EPRI (2003b); EPRI (2003c); Gellings, C. (2004a); Hauer, J. et al. (2002); and IEA (2004b).

Description

Technology

Software programs are being developed that employ probabilistic methods for planning and system management. Planning tools have been developed and tested with several US utilities. Development is continuing.

High-speed power system measurement units (eg phasor measurement units) are commercially available and are being used by several utilities. Research is progressing to examine use of these measurements for real-time control of the power system.

Commercial units are available to measure conductor sag allowing for dynamic transmission line limits. Dynamic transformer and cable measurement units are also commercially available.

Development status

Examples of Technologies with the Potential to Enhance Transmission System Security (Continued)

Table 8

LEARNING FROM THE BLACKOUTS

Various protocols for common information management and data transfer have been developed and implemented. Work is progressing to develop more integrated communications system architectures.

Improved methods of collecting, processing and sharing key data on the real-time status of integrated transmission systems can enhance system modeling and visualisation, enabling more effective planning and management of transmission system security.

Increasingly automated grid management designed to speed up reaction times for failure management.

Enhanced data management

Automatic control systems

Sources: US DOE (2002); EPRI (2003b); EPRI (2003c); Gellings, C. (2004a); Hauer, J. et al. (2002); and IEA (2004b).

On-line visualisation tools are being developed. One is currently being trialed by the Tennessee Valley Authority in the United States.

Improved real-time visualisation of dynamic system operating conditions has the potential to substantially improve operators’ situational awareness and capacity to respond to actual and potential emergency conditions.

Enhanced system visualisation

Such systems are under development. Solutions for secure procedures are a major challenge.

Development status

Description

Technology

Examples of Technologies with the Potential to Enhance Transmission System Security (Continued)

Table 8 3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

147

LEARNING FROM THE BLACKOUTS

Technology also has the potential to improve effective operator control over power flows, which would permit more flexible operation of transmission systems, allowing more effective real-time responses to alleviate congestion, manage emergency situations and enable timely service restoration. It may also support the development of more robust transmission systems that are able to continue to operate effectively even if regions within a larger integrated transmission system become disconnected – the so called ‘self healing’ transmission network. Technology also offers the potential to facilitate real-time co-ordination and more holistic management of transmission system security in transmission systems spanning multiple control areas. Such tools would greatly improve system operators’ capacity to manage operational uncertainty. Many technologies could be deployed to enhance transmission system security. Some examples are provided in Table 8. Technologies can be combined to create integrated technological solutions with the potential to substantially improve transmission system security. Examples include combining technologies to create Wide Area Measurement Systems (WAMS), Flexible Alternating Current Transmission Systems (FACTS), and real-time visual reliability and risk monitors, such as the Community Activity Room (CAR). Box 4 provides an overview of these technologies.

Box 4 . Integrating Technologies to Improve Transmission System Security Wide Area Measurement Systems (WAMS) WAMS technology is based on obtaining high-resolution power system measurements from sensors that are dispersed over wide areas of the grid, and synchronizing the data with timing signals from global positioning system satellites. System operators currently retrieve archived data to analyse grid disturbances and improve system models; in the future, they will be able to use these data in real time to assess the health of the grid. The real-time information available from WAMS may allow operators to detect and mitigate a disturbance before it can spread and enable greater utilisation of the grid by operating it closer to its limits while maintaining reliability. The capacity that is freed up is available to move larger amounts of power over the grid in response to competitive market transactions.

148

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Figure 28

UCTE Wide Area Measurement System Time synchr. by GPS

Measuring device 1st synchronous zone 2nd synchronous zone

A prototype WAMS network was installed in the North American Western Interconnection in 1995, while a more extensive pilot WAMS system covering continental Europe was installed by UCTE members in 1998. Both systems have provided useful information in the context of analysing system disturbances. However, considerable potential exists to expand and enhance these systems to achieve their full potential. Flexible Alternating Current Transmission Systems (FACTS) FACTS include a collection of electronic transmission power flow and control technologies that have extremely fast time response capabilities. FACTS devices are based on very high-power solid state electronic switches. They enable operators to exercise fast, precise and continuous active control of power flows on transmission systems to help relieve dynamic problems that may limit network use. FACTS also has the potential to increase the power-carrying capacity of individual transmission lines, providing an alternative to the construction of new transmission capacity.

149

LEARNING FROM THE BLACKOUTS

FACTS provides the potential to react almost instantaneously to alleviate system congestion and manage power disturbances by providing realtime control of power flow on key lines. FACTS can also counteract transient disturbances, allowing transmission lines to be loaded closer to their thermal limits. Most FACTS applications remain very expensive, making FACTS uneconomic for most transmission applications at present. Quick response controllable load and distributed generation has the potential to permit more flexible and effective deployment of FACTS, particularly in the context of managing emergency events. Figure 29

CAR Visualisation

Color Band = 200 MW Above is a CAR representation of a congested power grid where the “light bulb” (the red and white disc) has moved outside the “room”. The position for getting back inside the "room" is given by the green disc.

Real-time System Visualisation Tools The Community Activity Room (CAR) is a real-time dynamic visualisation display developed by EPRI. CAR converts the multidimensional transmission constraints for a transmission system into a threedimensional set of equations that describe the ‘walls’ of a ‘room’ that represents the boundaries of reliable system operation imposed by transmission system security requirements. Beyond these ‘walls’ colour codes indicate probability risk of an outage and the degree of congestion.

150

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Current operating conditions are shown by an illuminated point (or ‘light bulb’). Operators monitor the location of the ‘light bulb’ and are warned when it approaches a ‘wall’. If contingencies or market transactions move the ‘light bulb’ outside the ‘walls’, CAR shows the shortest path to move back inside the ‘room’. CAR is being tested by the Tennessee Valley Authority along with other EPRI software that together have the potential to turn CAR into an online probabilistic reliability monitor. This would allow system operators to understand the tradeoffs involved in operating their transmission systems in real time under a variety of conditions and how best to respond when system overloads or voltage violations occur. Sources: US DOE (2002); UCTE (2004a); EPRI (2003b); EPRI (2003c); Hauer, J. et al. (2002); Burns, R., Potter, S. and Witkind-Davis, V. (2004); and Gellings C (2004b).

Strategic combinations of these technologies may offer additional synergies with the potential to further increase transmission system security. For example, WAMS could be combined with FACTS to provide system operators with more accurate real-time system monitoring capability to guide more effective deployment of flexible system control tools in response to dynamic changes in transmission system security. This would allow operators to more readily realise the potential of FACTS, while at the same time help increase the benefits from implementing WAMS88. More integrated deployment of technologies has the potential to facilitate a transformation of power delivery systems, improving their flexibility and resilience while enabling more innovative solutions to improve reliability, power quality and market efficiency89. Technology, however, should not be viewed as a ‘magic bullet’. Increasing dependence on technologies also has the potential to increase vulnerability to technological failures. Technology failures in the North American case study contributed to a loss of situational awareness that was a direct cause of the blackout, while the spread of the failure reflected dependence on automatic controls. An appropriate balance between automatic and human control needs to be maintained, such that technology supports effective system operation and does not become a substitute for it. Potential barriers to efficient development and deployment of cost-effective technologies may also need to be addressed. For instance, development and 88. Hauer, J. et al. (2002). 89. Recent papers discussing these possibilities include Gellings, C. and Lordan, R. (2004); and EPRI (2003b).

151

LEARNING FROM THE BLACKOUTS

deployment of these technologies may suffer from being a shared responsibility, and a resource with some public good characteristics. An unbundled structure that encourages competition may not necessarily deliver the level of co-operation and co-ordination required through the value chain to achieve a timely and appropriate deployment. Uncertainty over relative roles and responsibilities in an unbundled environment and the potential to ‘free ride’ could exacerbate this challenge. Other uncertainties may also discourage development and deployment. Unbundling leads to greater direct exposure to the related financial risks and costs. Greater competition, more effective regulation and greater exposure to legal liability in this environment may also discourage development and deployment of cost-effective technologies. In some cases it is possible that commercial interests would not be served by certain technological developments, which for example, may reduce network congestion permitting more trade and competition. These interests may act as a disincentive for such investment. Uncertainty over the potential benefits, costs and risks associated with the deployment of new technologies, particularly in relation to how new technologies will interact with existing technologies, may create a further disincentive for investment. Regulation could also act as a barrier to efficient development and deployment of these technologies. For instance, regulatory incentives may generally focus on minimising costs at the expense of other activities, which may discourage investment in promising new technologies. There is also the potential that specific allowances for research and development or new investment may not be sufficient to enable timely deployment of cost-effective technologies. Regulatory or institutional barriers to efficient development and deployment of cost-effective technologies need to be identified and removed to the greatest extent possible. Effective processes will be needed to test and validate new technologies, to help reduce uncertainty and deployment risks. Consideration might also be given to government assistance for promising areas of research and development, where appropriate. Any assistance should be provided in a framework of strong co-operation with industry to ensure that outcomes are efficiently and effectively deployed.

Investing in People ............................................................... Skilled and experienced staff are essential for maintaining secure transmission system operation. Electricity market reform is creating a more dynamic operating environment, which on occasion requires immediate and incisive intervention to ensure continued, reliable and secure system operation. Highly 152

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

trained and experienced personnel are required in this more demanding operating environment where transmission systems are run for longer periods at or close to their security limits. Technology cannot at this stage be expected to replace operator judgment and skill in managing transmission system security, particularly during emergency or near-emergency events. The case studies illustrate the fundamental importance of effective system operation and the potential consequences of real-time operational decisions for transmission system security during emergency situations. In the Swiss-Italian event, for example, nine to ten minutes were used in an attempt to re-close the Mettlen-Lavorgo line even though it was not physically possible to do so given power flows on the system at that time. These minutes could have been spent exploring potentially more effective options. Debate over the appropriateness of the system operator actions following the failure of the Mettlen-Lavorgo line highlights the importance of highly trained and experienced staff capable of quickly recognising developing emergency conditions and effectively intervening to either relieve the conditions, or to return the system to a secure operating state within prescribed timeframes90. The case studies suggest that scope may exist to strengthen operators’ capacity to identify and respond to near and actual emergency events through appropriate training. More attention has focused on emergency training in the wake of the 2003 blackouts. Training programmes are being implemented in North America with greater emphasis on emergency simulations to sharpen these skills. NERC requires staff with responsibility for the real-time operation or reliability monitoring of the bulk transmission system to undertake a minimum of five days of training and drills each year focusing on managing emergency events. Training programs must use realistic simulations. System emergency training is undertaken in addition to other training requirements91. Italian and Swiss system operators implemented a similar program in January 2004. Under this program, system operators train together for a week each year to help improve their ability to communicate and to co-ordinate responses to emergency events in real time. The program aims to facilitate an exchange of knowledge and experience between operators92. Nordel also undertakes joint annual training programmes to improve cooperation and to facilitate better co-ordination by improving operator 90. CRE and AEEG (2004). 91. NERC (2004e). 92. UCTE (2004a).

153

LEARNING FROM THE BLACKOUTS

understanding of the characteristics of transmission systems in neighboring control areas. During these training programs system operators consider, among other things, operational issues relating to their shared management of transmission system security, including frequency regulation, managing power shortages and managing operational disturbances. Training employs case studies and uses desktop studies and computer simulators93. However, the more dynamic operating environment emerging with electricity market reform may suggest the need for a more fundamental review of current training programmes and practices. New training programmes may be needed to supplement experience, which may no longer be as relevant as it once was in the emerging more dynamic, real-time operating environment. Nordel’s Operation Committee recently noted that scope exists to improve its training programmes by adopting a more systematic approach to reviewing actual disturbances and applying simulators to emergency training. Potential exists to extend disturbance management training and to more effectively consider system restoration. A common certification process and training standards for new operators could also be considered to help operators develop a more holistic operational understanding of the integrated Nordic system94. In North America, NERC is also reviewing its training programs with a view to strengthening their effectiveness in a liberalised environment. A new operator training program is being developed and is scheduled for implementation in late 2005. Consideration is also being given to changing the basis for operator certification from a periodic assessment approach to a process of continual education and assessment95. Consideration could be given to reviewing the competencies needed to successfully manage transmission system security in liberalised electricity markets, and to examine whether current training and development programmes are sufficient to equip system operators for the more dynamic real-time operating challenges associated with electricity markets. Elkraft System is undertaking a competencies-mapping project to identify international best practices, new operator training requirements and opportunities for enhancing its operator training programmes.96 Training to ensure that system operators more fully understand the nature and operation of electricity markets should be considered in this context.

93. 94. 95. 96.

Nordel (2004b) and Hagman Energy AB (2005). Nordel (2004b). NERC (2005b). Elkraft System (2004b).

154

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Consideration could also be given to extending training to other relevant professionals involved in supporting secure system operation. These could include personnel managing related information technology and other technical or engineering staff. Training might also be extended to other parties whose actions could affect transmission system security, such as generator plant operators and technical regulators. NERC currently includes reliability co-ordinators in its operator training program. Access to skilled staff could also become more of an issue in the future in some regions where an ageing workforce and progressive retirement of experienced people may create an experience and knowledge deficit, particularly among control room personnel. Such a loss of corporate memory and skill can have negative implications for effective management of transmission system security, particularly for the quality of real-time intervention to manage emergency events. The US Department of Energy’s National Transmission Grid Study notes that an ageing power system engineering workforce combined with an apparent decline in power system engineering graduates may lead to a shortage of qualified system operating staff in the future in North America. It notes concerns about the remuneration for power system engineers and suggests that higher pay in recognition of the more demanding operating environment may be required to attract new graduates97. Regulatory pressures on utilities to cut operating costs have the potential to exacerbate the situation if they are translated into an undue reduction in control room staff and resources for managing transmission system security98. This could have the effect of increasing pressure on remaining operating staff at a time when the challenges associated with managing transmission system security are rising as a result of the more dynamic operating environment created by electricity market reform. Increasing pressure and stress levels may increase the risk of operating errors, particularly during emergency situations. Technical skill and staffing levels need to be carefully monitored, with appropriate action taken to identify and address any emerging deficits.

Asset Performance and Vegetation Management .............. Transmission system security is fundamentally dependent on predictable and reliable asset performance. Relatively small equipment failures can put 97. US DOE (2002). 98. Hauer, J., et al. (2002) notes that regulatory cost-cutting pressures may have already led to a reduction in staffing levels and resources devoted to system planning.

155

LEARNING FROM THE BLACKOUTS

transmission system security at risk. Effective asset management is required to deliver predictable and reliable asset performance. Key dimensions of asset management in this context include protection and other operational settings that affect equipment performance, equipment maintenance practices and vegetation management practices. Each of these can exert a considerable influence on asset performance and transmission system security.

■

Management and Maintenance of Capital Equipment The case studies highlight several examples of equipment that failed to operate as expected during emergency conditions and operated in a manner that helped to exacerbate the impact of the disturbance. The US-Canada Power System Outage Task Force noted that unduly conservative protection settings on some generating equipment and transmission lines may have led to premature tripping of key generation and network assets, which accelerated the spread of the blackout and magnified its impact. Similar issues emerged during the Swiss-Italian event in relation to some of the initial generator trips. The US-Canada Power System Outage Task Force also noted the failure of information technology equipment to function, or to function as expected, which served to degrade situational awareness among system operators and reliability co-ordinators, undermining a timely and effective emergency response. In the Australian event, failure of protection devices to operate according to specification at the Bayswater and Eraring power stations tripped those generators and lead to substantial under-frequency load shedding99. In the Swiss-Italian and in the Swedish-Danish case studies, the failure of some generation to successfully migrate to in-house operation combined with the failure of some black-start generators to operate effectively under emergency conditions limited system operators’ capacity to contain the impact of the event and slowed restoration. Unavailability of generation and transmission assets for programmed maintenance can also temporarily reduce transmission system security by decreasing available resources to help manage emergencies while also placing a transmission system closer to its security limits. Maintenance issues can be expected to become more significant as loading on transmission systems increases. Several transmission lines and generators were unavailable due to 99. NEMMCO (2005a).

156

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

programmed maintenance during the Swedish-Danish event, reducing the resources available to help manage the emergency. Effective maintenance programs are needed to meet these challenges. Scope may exist to apply more innovative maintenance practices. Transmission owners need to implement maintenance strategies that provide an appropriate balance between predictive and corrective maintenance. Predictive approaches using condition-based maintenance, for example, could be more fully utilised to target and optimise maintenance efforts, which could help to reduce maintenance down time while also minimising the risk of component failure at least cost100. Maintenance related risks to transmission system security could also be reduced by employing more integrated and coordinated maintenance cycles across integrated transmission systems. Nordel has proposed developing common procedures for co-ordinating planned transmission network outages101. Effective information exchange between transmission owners and operators will be required to support more integrated planned maintenance activities. The case studies highlight the need for effective verification and enforcement to ensure that equipment is appropriately maintained and operates predictably, particularly during emergency situations. The degree of independence associated with verification and enforcement may also need to be considered in this context given the growing trend toward outsourcing of maintenance activities. Maintenance requirements and performance standards may need to be clarified. Effective certification programs will help to strengthen verification and enforcement. Failure of contracted black-start generation in Denmark has led Elkraft System to implement new service contracts that incorporate routine and surprise inspections to test the preparedness of equipment contracted to provide black-start and other ancillary services. Incentive payments are linked to this verification program. Svenska Kraftnat is also reviewing its methodology and resource requirements to effectively manage outsourced maintenance, and is pursuing more effective enforcement of mandatory operational requirements for generators. NEMMCO has suggested that the legal framework underpinning system operation may need to be strengthened to provide it with authority to

100. Condition-based maintenance involves performing maintenance based on an assessment of the actual condition of components. It allows transmission owners to more closely align maintenance activities with observed wear and tear on equipment and probability of failure, to create more equipment-specific maintenance schedules designed to facilitate ‘just in time’ replacement. See Ray, C. (2004) and CIGRE (2005) for further discussion of innovative maintenance practices. 101. Nordel (2005a). 102. Elkraft System (2004b); Svenska Kraftnat (2004); and NEMMCO (2005a).

157

LEARNING FROM THE BLACKOUTS

demand similar assurances from National Electricity Market generators that their plant can withstand credible contingencies102. Protection settings may also need to be reviewed to ensure that they provide an appropriate balance between equipment protection and reliable system operation. In particular, protection settings should not unduly contribute to the risk of a cascading blackout following an N-1 event. An initial review of Zone 3 transmission protection systems undertaken by NERC revealed that around 70% of those systems that did not meet operational requirements could be fixed by changing protection settings. NERC is continuing to examine the potential for wider and more effective use of protection systems to support reliable system operation103. Regulators should also be mindful of the potential impact of regulatory decisions on incentives for efficient asset management. For instance, there is potential for incentive regulation to encourage cost-cutting at the expense of effective maintenance. Regulatory outcomes need to allow adequate costrecovery for prudent maintenance. These risks could be addressed though specific allowances for maintenance activities, possibly linked to some form of reliability performance incentive. Another regulatory dimension may relate to decisions regarding mergers and acquisitions. The US-Canada Power System Outage Task Force recommended that regulators take into account the reliability implications of these decisions104.

■

Vegetation Management Vegetation management is arguably the most critical component of any transmission asset management program. It is generally accepted that a flashover or contact with vegetation represents one of the most likely causes of a transmission line failure. Hence, effective vegetation management is vital for maintaining transmission system security. Vegetation management is the responsibility of transmission owners. Typically the rules governing vegetation management are more objective than prescriptive, leaving considerable scope for interpretation by transmission owners. Specific standards have generally not been prescribed in the past, in recognition of the need for flexibility to allow local transmission owners to adopt practices that are aligned to the nature and growth characteristics of local vegetation. Although some jurisdictions have adopted mandatory vegetation management standards, most are framed as guidelines and applied on a voluntary basis. 103. NERC (2005b). 104. United States-Canada Power System Outage Task Force (2004a).

158

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Specific practices adopted to manage vegetation also vary considerably between transmission owners, reflecting the unique circumstances pertaining to each transmission line easement. However, they tend to share certain common features. Practices are typically built around a vegetation management cycle that involves the progressive assessment and treatment of all easements over a period of years. Recent reports found that vegetation management cycles in North America can take from one year to over ten years to complete, with cycles of around five years relatively common practice105. Inspection and assessment of transmission easements is generally undertaken using aerial surveys and ground patrols. Treatment can involve tree pruning or removal, and vegetation control in the immediate vicinity of facilities. Manual, mechanical and chemical methods are used to control vegetation along easements. Sometimes ground covers are planted in easements to help control the growth of other less desirable forms of vegetation. Transplanting is also undertaken on occasion. Inspections are commonly undertaken following treatment to assess the treatment’s effectiveness. Vegetation management activities are sometimes supported by management systems such as vegetation inventories. Other activities can involve public education, and research and development. Vegetation management represents one of the largest recurring expenses for most transmission owners. Ideally, vegetation management programs should be guided by the nature and rate of growth of the vegetation in and around each easement. However, the high cost of vegetation management combined with a common lack of rigorous analysis to support work programmes can expose vegetation management to expedient commercial decisions, particularly where transmission owners are under pressure to cut costs. As a result, work programmes can be curtailed or delayed for financial reasons. Official investigations have raised concerns about the effectiveness of vegetation management standards and practices in the wake of the 2003 blackouts. Contact or flashover with vegetation was identified as a primary cause of the North American and Swiss-Italian blackouts featured in the case studies. In the North American event, three key 345 kV transmission lines failed due to contact with vegetation. Each of these lines failed before reaching their emergency maximum thermal ratings, with one failing with power flows at around 44% of its maximum rating. The US-Canada Power System Outage Task Force concluded that these failures were caused by contacts or flashovers with trees under the transmission lines which had been allowed to grow too 105. Cieslewicz, S. and Novembri, R. (2004); and United States -Canada Power System Outage Taskforce (2004a).

159

LEARNING FROM THE BLACKOUTS

tall rather than excessive conductor sag resulting from the overheating associated with large power flows. The Task Force recommended the establishment of enforceable standards for clearances in easements106. UCTE’s investigation listed contact with trees as a primary cause of the SwissItalian outage but did not examine vegetation management practices, which were beyond the scope of its review. Vegetation management within the UCTE area is subject to national regulation. An investigation by the Swiss Federal Inspectorate for Heavy Current Installations of the transmission lines that tripped due to contact or flashover with vegetation during the Swiss-Italian event concluded that vertical clearances on these transmission lines had been maintained in accordance with the relevant standards and regulations. However, the Swiss Federal Office of Energy’s investigation recommended that current standards be reviewed, particularly in relation to calculating conductor temperatures and the degree of sag associated with power flows107. Studies undertaken in the context of the North American blackout investigations have identified several best practices principles for vegetation management of transmission easements including108: ●

Management Framework. Transmission owners should establish a formal vegetation management plan incorporating practices, objectives and approved procedures. The plan should include a documented work program and appropriate measures to ensure effective outcomes, such as a quality assurance programme. It should be revised periodically.

●

Workload budgeting and scheduling should be based on accurate and current information about the vegetation under and adjacent to transmission lines. Appropriate information systems should be developed to support efficient and comprehensive vegetation management. Funding should be based on actual work requirements with flexibility to manage unanticipated events like removing a dead tree.

●

Operational Practices. Field inspections of vegetation conditions should occur on a frequent basis, and the schedule should be based on anticipated growth. Aerial patrols should be complemented by regular ground patrols.

●

Transmission easements should be maintained in accordance with ‘wire zone-border zone’ and integrated vegetation management strategies.

106. United States -Canada Power System Outage Task Force (2004a). 107. SFOE (2003). 108. Cieslewicz, S. and Novembri, R. (2004); and FERC (2004).

160

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

These practices employ environmentally sound control methods to promote desirable, stable, low-growing plant communities that will resist invasion by tall growing tree species. Figure 30 provides an overview of the ‘wire zone-border zone’ model. ●

Maintenance of easements should take into consideration issues including line sag and sway, expected weather conditions and the anticipated pruning response of specific trees.

●

Qualifications and Education. Transmission vegetation management programmes should be managed and implemented by appropriately qualified personnel. Internal training programmes should provide ongoing training for personnel directly involved with vegetation management. They should also aim to increase awareness and appreciation of vegetation management issues throughout a transmission owners organisation to promote support for these activities, particularly among senior management.

●

Utilities should have a comprehensive public education programme that provides the public, individual landowners and other agencies and groups with accurate information regarding vegetation management activities and practices, to help increase support for effective vegetation management. Figure 30

The Wire Zone – Border Zone Model

The wire zone-border zone model involves creating a predictable, lowgrowing vegetation environment immediately under and adjacent to transmission lines to prevent vegetation from growing into or falling on those lines. The concept is illustrated in the following diagram.

Border

Wire Zone

Border

It has proven effective in reducing, and in some cases eliminating, vegetation-related outages on transmission easements. It can also generate other benefits including reduced long-term maintenance costs, improved habitat for wildlife, greater biodiversity and improved fire mitigation. Sources: Cieslewicz, S. and Novembri, R. (2004); and FERC (2004). Diagram from Yahner, Bramble & Byrnes (2000).

161

LEARNING FROM THE BLACKOUTS

●

Research and Development. Transmission owners should undertake ongoing research and development to evaluate current and potential tools and practices.

Improved regulatory arrangements and more effective vegetation management practices offer the potential to significantly reduce the risk of vegetation-related transmission line outages. Several investigations have found that scope exists to create clearer, more consistent and enforceable vegetation management standards. NERC is in the process of developing a more comprehensive vegetation management standard109. The current draft standard proposes minimum clearances and incorporates several of the best practice principles noted above. It appears that considerable discretion will remain for transmission owners to define the precise measures under the standard; however the framework for verification and compliance will be prescribed. The considerable variation that can occur in vegetation and climatic conditions precludes absolute prescription in this context. A balance will need to be struck between flexibility to accommodate local variations and certainty to permit effective verification and enforcement. Mandatory standards alone are unlikely to eliminate vegetation-related transmission outages and, depending on their nature, may prove costly and difficult to enforce110. Standards need to be seen as part of the solution to improving vegetation management. Other regulatory arrangements can also influence the effectiveness of vegetation management. FERC has noted that potential exists to improve regulatory co-ordination to remove potential duplication, resolve conflicting requirements and streamline regulatory approval processes. Vegetation management is an expensive exercise and regulators need to ensure that regulatory arrangements permit recovery of prudent expenditures. They also need to ensure that general incentives encouraging cost cutting are not translated into inappropriate cuts to vegetation management budgets111. Land owner and other local objections to effective, environmentally sound vegetation management may also need to be addressed. Education and effective communication combined with appropriate regulatory appeals processes could help to reduce resistance to appropriate vegetation management practices. Adoption of improved vegetation management practices, including the best practice principles outlined above and shorter vegetation management cycles, would also help to increase the effectiveness of vegetation management programs. 109. NERC (2005a) and NERC (2005b). 110. Cieslewicz, S. and Novembri, R. (2004). 111. FERC (2004).

162

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Applying Market-based Approaches to Support Transmission System Security.............................................. Electricity market reform is helping to strengthen transmission system security. Greater inter-regional trade resulting from reform has encouraged more efficient integration of transmission systems and permitted more effective reserve sharing arrangements within systems spanning multiple control areas. More effective integration has also expanded the potential pool of frequency control ancillary service providers in some regions, effectively increasing the resources available to help manage disturbances and emergency events from a frequency control perspective. A key benefit has been more cost-effective delivery of these services. Some markets incorporate dynamic pricing of transmission congestion, such as locational marginal pricing, which is employed in several markets, especially in the Northeast of the United States. Dynamic congestion pricing provides greater transparency in relation to the operating state of a transmission system by time and location. It can create price incentives that are more closely aligned with the real-time operational conditions of a transmission system, encouraging market participants to respond in ways that support more efficient and costeffective management of transmission system security. It can also help system operators to more effectively target system security resources to those transmission paths which are of greatest value to market participants. Opportunities exist to strengthen transmission system security by employing more effective market-based mechanisms to improve transparency, strengthen cost-reflectivity and promote competitive service provision. Such mechanisms encourage more efficient, innovative and better-targeted provision of transmission system security at least cost. More transparent valuation of reliability and system security services may also provide a means of identifying and funding efficient new reliability-based investments, such as investments to provide reactive power. Market-based approaches also encourage more flexible and responsive use of the transmission system, which has the potential to complement system operator management of transmission system security by reducing pressures on transmission resources at times when systems are congested and operating at or near their security limits. Application of market mechanisms in these ways could enable more effective integration of network services with competitive markets, leading to more efficient and flexible provision of transmission system security reflecting 163

LEARNING FROM THE BLACKOUTS

market requirements. Markets may also provide a means of supporting more effective co-ordination of system security-related activities within the value chain and across transmission systems spanning multiple control areas, consistent with the shared nature of responsibility for transmission system security in unbundled electricity markets. Scope may exist to expand the role of market-based approaches to complement and, possibly in time, to replace regulatory methods for addressing transmission system security in some cases.

■

Markets for Ancillary Services Ancillary services are required to ensure secure operation of transmission systems. They include a range of products to control frequency, voltage and power flows to ensure reliable delivery of electricity services. They also incorporate services to facilitate system restart after an outage, so called black-start ancillary services. The potential benefits from employing market-based approaches to procure ancillary services have been recognised. At present, market-based methods typically involve open tendering and bilateral contracting processes. Some employ more dynamic market mechanisms that incorporate regular auction processes. Prices are set on the basis of competitive bids, with the marginal bid setting the system price for a particular trading period. In other cases, auction processes use cost-based regulated prices where competitive pressures are considered too weak to produce efficient system marginal prices. Cost recovery is typically averaged across users through some form of general charge or as part of the fixed component of transmission charges. In the United Kingdom, ancillary services are provided through a combination of mandatory and commercial services. Commercial ancillary services are typically procured using market-based, open tender processes. National Grid Transco (the system operator) purchases a wide range of ancillary services through commercial processes including frequency control, network control and black-start services112. In some cases mandatory services have been replaced with commercially procured services by agreement between the contracting parties. Commercial arrangements have led to a substantial reduction in the overall cost of ancillary services for maintaining transmission system security. The PJM Interconnection in the northeastern United States employs an auctionbased approach to purchase regulating power and spinning reserve services113. 112. Ancillary services that National Grid Transco currently procures through open tender processes include: maximum generation service, enhanced reactive services, commercial frequency response, fast reserve; standing reserve, warming; and emergency assistance. Further details are provided on the National Grid Transco website (http://www. nationalgrid.com/uk/). 113. PJM (2005).

164

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Both the regulating and spinning reserve markets are cleared on a real-time basis and co-optimised with the energy market to minimise the total cost. The markets are cleared simultaneously so that a generating unit can only be selected to provide either regulating power or spinning reserve but not both. PJM operates up to three regulating markets. Demand for regulating services is determined on the basis of technical requirements. Supply bids are accepted daily for each hourly trading period and can be modified on an hourly basis. In a market characterised by strong competition, the hourly price would be determined by the bid price of the marginal supplier, which would form the system marginal price received by all dispatched generators for a given trading period. However, the market is not considered sufficiently competitive to ensure efficient pricing outcomes at this stage. Hence, the market is currently cleared on the basis of administered prices that reflect the unit specific incremental cost plus a fixed opportunity cost margin per MWh supplied. These costs are calculated by PJM. PJM introduced spinning reserve markets in December 2002, and operates up to three of these markets. Each market has two operational levels. Tier 1 resources consist of generating units that are online following economic dispatch and are able to increase output at short notice. All units within PJM are considered potential Tier 1 suppliers unless they are assigned to Tier 2, which includes units that are operated specifically to provide spinning reserve. Tier 1 units are paid when activated, while Tier 2 units are paid on the basis of their availability, whether or not they are used. An auction mechanism with an administered pricing regime similar to the one used for the regulating power market is employed to clear the spinning reserve market, as it is not yet considered sufficiently competitive for market-based bidding to set the price. PJM’s Market Monitoring Unit undertakes market surveillance and advises PJM when administered pricing is required. In Scandinavia, Nordel has operated a common real-time balancing market across the synchronous part of the Nordic region since September 2002. Nordel is investigating the potential for extending this market to include manually activated operating reserves, possibly from 2006114. Since November 2000, Statnett, the Norwegian TSO, has also operated an options market for fast-acting operational reserves. Under this arrangement, Statnett purchases an option to dispatch regulating resources from generators and from large consumers. Options are active for a trading period equal to one week. Bids of at least 25 MW can be made for periods of up to 8 weeks in 114. Nordel (2005b).

165

LEARNING FROM THE BLACKOUTS

advance of a trading period and modified until the weekly gate closure (at noon on the Thursday before each trading week). Offers to reduce consumption by at least 25 MW must identify any restrictions regarding the period of activation and minimum interval between activations. Contracted reserves have to be offered to the balancing market between 6.00 a.m. and 10.00 p.m. each weekday during the trading period. Options that are activated must respond within 15 minutes. Dispatch is determined by the bid-based merit order, subject to network capability, with the system marginal price set by the bid (offer) of the marginal generator (consumer) dispatched. All dispatched entities are paid the system marginal price. The reserves options market has proven effective and resulted in a substantial volume of demand-side participation from large industrial consumers. Statnett is investigating options to modify these arrangements to encourage participation from smaller consumers115. Further opportunities could be explored to extend market-based approaches for purchasing ancillary services. For example, market-based procurement could be expanded to include a wider range of ancillary services, such as more effective coverage of network control ancillary services like reactive power. Potential may also exist to use more dynamic market-based models, like the auction models noted above, which have the potential to substantially improve the flexibility and efficiency of transmission system security by more closely aligning resource procurement with real-time requirements. Another possibility may involve moving toward cost allocation based on the causerpays principle rather then averaging costs across all users. One notable example is the Australian Frequency Control Ancillary Services (FCAS) market, which was launched by the National Electricity Market Management Company (NEMMCO) in September 2001. NEMMCO purchases all its FCAS requirements for the National Electricity Market from this market. Eight products consisting of six contingency and two regulation services116 are traded on the FCAS market. An auction mechanism is employed with the system marginal price determined by the bid of the marginal supplier for each product. Services are dispatched and settled on the same basis as the energyonly market, where bids are dispatched every 5 minutes with 30-minute settlement periods. This facilitates effective co-optimisation of the energy and 115. Walther, B. and Vognild, I. (2005); and Nilssen, G. and Walther, B. (2001). 116. Contingency services are required to ensure that the large frequency deviations caused by relatively infrequent, large contingency events such as the unexpected loss of generating units or disconnection of large load blocks do not cause the frequency to deviate outside defined limits. Regulation services are used to carry out the continuous fine-tuning required to maintain frequency within the normal operating band. This fine-tuning is required to manage small deviations in load consumption and generation output.

166

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

ancillary services markets. Small deviation costs are allocated on a causer-pays basis for each trading period. An overview of the FCAS market is provided in Box 5. Box 5 . Overview of the Australian Frequency Control Ancillary Services Markets The National Electricity Market Management Company (NEMMCO, the market and system operator for the Australian National Electricity Market) operates eight spot markets for frequency control ancillary services (FCAS). These include six markets for contingency services (6-second raise/lower; 60-second raise/lower and 5-minute raise/lower) and two markets for regulation service (raise/lower). Dispatch Registered service providers, including generation and load, can participate in the FCAS market by submitting an appropriate FCAS offer (load) or bid (generation) for that service, via NEMMCO’s market management systems. An FCAS offer or bid submitted for a raise service represents the amount of MWs that a participant can add to the system, in the given time frame, in order to raise the frequency. An FCAS offer or bid submitted for a lower service represents the amount of MWs that a participant can take from the system, in the given time frame, in order to lower the frequency. During each dispatch interval, NEMMCO’s dispatch engine enables a sufficient amount of each of the eight FCAS products, from the FCAS bids and offers submitted, to meet the FCAS MW requirement. The dispatch engine will enable MW FCAS bids/offers in merit order of cost. The highest cost bid/offer to be enabled will set the marginal price for the FCAS category. During periods of high or low demand, it may be necessary for the dispatch engine to move the energy target of a scheduled generator or load in order to minimise the total cost (of energy plus FCAS) to the market. This process is named co-optimisation and is inherent in the dispatch algorithm. Offers and Bids Offers and bids for the FCAS products take the form of the generic FCAS trapezium defined by enablement limits and breakpoints. The trapezium

167

LEARNING FROM THE BLACKOUTS

indicates the maximum amount of FCAS that can be provided (y axis) for a given MW output level for a generator, or given MW consumption level for a scheduled load (x axis). For example, a generator or load dispatched, in the energy market, at “n” MW could be enabled by the dispatch engine to provide up “N” MW of the relevant FCAS. Figure 31

The FCAS Trapezium MW Response

$X<$Y< ………<$Z

Lower Breakpoint

Band 10

Upper Breakpoint

$Z/MW/hr

Upper Enablement Limit

N MW Lower Enablement Limit

Band 2

$Y/MW/hr

Band 1 n MW

$X/MW/hr Unit MW Output Level

The FCAS offers and bids must comply with similar bidding rules that apply to the energy market: ●

Bids or offers can consist of up to ten price-quantity bands with nonzero MW availabilities;

●

Band prices must be monotonically increasing;

●

Band prices must be set by 12.30 p.m. on the day prior to the trading day for which the bid or offer applies; and

●

Band availabilities, enablement limits and breakpoints can be rebid under rules similar to those applying to the energy market.

An ancillary services plant dispatched between an enablement limit and a corresponding breakpoint can be moved in the energy market in order to obtain more FCAS. For example, if a generator was dispatched between the upper enablement limit and the upper breakpoint, the dispatch engine may constrain the unit in the energy market in order to obtain more FCAS, provided this led to the lowest overall cost. The generic trapezium shown above is altered to suit the various technologies that provide FCAS. NEMMCO is the sole purchaser of FCAS on behalf of market participants, and the costs are allocated to market participants on a causer-pays basis, where causers can be measured.

168

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Settlements For each FCAS market, clearing prices are set for each dispatch interval (5 minutes), to apply to all providers within a region. Prices are determined through the dispatch algorithm, and calculated to ensure that providers of both energy and FCAS are indifferent to the actual combination of energy and FCAS scheduled. That is, the prices for each FCAS service include an element of opportunity cost, to ensure providers are indifferent to using their equipment to provide either energy or FCAS. Payments to service providers are based on enablement of FCAS rather than actual use. Settlement is determined on the basis of the normal 30-minute trading period. FCAS costs for each trading period equal the sum of the costs accrued for each 5-minute dispatch interval within the trading period. All payments to FCAS providers are recovered from market participants. As contingency ‘raise’ requirements are set to manage the loss of the largest generator on the system, all payments for these three services are recovered from generators. On the other hand, as contingency lower requirements are set to manage the loss of the largest load/transmission element on the system, all payments for these three services are recovered from customers. Recovery for contingency services is pro-rated over participants based on the energy generation or consumption in the trading interval. The recovery of payments for regulation services is based on the causerpays principal. NEMMCO uses its supervisory control and data acquisition (SCADA) data to determine the contribution of each measured generator and load to frequency deviations for each dispatch interval. Causer-pays factors are calculated for each measured generator and load. Participants whose responses assist in the correction of frequency deviations are assigned a low causer-pays factor while those whose responses exacerbated the frequency deviation are assigned a high factor. Costs are recovered accordingly. All non-measured entities (customers without SCADA) are assigned causer-pays factors based on the remainder (causers not accounted for by measured entities) and pro-rated based on their energy consumption in the trading interval being settled. Sources: NEMMCO (2001); and ACCC (2001).

169

LEARNING FROM THE BLACKOUTS

The National Electricity Code Administrator117 reported that total FCAS costs fell from around AUD 110 million during the first full year of FCAS market operation (i.e. year ending September 2002) to around AUD 27 million during 2003-04, representing a reduction of around 75% in total annual costs over the period. The median weekly cost of FCAS in 2003-04 was around AUD 335 000, which represented around 0.5% of the total value of spot market turnover. The FCAS market is being refined to incorporate more effective regional cost allocation during periods when network congestion or disruptions separate the national market into its component regions.

■

Harnessing Demand Response Demand-side participation has the potential to support more flexible, innovative and efficient delivery of transmission system security at least cost. Demand reductions in response to high prices tend to occur when transmission systems are operating close to their security limits. Effective harnessing of demand response in these situations has the potential to significantly reduce pressure on system security and improve reliability by improving the balance between generation and load. It would provide a more flexible and efficient alternative to mandatory load shedding during emergency situations. Greater demand flexibility may also reduce the volume of operating reserves system operators need to acquire to meet system security requirements. It may also be harnessed as an alternative source of operating reserve, helping to deepen the pool of reserves and increasing competition to provide reserve services. This could have the effect of lowering ancillary services costs while improving overall system security. Greater demand flexibility may also support the development of markets that value reliability and system security. Such markets may open opportunities for the development of more flexible and innovative products and services to support transmission system security. In principal, they would also offer the potential to more closely integrate reliability with energy markets, providing a transparent price signal to help guide system operators toward more flexible and efficient management of system security that reflects consumers’ valuation of reliability. They may also support the development of more efficient and dynamic methods for allocating transmission capacity for trade and reliability. Such markets could help complement existing regulatory measures to manage system security, and may have the potential to replace some of them over time. For instance, potential may exist to largely replace the use of mandatory load shedding with voluntary reductions based on some form of interruptible contract if sufficiently deep and competitive markets can be established. 117. NECA (2004).

170

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

The potential benefits of harnessing demand responsiveness to strengthen system security have been recognised. Typically these efforts have relied on administrative approaches such as demand management. Public appeals to voluntarily reduce consumption have also been used, particularly during emergency situations. More effective market-based approaches are emerging. A variety of interruptible contracts with large industrial users are already used in North America, the United Kingdom, Scandinavia and Australia. Table 9 provides some examples of emergency load response programmes used in North America and the United Kingdom. Probably the most substantial progress in harnessing demand responsiveness for system security to date has been made in the Nordic region and in the United Kingdom where significant amounts of demand response have been secured from large industrial users to provide regulating power reserves and operational reserves. An example from the Nordic market was described in the previous section. In the United Kingdom, National Grid Transco has attracted over 4 000 MW of demand response for frequency control and fast response operating reserves118. However, it is likely that considerable potential remains to be developed. Substantial voluntary demand reductions of up to 4 000 MW were recorded in Ontario during the emergency restoration phase following the 14 August 2003 blackout in response to public appeal119. Nordel has estimated the total potential demand response in the Nordic market at up to 12 000 MW, with every 1 200 MW of realised demand response having the potential to shave up to 2% off peak demand120. There are a range of potential barriers that may hinder the development of demand response for transmission system security. The real-time, instantaneous nature of system security management and the corresponding need for immediate responsiveness may practically preclude wider use of demand response beyond larger users. Technical and commercial arrangements used by system operators to procure operational reserves often tend to favour generation over demand, reflecting lower associated risks and transaction costs from a system operator’s perspective, which may act as barrier to greater use of demand response121.

118. 119. 120. 121.

IEA (2005). IMO (2004). Nordel (2005c). IEA (2003).

171

172

15% reduction from maximum demand, at least 100 kW

Demand Relief Programme

California Independent System Operator

San Diego Gas & Electric Rolling Blackout Reduction Programme

Source: IEA.

New York Independent System Operator

1 MW load reduction

Emergency Load Response Programme

PJM Interconnection (US)

Financial Penalty

None Performance-based capacity payment None

Higher of USD 500 MWh or zonal LMP USD 20 000/MW-month and USD 500/MWh USD 200/MWh

None

USD 6 000 MWh of excess energy

Exemption from rotating outages

Cost-reflective, real-time rate None

Price Incentives

Emergency Demand Response 100 kW reduction per Greater of real-time price zone (aggregated) or USD 500/MWh Programmes

100 kW

Optional Binding Mandatory Curtailment Programme

US state utilities

15% reduction on entire circuit, in 5% increments

Emergency Demand Response N.A. Programme

Independent Electricity Market Operator (Ontario)

Minimum Size

Programme

Company

Examples of Emergency Load Response Programmes

Table 9

LEARNING FROM THE BLACKOUTS

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

Verification of response is a critical issue in this context. Appropriate interval metering with remote telecommunications capability is necessary given the need to be able to precisely verify demand reductions in real time. Large industrial users typically possess appropriate metering; however, other users with potential to respond may not. Despite falling costs, installation of these meters remains relatively expensive and may limit further development of demand response for system security. Efforts are being made to address some of these barriers and to expand the potential pool of demand response available to support system security. For instance, in the United States, the Federal Energy Regulatory Commission has proposed changes to market design rules to encourage greater demand response. In the United Kingdom, the system operator (National Grid Transco) is implementing new procurement guidelines with a view to engaging more demand-side participation122. Nordic TSOs have also proposed a range of action plans to enhance demand response, including improving its potential to contribute to system security123. These plans include: ●

measures to enhance demand bidding for operational reserves and the balancing market;

●

undertaking related research and development projects of common interest for market design and power system planning; and

●

developing communication and information strategies to encourage market players and other stakeholders to become more active demand-side participants, or to take appropriate action to encourage further demand response.

These action plans will focus on encouraging greater demand-side participation among medium-sized industrial and commercial users, and in relation to the larger components of residential use, such as household electric heating. Innovative financial products have the potential to support more effective harnessing of demand side participation. For instance, retail products could be developed that offer consumers differing degrees of service quality. Premiumpriced products could offer minimum service interruption during power shortages or emergencies, with less expensive products involving greater exposure to service disruption during emergencies. Alternatively, products could be differentiated on the basis of compensation paid in the event of a service disruption. 122. IEA (2003). 123. Nordel (2005a); and Nordel (2005c).

173

LEARNING FROM THE BLACKOUTS

Such products would allow customers to choose the level of service for which they are willing to pay. They could also be used to identify consumers that are more willing to experience service interruptions. This could support more efficient targeting of certain emergency interventions, such as load shedding, to protect system security at least overall cost to the community124. Another possibility involves encouraging more active participation from distributed generation. These generators are typically self-dispatching and tend to act like demand reduction from a system operation perspective. They are usually aggregated with demand in electricity market management systems. Nordic TSOs are investigating options to more effectively access and integrate distributed generation to deepen the market for system security services125. Some Challenges and Limitations Although market-based mechanisms have considerable potential to help improve the flexibility and efficiency of transmission system security management, care needs to be exercised in their development and deployment in this context. System security exhibits some characteristics consistent with a public good. For instance, the potential for system collapse may make operating reserves a public good where network users consider transmission system security external to their own activities and are therefore unwilling to pay for it. In these circumstances it is possible that a pure market solution may produce insufficient operating reserves to maintain system security126. Similar problems may emerge in relation to shared security services, such as frequency control, which can be provided at any point in an integrated transmission system. In transmission systems that span multiple control areas, there may be an incentive for individual system operators to ‘free ride’ to a degree on neighbouring system operators, relying on them to manage frequency deviations with their operating reserves. Were several system operators within an integrated transmission network to do this, then there is a risk that insufficient reserves may be procured in aggregate to meet credible system security contingencies127. The real-time nature of system security management may limit the application of market-based mechanisms. For instance, management of small deviations in voltage and frequency require immediate and largely automatic responses that

124. For further discussion of possible products built on the notion of ‘priority insurance’ see Giberson, M. and Kiesling, L. (2004). 125. See Chang, J., Murphy, D. and Graves, F. (2003) for further discussion of these issues. 126. See Joskow, P. and Tirole, J. (2004) for further discussion of this issue. 127. See Kirby, B. and Hirst, E. (2002) for further discussion of this issue.

174

3 POLICIES TO STRENGTHEN TRANSMISSION SYSTEM SECURITY

would naturally preclude a market-based approach relying on price signals to ellicit responses. Models based on tenders or auctions that provide services for an agreed period would be better suited for providing such services. Issues relating to the strength and depth of competition to provide services may also have implications for the deployment of market-based mechanisms. PJM’s experience with market-based procurement of regulating and spinning reserves where the number of potential providers is relatively thin illustrates this point. Similar issues may emerge in the context of procuring services which are localised in nature, like reactive power. Reactive power rapidly diminishes over relatively short distances, which means that it must be provided close to source. As a result, it is likely that the underlying number of potential providers would be relatively small, possibly limited to just one or two suppliers in some cases, which might create potential for market power abuse128. Conversely, mandated provision combined with poor transparency may create a monopsony opportunity for system operators, as single buyers, to purchase these services below their market value. Regulation has the potential to intensify such behaviour where it creates inappropriate incentives for system operators to cut operating costs129. This could discourage efficient investment to provide reactive power in the longer term. Effective market-based models rely fundamentally on the quality and timeliness of information. Information needs to be accurate over time and by location. Most importantly, it needs to be made available to all market participants quickly and transparently. This would be especially important with market models relying on near real-time responses to price signals, such as hourly auction models. Accurate information is also critical for effective verification of actions, which is particularly important for managing system security. It is also critical for effective regulation in this context. Effective realtime metering, information management systems and control equipment are required to meet this challenge. However, installation of these systems can be relatively expensive and coverage may therefore be limited to relatively few market participants. This could represent a practical barrier to deepening and broadening participation in the short term, particularly in the context of demand response as discussed previously. Effective governance and regulatory arrangements that clearly assign authority and accountability would greatly assist the introduction of marketbased mechanisms in this context. Related issues have been discussed in previous sections. 128. See Chang, J., Murphy, D. and Graves, F. (2003) for further discussion of these issues. 129. Giberson, M. and Kiesling, L. (2004).

175

LEARNING FROM THE BLACKOUTS

These and other practical, technical and institutional issues need to be carefully considered in the context of developing and deploying market-based mechanisms and products to help manage transmission system security. The combination of these issues places some practical limitations on the potential to deploy market-based mechanisms and products at this time. Regulatory approaches will continue to be needed to ensure that transmission system security is not jeopardised. However, market-based mechanisms and products offer considerable potential to enhance and complement regulatory arrangements to strengthen transmission system security at least cost.

176

ANNEX 1

ANNEX 1

177

178

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

EPA provisions provide for the establishment of an independent FERC to finalise Electric Reliability Organisation (ERO) to develop and enforce criteria for EROs by reliability rules, subject to FERC regulatory supervision. Application February 2006 across North America to be secured through international agreement between the US, Canada and Mexico. NERC has reformed its arrangements to make them consistent with the expected requirements for an ERO. FERC proposals are contained in its NOPR RM05-30-000 of 1 September 2005.

Strengthen the institutional framework for reliability management in North America.

#3

FERC to finalise mechanism by February 2006

EPA provisions empower FERC to oversee and approve funding arrangements. FERC proposals are contained in its NOPR RM05-30-000 of 1 September 2005.

Develop a regulator-approved fundin mechanism for NERC and the regional reliability councils, to ensure their independence from the parties they oversee.

#2

Implementation Status

Energy Policy Act 2005 (EPA) implements provisions for the FERC to finalise rules development and application of mandatory and enforceable reliability by February 2006 standards. The Federal Energy Regulatory Commission (FERC) issued a draft Notice of Proposed Rulemaking (NOPR Docket No. RM05-30000) on 1 September 2005, which incorporates its proposals for establishing mandatory and enforceable reliability standards.

Key Implementation Actions

Make reliability standards mandatory and enforceable, with penalties for noncompliance.

Description

#1

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status

Annex 1

LEARNING FROM THE BLACKOUTS

EPA provisions empower an ERO to apply and enforce reliability standards with all relevant parties managing or using the bulk transmission system. FERC proposals are contained in its NOPR RM05-30-000 of 1 September 2005.

Require any entity operating as part of the bulk power system to be a member of a regional reliability council if it operates within the council’s footprint.

#7

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

FERC to finalise rules by February 2006

FERC policy statement of 19 April 2004 reaffirmed its commitment Complete to take reliability issues into consideration before authorising new ISOs or RTOs.

Confer FERC approval on new RTOs or ISOs only after they have met minimum functional requirements.

#6

#5

Ongoing

#5

#4

NERC and regional reliability councils continue to track implementation and issue regular progress reports. NERC is developing arrangements for ongoing reliability performance monitoring.

Implementation Status

Track implementation of recommended actions to improve reliability.

Key Implementation Actions

FERC policy statement of 19 April 2004 reaffirmed its commitment Complete to approve applications that recover prudent expenditures.

Description

Clarify that prudent expenditures and investments for bulk system reliability (including investments in new technologies) will be recoverable through transmission rates.

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1 3 DEMAND RESPONSE MARKETS AND ECONOMICS

179

180

Ongoing, EEMC Task Force to report by August 2006

Commission an independent study of the DOE and Natural Resources Canada have initiated this study. EPA relationships among industry provisions require an Electric Energy Market Competition (EEMC) restructuring, competition and reliability. Task Force to report to Congress on competition issues.

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#12

Ongoing

Joint DOE/FERC report due by February 2006

Ongoing

Not active

Implementation Status

Establish requirements for collection and NERC’s Board of Trustees approved proposed standards for reporting of data needed for postdisturbance monitoring equipment and requirements for timeblackout analyses. synchronised fault-recording devices in May 2005.

EPA provisions require the US DOE and FERC to report on establishing a system to provide real-time information on the functional status of all transmission lines within each interconnection.

Establish an independent source of reliability performance information.

#10

#15

FERC has created a reliability division to support assessment of reliability impacts.

Integrate a “reliability impact” consideration into the regulatory decision-making process.

#9

#11

No direct action to date. Could be addressed in context of strengthening institutional frameworks and prescribing reliability standards.

Key Implementation Actions

Shield operators who initiate load shedding pursuant to approved guidelines from liability or retaliation.

Description

#8

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1

LEARNING FROM THE BLACKOUTS

NERC is in the final stages of developing a measurable standard New standard is that will specify the minimum clearances between energised high expected by the end voltage lines and vegetation. EPA provisions will make minimum of 2005 vegetation management standards mandatory and enforceable. NERC Reliability Regions submit annual reports on vegetation related line trips to NERC – first report was provided in March 2005.

Establish enforceable standards for maintenance of electrical clearances in right-of-way areas.

#4

#16

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

FE, MISO and PJM have undertaken remedial actions specified by NERC.

Correct the direct causes of the 14 August 2003 blackout.

#1

#15

Complete, with some related ongoing activities

NERC developed blackout disturbance response procedures, including analysis requirements, which were accepted by the board in May 2005. Work to develop the procedures is proceeding. DOE has commissioned a report and work is progressing.

Establish a standing framework for the conduct of future blackout and disturbance investigations.

#15

Progressing

Implementation Status Ongoing

Key Implementation Actions

Expand DOE’s research programs on DOE is working with NERC through the Consortium of Electric reliability-related tools and technologies. Reliability Technology Solutions on the development of these tools.

Description

#14

#13

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1 ANNEX 1

181

182

#3

#6

#18

#19

All reliability co-ordinators, control areas and transmission operators to provide 5 days of system emergency management training for system operators each year, with initial training completed by 30 June 2004. NERC is revising its operator training programme with new standards and accreditation requirements to be implemented shortly.

Improve near-term and long-term training and certification requirements for operators, reliability co-ordinators and operator support staff.

Ongoing; new training standards and accreditation to be implemented in 2005

Substantial progress has been made toward the goal of completing Complete. reliability audits for all reliability co-ordinators and control areas Implementation within 3 years. The 20 highest priority regions have been completed. ongoing NERC conducted reliability readiness audits of 61 control areas and 6 reliability co-ordinators during 2004. A further 55 entities are scheduled for audit in 2005.

Support and strengthen NERC’s Reliability Readiness Audit Programme.

Largely complete; compliance enforcement programme subject to ongoing development

Implementation Status

NERC has implemented reforms to improve its compliance audit programme including updating compliance templates, disclosure of violators and reorganisation of its compliance committee. New guidelines for disclosure were adopted in June 2004. NERC is developing a methodology for ranking violations to guide the development of sanctions. EPA will help strengthen compliance and enforcement.

Key Implementation Actions

Strengthen the NERC Compliance Enforcement Programme.

Description

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#2

#17

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1

LEARNING FROM THE BLACKOUTS

#8

#10

#21

#22

NERC is working with the Consortium for Electric Reliability Ongoing Technology Solutions on an industry survey of best practices, and on the Eastern Interconnection Phasor Project. This program together with the Western Electricity Co-ordinating Council’s Wide Area Measurement System, will include installation and assessment of high-speed measurement and analysis tools.

Evaluate and adopt better real-time tools for operators and reliability co-ordinators. NERC has established a Real-time Tools Best Practise Task Force focusing on identifying tools to improve situational awareness. Work should be completed by late 2005.

Partially completed; follow-up work progressing

Regional Reliability Councils have completed initial reviews of the need for under-voltage load shedding. Final recommendations are expected in December 2005. NERC’s review of protection and control standards is progressing.

Make more effective and wider use of system protection measures.Initial review of Zone 3 relay setting on 230 kV+ transmission lines completed in September 2004.

Ongoing

Implementation Status

NERC has developed draft definitions. The draft is undergoing refinement. Once completed, it will be incorporated into the Reliability Co-ordinator Information System to improve consistency of communications between reliability co-ordinators.

Key Implementation Actions

Establish clear definitions for normal, alert and emergency operational system conditions. Clarify roles, responsibilities, and authorities of reliability co-ordinators and control areas under each condition.

Description

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#9

#20

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1 ANNEX 1

183

184

#14

#13a #13b #16

#16

#24

#25

#26

Key Implementation Actions

Implementation Status

Tighten communications protocols, especially for communications during alerts and emergencies. Upgrade communication system hardware where appropriate.

NERC has adopted new protocols for hotline calls between reliability Ongoing co-ordinators, and has installed new teleconference facilities. NERC is upgrading its real-time Reliability Co-ordinator Information System.

Complete

Reliability regions are reviewing their models, criteria and procedures Ongoing for model validation and benchmarking. NERC is monitoring regional activities to improve modeling and data exchange. NERC has initiated a project to identify best practices and to propose recommendations for improving current standards and practices.

NERC should reevaluate its existing NERC adopted new reliability standards in February 2005. Under reliability standards development the EPA, FERC will be required to review the reliability standards process and accelerate the adoption of adopted by an ERO. enforceable standards.

Improve quality of system modeling data and data exchange practices.

Strengthen reactive power and voltage NERC has reviewed current practices. New and revised standards Ongoing control practices in all NERC regions. and procedures are being implemented. Regional studies are being undertaken into the feasibility of an under-voltage load shedding program. These studies are expected to be completed before 2006. Several regions have initiated processes to develop regional voltage control/reactive power criteria.

Description

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#7

#23

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1

LEARNING FROM THE BLACKOUTS

#12

#11

#28

#29

Evaluate and disseminate lessons learned NERC regional reliability councils most affected by the blackout Ongoing during system restoration. have prepared restoration reports, which have been circulated to members. Key recommendations from these investigations are being implemented. Ontario’s IMO published its restoration evaluation report in February 2004. All NERC regions are reviewing their blackstart and system restoration plans and procedures in the light of these experiences.

NERC has received standards and current practices. New standards Ongoing are being developed to improve disturbance monitoring capabilities. Criteria for selecting and locating disturbance recording devices are being developed with regional councils. Protocols for exchanging information are also being developed. FERC policy statement of 19 April 2004 reaffirmed its commitment to approve applications that recover prudent expenditures.

Require use of time-synchronised data recorders.

Ongoing

Implementation Status

NERC is developing a transmission facility ratings reliability standard. EPA provision will empower an ERO to enforce such standards.

Key Implementation Actions

Develop enforceable standards for transmission line ratings.

Description

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#13c

#27

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1 ANNEX 1

185

186

#17

#35

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

#17

#34

New permanent standard is expected to be implemented in 2006

NERC implemented a temporary cyber-security standard and procedures in August 2003. New permanent standards and procedures are being developed and are expected to be completed before 2006. Implementation will proceed in 2006. EPA will enable mandatory application and enforcement of standards.

Implement NERC IT standards. Develop and deploy IT management procedures. Develop corporate-level IT security governance and strategies. Implement controls to manage system health, network monitoring and incident management.

#17 #17

#32 #33

Complete

Clarify that the transmission loading NERC has revised its operating policies in relation to TLRs and relief (TLR) process should not be used in incorporated changes into the latest reliability standards. situations involving an actual violation of an operating security limit. Streamline the TLR process.

#31

Implementation Status

NERC has enhanced its RCIS and system data exchange applications to Ongoing improve dissemination of information on forced outages. Work is continuing on criteria for identifying and disseminating information on the operational status of other critical facilities. Joint DOE/FERC study about dissemination of real-time information on the functional status of all transmission lines is also relevant in this context (see Recommendation10).

Key Implementation Actions

Clarify criteria for identification of operationally critical facilities, and improve dissemination of updated information on unplanned outages.

Description

#30

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1

LEARNING FROM THE BLACKOUTS

187

Develop capability to detect wireless and NERC has developed guidelines on cyber-security intrusion detection. Guidelines complete; remote wireline intrusion and The permanent standard will incorporate requirements for development surveillance. ongoing monitoring and controlling access.

Control access to operationally sensitive NERC has developed and implemented the relevant security equipment. NERC should provide guidance guideline. on employee background checks.

#17

#17

#17

#39

#40

#41

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

NERC has developed several security guidelines relating to process Ongoing control systems (PCS) and Supervisory Control and Data Acquisition (SCADA) systems. Work is progressing to promote the development of new technologies to enhance security, safety and reliability of PCS and SCADA systems.

Assess IT risk and vulnerability at scheduled intervals.

#17

#38

Complete

Actions to identify and adopt best practices are an ongoing part of Ongoing NERC operational practice.

Improve IT forensic and diagnostic capabilities.

#17

#37

Risk assessment methodology is complete; implementation is ongoing

The US Department of Homeland Security, DOE, NERC and Public Safety & Emergency Preparedness Canada have jointly developed a security risk assessment methodology, which is being applied to specific infrastructure components. Several reviews have been completed.

Implementation Status

Initiate US-Canada risk management study.

Key Implementation Actions

#17

Description

#36

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1 ANNEX 1

188

Complete

Complete

Sources: US-Canada Power System Outage Task Force (2004a); US-Canada Power System Outage Task Force (2004b); NERC (2005b) and FERC (2005).

Capacity was installed in August 2004.

CNSC to purchase and install backup generation equipment.

#46

NERC has developed a security guideline addressing this issue. The permanent standards will include requirements for data and information to be classified according to confidentiality.

Develop procedures to prevent or mitigate inappropriate disclosure of information.

Ongoing

#17

#44

New permanent standard is expected to be implemented in 2006

Establish clear authority for physical and NERC’s new cyber-security standards will address this issue. New cyber security. permanent standards and procedures are being developed and are expected to be completed before 2006. Implementation will start in 2006.

Canadian Nuclear Safety Commission CNSC is responding through its licensing program. (CNSC) request Ontario Power Generation and Bruce Power to review operating procedures and operator training associated with the use of adjuster rods.

#17

#43

Implementation Status Complete

Key Implementation Actions

Confirm that NERC’s Electricity Sector Information Sharing and Analysis Center is the central point for sharing security information and analysis.

Description

#45

#17

#42

Task Force NERC

Rec. Number

US-Canadian Power System Outage Task Force and NERC Investigation Recommendations and Implementation Status (continued)

Annex 1

LEARNING FROM THE BLACKOUTS

ANNEX 2

ANNEX 2

189

190

Reliability • Enforce operational reliability requirements Authority • Monitor all reliability-related parameters within the Reliability Authority Area, including generation dispatch and transmission maintenance plans • Direct revisions to transmission maintenance plans as required and as permitted by agreements • Request revisions to generation maintenance plans as required and as permitted by agreements • Develop Interconnection Reliability Operating Limits (to protect from instability and cascading outages). • Perform reliability analysis (actual and contingency) for the Reliability Authority Area • Approve or deny bilateral schedules from the reliability perspective • Assist in determining Interconnected Operations Services requirements for balancing generation and load, and transmission reliability (e.g., reactive requirements, location of operating reserves). • Identify, communicate, and direct actions to relieve reliability threats and limit violations in the Reliability Authority Area • Direct implementation of emergency procedures • Direct and coordinate System Restoration

Ensures the real-time operating reliability of the interconnected bulk electric transmission systems within a Reliability Authority Area.

Operating Reliability*

• Transmission Owners • Transmission Planners • Transmission Operators • Transmission Service Providers • Generator Owners • Generator Operators • Load-serving Entities • Distribution Providers • Balancing Authorities • Interchange Authorities • Planning Authorities • External Reliability Authorities • ERO/NERC

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model

Annex 2

LEARNING FROM THE BLACKOUTS

Planning • Develop and maintain transmission and resource (demand and Authority capacity) system models to evaluate transmission system performance and resource adequacy. • Maintain and apply methodologies and tools for the analysis and simulation of the transmission systems in the assessment and development of transmission expansion plans and the analysis and development of resource adequacy plans. • Define and collect or develop information required for planning purposes • Evaluate plans for customer requests for transmission service. • Review and determine TTC values (generally one year and beyond) as appropriate. • Assess, develop, and document resource and transmission expansion plans. • Provide analyses and reports as required on the long-term resource and transmission plans for the Planning Authority Area. • Monitor transmission expansion plan and resource plan implementation. • Co-ordinate projects requiring transmission outages that can impact reliability and firm transactions. • Evaluate the impact of revised transmission and generator in-service dates on resource and transmission adequacy.

Ensures a long-term (generally one year and beyond) plan is available for adequate resources and transmission within a Planning Authority Area. It integrates and assesses the plans from the Transmission Planners and Resource Planners within the Planning Authority Area to ensure those plans meet the reliability standards, and develops and recommends solutions to plans that do not meet those standards.

Planning Reliability*

• Transmission Owners • Transmission Planners • Transmission Operators • Transmission Service Providers • Generator Owners • Generator Operators • Load-serving Entities • Distribution Providers • External Resource Suppliers • Planning Authorities • Resource Planners • External Reliability Authorities

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2 ANNEX 2

191

192

Balancing • Must have control of any of the following combinations within a Balancing Authority Area: Load and Generation (an isolated system); Authority Load and Scheduled Interchange; Generation and Scheduled Interchange; or Generation, Load, and Scheduled Interchange. • Calculate Area Control Error within the Balancing Authority Area. • Review generation commitments, dispatch, and load forecasts. • Formulate an operational plan (generation commitment, outages, etc) for reliability assessment • Approve Interchange Transactions from ramping ability perspective • Implement interchange schedules by entering those schedules into an energy management system • Provide frequency response • Monitor and report control performance and disturbance recovery • Provide balancing and energy accounting (including hourly checkout of Interchange Schedules and Actual Interchange), and administer Inadvertent energy paybacks • Determine needs for Interconnected Operations Services • Deploy Interconnected Operations Services. • Implement emergency procedures

Integrates resource plans ahead of time, and maintains loadinterchange-generation balance within a Balancing Authority Area and supports interconnection frequency in real time.

Balancing*

• Transmission Operators • Transmission Service Providers • Generator Operators • Load-serving Entities • Balancing Authorities • Interchange Authorities • Reliability Authority

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2

LEARNING FROM THE BLACKOUTS

Integrates energy, capacity, • Administer a market that provides capacity, energy, balancing resources, and balancing, and other Ancillary Services subject to system requirements and constraints. transmission resources to • Arrange resources for congestion management. achieve an economic, • Provide dispatch plans. reliability-constrained dispatch of resources. The dispatch may be either cost-based or bid-based.

Market Operations

• Transmission Service Providers • Purchasing-Selling Entities • Balancing Authorities • Reliability Authorities

Market Operator • Market Operator tasks and relationships are specific to a or Resource particular Market Operator and Dispatcher (nonwill depend on the market market structure over which the Market environment) Operator presides. • A Resource Dispatcher performs the same dispatch duties as a Market Operator, but in a nonmarket environment.

• Determine valid, balanced, Interchange Schedules (validation of sources and Interchange sinks, transmission arrangements, interconnected operations services, etc.). Authority • Verify ramping capability of the source and sink Balancing Authority Areas for requested Interchange Schedules • Collect and disseminate Interchange Transaction approvals, changes, and denials • Authorize implementation of Interchange Transactions • Enter Interchange Transaction information into Reliability Assessment Systems (e.g., the Interchange Distribution Calculator in the Eastern Interconnection) • Maintain record of individual Interchange Transactions

Authorizes implementation of valid and balanced Interchange Schedules between Balancing Authority Areas, and ensures Interchange Transactions are properly identified for reliability assessment purposes.

Interchange*

Responsible Related Entities that serve or are served by Entity the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2 ANNEX 2

193

194

• Maintain resource models and apply appropriate tools for the Resource development of adequate resource plans. Planner • Define and collect or develop demand and resource information required for planning purposes. • Provide capacity resource information to planning and operating functions and service functions. • Assist in the evaluation of the deliverability of resources to customers. • Include consideration of generation capacity from resources both within and outside of the Planning Authority Area. • Develop and report, as appropriate, on its resource plans to others for assessment and compliance with reliability standards. • Monitor and report, as appropriate, on its resource plan implementation.

Develops a long-term (generally one year and beyond) plan for the resource adequacy of specific loads (customer demand and energy requirements) within a Planning Authority Area.

Resource Planning

• Transmission Owners • Transmission Planners • Transmission Service Providers • Generator Owners • Generator Operators • Load-serving Entities • Interchange Authorities • Planning Authority • Resource Planners • Reliability Authorities

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2

LEARNING FROM THE BLACKOUTS

• Maintain reliability of the transmission area in accordance with Transmission Reliability Standards. Operator • Provide detailed maintenance schedules (dates and times) • Adjust DC ties within the transmission area for those Interchange Transactions that include the DC tie in the transmission path • Maintain defined voltage profiles. • Define operating limits, develop contingency plans, and monitor operations of the transmission facilities. • Provide telemetry of transmission system information • Maintain transmission system models (steady-state, dynamics, and Transmission short circuit) and apply appropriate tools for the development of Planner transmission plans. • Define and collect transmission information and transmission facility characteristics and ratings. • Develop plans within defined voltage and stability limits and within appropriate facility thermal ratings. • Define system protection and control needs and requirements, including special protection systems (remedial action schemes), to meet reliability standards. • Determine total transfer capacity values as appropriate.

Operates or directs the operations of the transmission facilities.

Develops a long-term (generally one year and beyond) plan for the reliability (adequacy) of the interconnected bulk electric transmission systems within its portion of the Planning Authority Area

Transmission Operations

Transmission Planning

• Transmission Owners • Other Transmission Planners • Transmission Operators • Transmission Service Providers • Generator Owners • Generator Operators • Load-serving Entities • Distribution Providers

• Transmission Owners • Transmission Service Providers • Generator Owners • Generator Operators • Distribution Providers • Planning Authorities • Reliability Authority

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2 ANNEX 2

195

196

Description

Administers the transmission tariff. Provides transmission services to qualified market participants under applicable transmission service agreements.

Function

Transmission Service

• Other Transmission Service Providers • Purchasing-Selling Entities • Generator Owners • Load-serving Entities • Interchange Authorities • Planning Authority • Reliability Authorities

• Planning Authorities • Resource Planners • Reliability Authorities

Responsible Related Entities that serve or are served by Entity the Responsible Entity

• Receive transmission service requests and process each request for Transmission service according to the requirements of the tariff. Service Provider • Approve or deny transmission service requests. • Approve Interchange Transactions from transmission service arrangement perspective. • Determine and post available transfer capability values. • Allocate transmission losses (MWs or funds) among Balancing Authority Areas.

• Notify others of any planned transmission changes that may impact their facilities. • Evaluate and plan for transmission service and interconnection requests beyond one year. • Develop and report, as appropriate, on its transmission expansion plan for assessment and compliance with reliability standards. • Monitor and report, as appropriate, on its transmission expansion plan implementation.

Key Tasks

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2

LEARNING FROM THE BLACKOUTS

Transmission Owner

Distribution Provider

• Install and maintain transmission facilities according to prudent utility practice. • Establish ratings of transmission facilities. • Develops interconnection agreements.

Owns and maintains transmission facilities.

Provides and operates • Provide the interface between the transmission system and the the “wires” between the end-use customer. transmission system and • Provide localised voltage reduction and load shedding as the end-use customer. necessary.

Transmission Ownership

Distribution

• Transmission Planners • Transmission Operators • Load-serving Entities • Planning Authorities

• Other Transmission Owners • Transmission Planners • Transmission Operators • Transmission Service Providers • Generator Owners • Load-serving Entities • Planning Authorities • Reliability Authorities

Responsible Related Entities that serve or are served by Entity the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2 ANNEX 2

197

198

Key Tasks

• Operate generators to provide energy or Interconnected Operations Services (or both) per contracts or arrangements. • Formulate daily generation plan. • Report operating and availability status of units and related equipment, such as automatic voltage regulators. • Develop annual maintenance plan for generating units and performs the day-to-day generator maintenance. • Establish generating unit ratings, limits and operating requirements. • Maintain generation facilities according to prudent utility practices. • Verify generating unit performance characteristics. • Purchase and sell generation or capacity. • Arrange Interchange Transactions . • Arrange for transmission service (as required by tariffs). • Purchase and sell Interconnected Operations Services. • Request implementation of Interchange Transactions.

Description

Operates generating unit(s) and performs the functions of supplying energy and Interconnected Operations Services.

Owns and maintains generating units.

Purchases or sells energy, capacity and all necessary Interconnected Operations Services as required. PurchasingSelling Entities may be Marketers or Merchant Affiliates.

Function

Generator Operation

Generator Ownership

PurchasingSelling

PurchasingSelling Entity

Generation Owner

Generator Operator

• Transmission Service Providers • Generator Owners • Generator Operators • Other Purchasing-Selling Entities • Load-serving Entities • Interchange Authorities

• Transmission Planners • Transmission Operators • Purchasing-Selling Entities • Load-serving Entities • Planning Authorities

• Transmission Operators • Purchasing-Selling Entities • Load-serving Entities • Balancing Authorities • Planning Authorities • Resource Planners • Reliability Authorities

Responsible Related Entities that Entity serve or are served by the Responsible Entity

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2

LEARNING FROM THE BLACKOUTS

Monitors, reviews, and • Audit and document compliance of all registered Responsible ensures compliance with Entities to Reliability Standards. Reliability Standards • Recommend sanctions or penalties for non-compliance with and administers Reliability Standards. sanctions or penalties for non-compliance to the standards.

Compliance Monitoring

Compliance Monitor

• Collect individual and develop overall load profiles and forecasts Load-Serving of end-user energy requirements. (daily, weekly, monthly, annually Entity etc…). • Identify and provide facilities for load curtailment. • Identify and provide facilities for self-provided Interconnected Operations Services. • Negotiate agreements for needed energy, transmission service, and Interconnected Operations Services. • Manage resource portfolios to meet demand and energy requirements of end-use customers.

Secures energy and transmission service (and related Interconnected Operations Services) to serve the end-use customer.

Load-Serving

• All Responsible Entities

• Transmission Planners • Transmission Operators • Transmission Service Providers • Generator Owners • Generator Operators • Purchasing-Selling Entities • Distribution Providers • Market Operator • Balancing Authorities • Planning Authorities • Resource Planners

Responsible Related Entities that Entity serve or are served by the Responsible Entity

Key Tasks

Description

Function

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2 ANNEX 2

199

200

Key Tasks

• Develop Reliability Standards for the planning and operation of the interconnected bulk electric transmission systems that serve the United States, Canada, and Baja California Norte, Mexico. • Develop compliance measurement and enforcement procedures for each Reliability Standard. • Develop Criteria and Certification Procedures for Balancing, Interchange, and Reliability Authorities, Transmission Operators, and others as needed. • Provide for appeals and dispute resolution procedures.

Description

Develops, maintains, and implements Reliability Standards to ensure the reliability of the interconnected bulk electric transmission systems. In North America this would include the United States, Canada, and Baja California Norte, Mexico.

Function

Standards Development Standards Developer

• Coordinates with other standards approving entities.

Responsible Related Entities that Entity serve or are served by the Responsible Entity

North American Electric Reliability Council Draft Functional Model (continued)

Annex 2

LEARNING FROM THE BLACKOUTS

Source: NERC (2004b).

Transmission Planning Area. That area under the purview of the Transmission Planner.

Transmission Arrangements. An agreement between a Transmission Service Provider and Transmission Customer (Purchasing-Selling Entity, Generator Owner, Load-Serving Entity) for transmission services.

Transaction. An agreement arranged by a Purchasing-Selling Entity to transfer energy from a seller to a buyer.

Total Transfer Capacity (TTC). This refers to the amount of electric power that can be moved or transferred reliably from one area to another area of an interconnected transmission system by way of all transmission lines (or paths) between those areas under specified system conditions.

Task. One of the elements that make up a Function in the Functional Model.

Responsible Entity. The label that NERC applies to an organisation that is responsible for carrying out the tasks within a Function.

Reliability Authority Area. The collection of generation, transmission, and loads within the boundaries of the Reliability Authority. Its boundary coincides with one or more Balancing Authority Areas.

Planning Authority Area. That area under the purview of the Planning Authority. It will include one or more Transmission Planning Areas.

Function. A group of tasks that can not be logically subdivided into other groups.

End-use Customer. The customer served by a Load-Serving Entity.

Customer. A Purchasing-Selling Entity, Generator Owner, Load-Serving Entity, or End-user.

Balancing Authority Area. The collection of generation, transmission, and loads within the metered boundaries of the Balancing Authority. The Balancing Authority maintains load-resource balance within this area.

Authority. The highest level of responsible entity for a particular function. The Reliability Authority is the highest level of all responsible entities.

Key Definitions:

* = Functions NERC considers especially critical for reliability, including transmission system security. ANNEX 2

201

REFERENCES

REFERENCES ACCC (Australian Competition and Consumer Commission) (2001), Final Determination: NEC - Ancillary Services Amendments (Application # A90742, A90743 & A90744), ACCC, Canberra. Alvarado, F. and Oren, S. (2002), “Transmission System Operation and Interconnection”, National Transmission Grid Study – Issue Papers, United States Department of Energy, Washington DC. Bialek J. (2004), “Recent Blackouts in US and Continental Europe: Is Liberalisation to Blame?”, Cambridge Working Papers in Economics, No. CWPE 407, University of Cambridge, Dept. of Applied Economics. Cambridge. Burns, R., Potter, S. and Witkind-Davis, V. (2004), “After the Lights Went Out”, The Electricity Journal, Vol. 17. No. 1, Elsevier Inc., New York. Carreras, B., et al. (2003), “Blackout Mitigation Assessment in Power Transmission Systems”, proceedings of the 36th Annual Hawaii International Conference on System Sciences, IEEE Computer Society Press, Hawaii, 6-9 January. This paper can be viewed at http://csdl2.computer.org/comp/ proceedings/hicss/2003/1874/02/187420065b.pdf. CEER (Council of European Energy Regulators) (2004), CEER Security of Electricity Supply Report 2004, CEER. This publication can be viewed at www.ceer-eu.org/. Chang, J., Murphy, D. and Graves, F. (2003), “Transmission Management in the Deregulated Electric Industry: A Case Study on Reactive Power”, The Electricity Journal, Vol. 16. No. 8, Elsevier Inc., New York. Cieslewicz, S. and Novembri, R. (2004), Utility Vegetation Management Final Report, prepared by CN Utility Consulting LLC for the United States Federal Energy Regulatory Commission, FERC, Washington DC. CIGRE (International Council on Large Electricity Systems) (2005), “Maintenance for HV Cables and Accessories”, Electra, No.221, CIGRE, Paris. CRE (Commission de Regulation de l’Energie) and AEEG (Autorita per l’Energia Elettrica e il Gas) (2004), Report on the Events of September 28th, 2003 Culminating in the Separation of the Italian Power System from the other UCTE Networks, CRE and AEEG. This publication can be viewed at www.autorita.energia.it/inglese/press/eng_index.htm Dy Liacco, T. E. (1967), The Adaptive Reliability Control System, IEEE Trans. on PAS, vol. PAS-86. 203

LEARNING FROM THE BLACKOUTS

ELCON (Electric Consumers Resource Council) (2004), “ELCON Study Reviews Impact of August 2003 Blackout on Facilities”, ELCON Report (No.1 2004), Washington DC. Elkraft System (2003), Power Failure in Eastern Denmark and Southern Sweden on 23 September 2003: Final Report on the Course of Events, Elkraft System, Ballerup, Denmark. Elkraft System (2004a), 2003 Annual Report, Elkraft System, Ballerup, Denmark. Elkraft System (2004b), System Plan 2004, Elkraft System, Ballerup, Denmark. EPRI (Electric Power Research Institute Inc.) (2001), Assessment Methods and Operating Tools for Grid Reliability Under Deregulation: The Transmission Program of EPRI’s Power Delivery Reliability Initiative, EPRI, Palo Alto, CA, (Product No. 2001.100140). EPRI (2003a), Electricity Technology Roadmap: 2003 Summary and Synthesis – Power Delivery and Markets, EPRI, Palo Alto, CA, (Product No. 2003. 1002639). EPRI (2003b), Online Probabilistic Reliability Monitor, EPRI, Palo Alto, CA, (Product No. 2003.1002262). EPRI (2003c), Moving Toward Probabilistic Reliability Assessment Methods: A Framework for Addressing Uncertainty in Power System Planning and Operation, EPRI, Palo Alto, CA, (Product No. 2003.1002639). Eurelectric (2004), Power Outages in 2003 - Task Force Power Outages (Ref: 2004-181-0007), Eurelectric, Brussels. European Parliament (2005), European Parliament legislative resolution on the proposal for a Directive of the European Parliament and of the Council concerning measures to safeguard security of electricity supply and infrastructure investment, first reading, [P6_TA(2005)0267], European Parliament, Strasbourg, July 2005. FERC (Federal Energy Regulatory Commission) (2004), Utility Vegetation Management and Bulk Electric Reliability – Report from the Federal Energy Regulatory Commission, FERC, Washington DC. FERC (2005), Rules Concerning Certification of the Electric Reliability Organisation; and Procedures for the Establishment, Approval, and Enforcement of Electric Reliability Standards, FERC, Washington DC, (Notice of 204

REFERENCES

Proposed Rulemaking of 1 September 2005, 18 CFR Part 38, Docket No. RM05-30-000). Gellings, C. Lordan, R. (2004), “The Power Delivery System of the Future”, The Electricity Journal, Vol. 17. No. 1, Elsevier Inc., New York. Gellings, C. (2004a), “A Power Delivery System to Meet Society’s Future Needs”, paper presented to the International Energy Agency Workshop on Transmission Network Reliability in Competitive Electricity Markets, International Energy Agency, Paris, 29-30 March. Gellings, C. (2004b), “R&D Needs in Electric Transmission and Distribution”, keynote address to the International Energy Agency Workshop on Electricity Transmission and Distribution Technology and R&D, International Energy Agency, Paris, 4-5 November. Giberson, M. and Kiesling, L. (2004), “Analysing the Blackout Report’s Recommendations: Alternatives for a Flexible, Dynamic Grid”, The Electricity Journal, Vol. 17. No. 6, Elsevier Inc., New York. Hagman Energy AB (2005), Survey of System Responsibility in the Nordic Countries – Final Report, prepared for Nordel (the Nordic council of transmission system operators). The publication can be viewed on the Nordel website at www.nordel.org/Publications/Reports. Haubrich, H-J. and Fritz, W. (1999), Study of Cross-Border Electricity Transmission Tariffs, prepared for the European Commission, DG XVII/C1, Aachen. Hauer, J., et al. (2002), “Advanced Transmission Technologies”, National Transmission Grid Study – Issue Papers, United States Department of Energy, Washington DC. IEA (International Energy Agency) (2003), Power to Choose – Demand Response in Liberalised Electricity Markets, Energy Market Reform, OECD/IEA, Paris. IEA (2004a), Transmission Network Reliability in Competitive Electricity Markets Workshop (various papers), OECD/IEA, Paris, 30-31 March. IEA (2004b), Electricity Transmission and Distribution Technology and R&D Workshop (various papers), OECD/IEA, Paris, 4-5 November. IEA (2004c), Transmission Network Performance in Competitive Electricity Markets Workshop (various papers), OECD/IEA, Paris, 18-19 November. IEA (2005), Lessons from Liberalised Electricity Markets, IEA, Paris. 205

LEARNING FROM THE BLACKOUTS

IMO (Independent Electricity Market Operator) (2003), Quick Takes – Market Suspension Update: Special Edition (various editions between 14-22 August 2003), IMO, Toronto, Canada. IMO (2004), August 2003 Blackout – Restoration Evaluation Report (Issue 1.0), IMO, Toronto, Canada. Joskow, P. and Tirole, J. (2004), “Reliability and Competitive Electricity Markets”, Center for the Study of Energy Markets Working Paper Series (CSEM WP 129), University of California Energy Institute, Berkeley, CA. Kirby, B. and Hirst, E. (2002), “Reliability Management and Oversight”, National Transmission Grid Study – Issue Papers, United States Department of Energy, Washington DC. Knight, U. G. (2001), Power Systems in Emergencies: From Contingency Planning to Crisis Management, Wiley, Chichester, England. Kirschen, D. and Stribac, G. (2004), “Why Investments Do Not Prevent Blackouts”, The Electricity Journal, Vol. 17. No. 2, Elsevier Inc., New York. Larsson, S. (2004), “The Blackout in Southern Sweden and Eastern Denmark, 23 September 2003”, paper presented to the International Energy Agency Transmission Reliability in Competitive Electricity Markets Workshop, IEA, Paris 30-31 March. Masiello, R., Spare, J. and Roark, A. (2004), “Reliability @ Risk: A New Paradigm for Assessing Reliability”, The Electricity Journal, Vol. 17. No. 10, Elsevier Inc., New York. Moss, D. (2004), “Competition or Reliability in Electricity? What the Coming Policy Shift Means for Restructuring”, The Electricity Journal, Vol. 17. No. 2, Elsevier Inc., New York. National Grid Transco (2004), GB Security and Quality of Supply Standard (Version 1.0), National Grid Transco. This material can be viewed on the National Grid Transco website at www.nationalgrid.com/uk/library/ documents/pdfs/GB_SQSS_V1.pdf. National Grid Transco (2005), The System Operator Transmission Owner Code (Version 2), National Grid Transco. This material can be viewed on the National Grid Transco website at www.nationalgrid.com/NR/rdonlyres/B056F1104B16-404D-9A91-E6DD408CB981/3834/STCdesignationcleanv2_rev8.pdf. NECA (National Electricity Code Administrator Ltd.) (2004), Annual Report 2003-04, NECA, Adelaide. 206

REFERENCES

NEMMCO (2001), Guide to Ancillary Services in the National Electricity Market, NEMMCO, Melbourne. This publication can be viewed on the NEMMCO website at http://www.nemmco.com.au/ancillary_services/160-0056.htm. NEMMCO (National Electricity Market Management Company Ltd.) (2004), Power System Incident Report – Friday 13 August 2004: Preliminary Report (Version 1.0), NEMMCO, Melbourne. NEMMCO (2005a), Power System Incident Report – Friday 13 August 2004: Final Report (Version 1.0), NEMMCO, Melbourne. NEMMCO (2005b), Power System Incident Report – Friday 13 August 2004: Load Shedding Report (Version 2), NEMMCO, Melbourne. NERC (North American Electric Reliability Council) (2004a), August 14, 2003 Blackout: NERC Actions to Prevent and Mitigate the Impacts of Future Cascading Blackouts, NERC, Princeton, New Jersey. NERC (2004b), NERC Reliability Functional Model – Function Definitions and Responsible Entities (Version 2), NERC. The publication can be viewed on the NERC website at www.nerc.com/~filez/functionalmodel.html. NERC (2004c), Technical Analysis of the August 14, 2003, Blackout: What Happened, Why, and What Did We Learn?, NERC, Princeton, New Jersey. NERC (2004d), Status Report on NERC Implementation of August 14, 2003, Blackout Recommendations, NERC. The publication can be viewed on the NERC website at www.nerc.com/~filez/blackout.html. NERC (2004e), Status Report: Recommendation 6a - Improve Operator and Reliability Coordinator Training, NERC. The publication can be viewed on the NERC website at www.nerc.com/~filez/blackout.html. NERC (2005a), Standards FAC-003-1 – Transmission Vegetation Management Program, NERC. The publication can be viewed on the NERC website at www.nerc.com/~filez/standards/Vegetation-Management.html. NERC (2005b), North American Electric Reliability Council Status of August 2003 Blackout Recommendation, NERC, Princeton, New Jersey. Nilssen, G. and Walther, B. (2001), “Market-based Power Reserve Acquirement”, paper presented to the Methods to Secure Peak Load Capacity in Deregulated Electricity Markets Conference, Stockholm, 7-8 June. Nordel, Annual Reports (various years), Nordel. These publications can be viewed on the Nordel website at www.nordel.org/Publications/Annual Reports. 207

LEARNING FROM THE BLACKOUTS

Nordel (2002), Reliability Standards and System Operating Practices in Nordel, Nordel. The publication can be viewed on the Nordel website at www.nordel.org/ Operations Committee/Publications/Reliability Standards and System Operating Practices in Nordel. Nordel (2004a), System Operation Agreement 2004, Nordel. The publication can be viewed on the Nordel website at www.nordel.org/Operations Committee/Publications/System Operation Agreement 2004. Nordel (2004b), Report on System Operation – A Background Report Prepared by the Nordel Operation Committee/OPG in the Nordel Project for Enhancing Efficient Functioning of the Nordic Electricity Market, Nordel. The publication can be viewed on the Nordel website at www.nordel.org/Publications/ Reports/Background Report System Operation, 2005. Nordel (2005a), Enhancing Efficient Functioning of the Nordic Electricity Market, Nordel. A Report prepared for the Nordic Council of Ministers. The publication can be viewed on the Nordel website at www.nordel.org/ Publications/Reports/Reprot: Enhancing Efficient Functioning of the Nordic Electricity Market. Nordel (2005b), Report on Peak Load Mechanisms – A background report prepared by the Nordel Operation Committee/OPG in the Nordel Project Enhancing efficient functioning of the Nordic Electricity Market, Nordel. The publication can be viewed on the Nordel website at www.nordel.org/ Publications/Reports/Background Report Peak Load Mechanisms. Nordel (2005c), Nordic TSOs’ Action Plans in Enhancing and Monitoring Demand Response, Nordel. The publication can be viewed on the Nordel website at www.nordel.org/Publications/Reports/Background Report: Action Plans in enhancing and Monitoring Demand Response. OXERA (Oxford Economic Research Associates) (2005), “The Cost of Failure”, The Utilities Journal (Vol. 8, January), OXERA, Oxford. PJM (PJM Interconnection) (2005), 2004 State of the Market, Market Monitoring Unit, PJM. This publication can be viewed on the PJM website at www.pjm.com/ markets/market-monitor/downloads/mmu-reports/pjm-som-2004.pdf Ray, C. (2004), Power System Planning: System Development – Maintaining Security, paper presented to the IEA Transmission Reliability in Competitive Electricity Markets Workshop, IEA, Paris 30-31 March. SFOE (Swiss Federal Office of Energy) (2003), Report on the Blackout in Italy on 28 September 2003, SFOE. This publication can be viewed on the SFOE 208

REFERENCES

website at www.energie-schweiz.ch/imperia/md/content/energiemrkteetr gertechniken/elektrizitt/strompanne03/3.pdf. Statistics Canada (2003a), Gross Domestic Product by Industry, August 2003 (Cat.# 15-001), Federal Publications Inc., Toronto, Canada. Statistics Canada (2003b), Labour Force Survey; and Monthly Survey of Manufacturing, August 2003 (Cat.# 31-001), Federal Publications Inc., Toronto, Canada. Stoft, S. (2002), Power System Economics – Designing Markets for Electricity, IEEE Press, Piscataway, New Jersey. Svenska Kraftnät (2003), The black-out in southern Sweden and eastern Denmark, September 23, 2003 – Preliminary Report, Svenska Kraftnät, Vällingby, Sweden. Svenska Kraftnät (2004), The black-out in southern Sweden and eastern Denmark, September 23, 2003, Svenska Kraftnät. This publication can be viewed on the Svenska Kraftnät website at www.svk.se/upload/3813/ IEEE_engelskpresentation_jul2003_2230.pdf. Swiss Federal Inspectorate for Heavy Current Installations (2003), Report of the Federal Inspectorate for Heavy Current Installations on the blackout that occurred in Italy and border regions of Switzerland on 28 September 2003, SFOE. This publication can be viewed on the SFOE website at www.energieschweiz.ch/imperia/md/content/energiemrkteetrgertechniken/elektrizitt/st rompanne03/11.pdf. UCTE (Union for the Co-ordination of Transmission of Electricity) (2004a), Final Report of the Investigation Committee on the 28 September 2003 Blackout in Italy, UCTE. This publication can be viewed on the UCTE website at http://www.ucte.org/pdf/News/20040427_UCTE_IC_Final_report.pdf. UCTE (2004b), Operation Handbook, UCTE. This publication can be viewed on the UCTE website at www.ucte.org/pdf/ohb/Operation_Handbook_ 20.07.2004.pdf. UCTE (2005a), UCTE Reliability Standards Become Binding, (press release of 13 May 2005), UCTE. This publication can be viewed on the UCTE website at http://www.ucte.org/pdf/News/20050512_MLA_V2.pdf. UCTE (2005b), Seven Actions for a Successful Integration of Wind Power into European Electricity Systems, (press release of 17 May 2005 – Long Version), UCTE. This press release can be viewed on the UCTE website at www.ucte.org/pdf/News/20050517_Actions_WIND_short&long.pdf. 209

LEARNING FROM THE BLACKOUTS

UCTE (2005c), UCTE inter-TSO Multilateral Agreement entered into force on 1 July 2005, (press release of 19 July 2005), UCTE. This press release can be viewed on the UCTE website at www.ucte.org/pdf/News/20050701_MLA_ in_force_v5.pdf. US DOE (United States Department of Energy) (2002), National Transmission Grid Study, US DOE, Washington DC. This publication can be viewed on the US DOE website at www.eh.doe.gov/ntgs/gridstudy/main_print.pdf. United States Congress (2005), Electricity Modernisation Act 2005 (Title XIIElectricity of the Energy Policy Act of 2005), Washington. United States-Canada Power System Outage Task Force (2004a), Final Report on the August 14, 2003 Blackout in the United States and Canada: Causes and Recommendations, Natural Resources Canada and US DOE. This publication can be viewed at https://reports.energy.gov/BlackoutFinal-Web.pdf. United States -Canada Power System Outage Task Force (2004b), The August 14, 2003 Blackout One Year Later: Actions Taken in the United States and Canada To Reduce Blackout Risk, Natural Resources Canada and US DOE. This publication can be viewed at www.electricity.doe.gov/documents/blackout_ oneyearlater.pdf. Vandenberghe, F. (2004), “Development of Interconnections and Reliability Standards”, Electricity Trade in Europe – Review of the Economic and Regulatory Changes, International Energy and Resources Law and Policy, Kluwer Law International, The Hague, The Netherlands. Walther, B. and Vognild, I. (2005), “Statnett’s Option Market for Fast Operating Reserves”, Demand Response Dispatcher (Vol.1, Issue 6), International Energy Agency - Demand Side Management Programme Task XIII. This publication can be viewed at www.demandresponseresources.com/Portals/0/Demand% 20Response%20Dispatcher%20Vol%201%20Issue%206.pdf

210

The Online Bookshop International Energy Agency All IEA publications can be bought online on the IEA Web site: w w w. i e a . o r g / b o o k s You can also obtain PDFs of all IEA books at 20% discount. Books published before January 2004 - with the exception of the statistics publications can be downloaded in PDF, free of charge, on the IEA website.

IEA BOOKS Tel: +33 (0)1 40 57 66 90 Fax: +33 (0)1 40 57 67 75 E-mail: [email protected]

CUSTOMERS IN NORTH AMERICA

International Energy Agency 9, rue de la Fédération 75739 Paris Cedex 15, France

You can also send

CUSTOMERS IN THE REST OF THE WORLD

your order Turpin Distribution The Bleachery 143 West Street, New Milford Connecticut 06776, USA Toll free: +1 (800) 456 6323 Fax: +1 (860) 350 0039 [email protected] www.turpin-distribution.com

to your nearest OECD sales point or through the OECD online services: www.oecdbookshop.org

Turpin Distribution Services Ltd Stratton Business Park, Pegasus Drive, Biggleswade, Bedfordshire SG18 8QB, UK Tel.: +44 (0) 1767 604960 Fax: +44 (0) 1767 604640 [email protected] www.turpin-distribution.com

IEA PUBLICATIONS 9, rue de la Fédération, 75739 Paris Cedex Pre-press by Linéale Production Printed in France by JOUVE Code 612005331p1 ISBN 92 64 10961 7 December 2005

INTERNATIONAL

ENERGY MARKET EXPERIENCE

ENERGY

AGENCY

This pdf-ﬁle is for the personal use of the purchaser only. The information contained in it is subject to copyright by the OECD/IEA. This material is for sale; therefore, the User may not copy, republish, lend, post on any server (unless a multi-user licence is purchased for use on the intranet, in which case the multi-user license terms apply), use on Web sites, transmit or redistribute to any mailing lists any part of the material for commercial purposes, for compensation of any kind or simple dissemination.

Lessons from Liberalised Electricity Markets

INTERNATIONAL ENERGY AGENCY The International Energy Agency (IEA) is an autonomous body which was established in November 1974 within the framework of the Organisation for Economic Co-operation and Development (OECD) to implement an international energy programme. It carries out a comprehensive programme of energy co-operation among twenty-six of the OECD’s thirty member countries. The basic aims of the IEA are: • to maintain and improve systems for coping with oil supply disruptions; • to promote rational energy policies in a global context through co-operative relations with non-member countries, industry and international organisations; • to operate a permanent information system on the international oil market; • to improve the world’s energy supply and demand structure by developing alternative energy sources and increasing the efficiency of energy use; • to assist in the integration of environmental and energy policies. The IEA member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Japan, the Republic of Korea, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, the United States. The European Commission takes part in the work of the IEA.

ORGANISATION FOR ECONOMIC CO- OPERATION AND DEVELOPMENT The OECD is a unique forum where the governments of thirty democracies work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies. The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The European Commission takes part in the work of the OECD.

© OECD/IEA, 2005 No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications should be sent to: International Energy Agency (IEA), Head of Publications Service, 9 rue de la Fédération, 75739 Paris Cedex 15, France.

FOREWORD

FOREWORD In ever more globalised and automated economies the role of electricity is increasingly important as a driver for economic prosperity. Reliable and affordable supply of electricity is essential for the competitiveness of global industrial product markets and a necessary ingredient in the daily workings of modern societies. At the same time, environmental impacts of energy usage are one of the most difficult global policy challenges. Reliable access to affordable electricity supply with acceptable environmental impacts is only achieved with comprehensive and carefully balanced policy actions to establish the necessary incentive-based framework. To that end, liberalisation of electricity markets is a development path and policy option that has been implemented or considered in all IEA member countries. Through competition in liberalised markets incentives are created to drive for more efficient operation of electricity systems and more efficient investment decisions in terms of timing, sizing, siting and choice of technology. Even if liberalised markets leave critical policy challenges un-resolved, the transparency created by competition tends to improve the framework for targeted policy actions to address issues such as environmental quality and reliability. After up to ten years’ experience with liberalised electricity markets and even longer in some cases important lessons can now be drawn from some pioneering countries and regions. Theoretical principles for successful liberalisation can now be augmented with more qualified policy prescriptions based on real-world experience. Even if some pioneering markets have operated with considerable success for a number of years, liberalisation has shown it self not to be a single event, but rather a long process that requires on-going government commitment. No markets are perfect, and they will continue to evolve and develop to match the needs of electricity systems - systems that are at the same time undergoing considerable change. This book addresses the main principles of successful liberalisation with actual experiences and outcomes, hopefully providing decision makers within government and industry with policy prescriptions on the key issues. One point of departure is to ask whether liberalisation of electricity markets is feasible – that it, has it been possible to develop a functional market without jeopardizing reliability and other public policy priorities. Secondly, if liberalisation has worked in that it has been able to achieve this balance, has it delivered the expected outcome in terms of real economic benefit. An affirmative answer to these questions leads to the books’ focus on what issues are critical and what 3

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

approaches best lead to successful electricity markets, allowing the book to point to best practices. This book is the first in the IEA series on electricity market experiences. It is published under my authority as Executive Director of the International Energy Agency. Claude Mandil Executive Director

4

ACKNOWLEDGEMENTS

ACKNOWLEDGEMENTS The principal author of this book is Ulrik Stridbaek of the Energy Diversification Division, working under the direction of Ian Cronshaw, Head of Division, and Noé van Hulst, Director of the Office for Long-Term Co-operation and Policy Analysis. This book has benefited greatly from suggestions and comments by Doug Cooke, from the IEA, who particularly contributed to issues of transmission investment and performance. Dong-wook Lee and Josef Sangiovanni from the IEA contributed to the annexes, in particular. Valuable comments were provided by Jolanka Fisher, from the IEA, Mats Nilsson, from the Swedish Energy Markets Inspectorate, Henning Parbo, from Energinet.dk and delegates from the IEA’s Standing Group on Long-Term Co-operation. Muriel Custodio managed the production of the book, Corinne Hayworth, designed the layout and the front cover, and Marilyn Smith edited the book.

5

TABLE OF CONTENTS

TABLE OF CONTENTS Executive Summary .......................................................11 Introduction.................................................................27 Chapter 1 Electricity market liberalisation has delivered long term benefits .......................31

Indicators of Success..........................................................................................................32 Market Liberalisation is a Process rather than an Event.......................................42 Distributing the Benefits of Liberalisation.................................................................44 Chapter 2 Competition is the fuel for effective markets ..........................................47

Competitive Markets Replace Vertically Integrated Utilities ...............................48 Legislative and Regulatory Framework for Effective Competition.....................57 Regulating Competition ...................................................................................................63 Chapter 3 Price signals are the glue...................................71

Prices to Reflect the Inherently Volatile Nature of Electricity .............................71 Locational Pricing................................................................................................................76 Markets for Ancillary Services.........................................................................................87 Cross-border Trade Creates Benefits .............................................................................88 Market Models .....................................................................................................................94 Chapter 4 Risk Management and Consumer Protection...........................................99

Power Utilities Use Contracts to Manage Risks.....................................................101 Contracts Offer Protection for Consumers ...............................................................108 Retail Competition............................................................................................................112

7

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Chapter 5 Investment in Generation and Transmission .......................................................117

Market-driven Investment in Generation...................................................................118 Demand Participation as an Alternative..................................................................133 Planned Investment and Capacity Measures..........................................................139 Investment in Transmission Networks .......................................................................144 Co-ordination of Transmission Investment...............................................................152 Chapter 6 When do liberalised electricity markets fail? ......................................................................155

Reliability of Supply in Liberalised Electricity Markets........................................156 Addressing Environmental Issues and Climate Change .....................................163 Annex 1 British Electricity Trading and Transmission Arrangements ..............................171 Annex 2 The Nordic electricity market...........................181 Annex 3 Australian National Electricity Market ..........193 Annex 4 Pennsylvania - New Jersey - Maryland Interconnection (PJM) ........................................203 References .................................................................213 List of Tables

1 • Models for unbundling of transmission system operation......................51 2 • Liquidity in various electricity markets: turnover in different market segments as share of total consumption, 2004........................104 3 • Switching rates amongst contestable customers: number of customers no longer supplied by incumbent retailer as share of total number of residential and non-residential customers.............113 4 • Price caps in PJM, Australian, British and Nordic markets. ....................120

8

TABLE OF CONTENTS

5 • Demand participation: committed by system operators with minimum additional assessed and observed demand participation ..137 List of Figures

1 • Electricity prices paid by end-consumers .......................................................33 2 • Flow over interconnectors between western Denmark and Norway, and western Denmark and Sweden ...............................................36 3 • Employment in electricity, gas and water .....................................................37 4 • A new model for the electricity sector............................................................48 5 • Timeline for planning and operation..............................................................53 6 • Supply and demand in liberalised electricity markets...............................74 7 • Electricity trade between two areas ................................................................82 8 • Reservoir levels in Norway and Sweden, trade between Norway and Sweden, and Nord Pool spot price, 2002-2004 ................................93 9 • Year in which all consumers are allowed to switch retailer..................112 10 • Average monthly prices in national electricity market, Australia.......123 11 • Installed capacity principal power stations in National Electricity Market, Australia..................................................................................................124 12 • Price duration curves for the highest percent in South Australia .....125 13 • Price duration curves: national electricity market, Australia................128 14 • Nord Pool forward prices, yearly contract, three years ahead .............132 15 • Demand participation improves market and system performance....134 16 • Annual average increase in length of 220-400 kV transmission lines in 16 European countries .......................................................................145 17 • The value chain for reliable electricity supply: Energy security, adequacy and system security ........................................157 18 • Wind power production and Nord Pool day-ahead prices in western Denmark, December 2003.........................................................168 19 • Wind power production in western Denmark and trade with Nordic neighbours.....................................................................................169

9

EXECUTIVE SUMMARY

EXECUTIVE SUMMARY Over the past decade, several IEA member countries embarked on a policy path of market liberalisation of the electricity supply industry. Pioneers in electricity market reform have now been operating with considerable success for a number of years, delivering substantial benefits to economies. Finding the most effective way to develop competitive electricity markets that fulfil the goals of real economic benefits has not been clear, however. Scepticism and concerns are voiced in many countries, and debate continues on several key issues. The sceptics point to the California crisis and market breakdown in 2001, only a few years after the new market was launched under much publicity and the subsequent and spectacular bankruptcy of the large energy company Enron. The widespread blackouts in North America, Italy and Scandinavia in 2003 are also sometimes used to argue that electricity market liberalisation is a failed concept. Today, extensive expert opinion and research material identifies the root causes of some of the past failures. The California crisis can be attributed to a wide range of factors, including important regulatory failures in the set-up of the California market. Official investigations into the 2003 blackouts do not blame liberalised markets for being the root causes of those events. Learning from these blackouts, a recent IEA study on transmission system security points out that liberalisation has fundamentally changed the use of transmission systems. The management of transmission systems still needs to adapt to these changes. With that adjustment in place, liberalised markets can provide a framework that improves system security, largely through increased co-operation amongst jurisdictions. While the public has focused on the remarkable failures of the past decade, several electricity markets have been operating successfully and have developed into robust markets during the same period. In all IEA member countries, the liberalisation process has progressed at varying speeds. Despite the fact that no straightforward path to success has emerged, there is a general lesson to be learned: Electricity market liberalisation is not an event. It is a long process that requires strong and sustained political commitment, extensive and detailed preparation, and continuous development to allow for necessary improvements while sustaining on-going investment. It is, in fact, a process that has not yet been completed anywhere in the world – nor will it be in the foreseeable future. This book focuses on the experiences and lessons learned in several pioneering markets that have now operated with considerable success for a decade or more. While discussing prescriptions for successful electricity market 11

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

liberalisation and drawing on real-world experiences, three questions are pursued: Has electricity market liberalisation been practicable? Has it delivered real economic benefits? If so, what are the key issues for success? The key messages, conclusions and recommendations following from the findings are presented in this Executive Summary.

Electricity Market Liberalisation Delivers Long-term Benefits .............................................................. Traditionally, electricity sectors developed and operated within strictly regulated frameworks in which vertically integrated utilities have handled most or all activities – from generation to transport to distribution. Moreover, it has been a centrally planned activity, wherein needs are assessed and fulfilled by electricity system planners and all associated costs are passed on to consumers. But traditional, vertically integrated utilities tend to create substantial overcapacity, a fact that became more obvious when electricity demand growth slowed during the 1980s and 1990s in many IEA member countries. In addition to reducing this overcapacity, liberalisation has also been shown to provide large potential gains from improved efficiency in the operation of generation plants, networks and distribution services. Monitoring of electricity rates paid by different customer classes is one basic way to assess the performance of liberalised electricity markets. Indeed, many countries promised falling prices prior to launching liberalisation processes. In markets that have liberalised successfully, there is a clear trend of falling electricity prices for industrial consumers, in both nominal and real terms. The trend is less clear, and certainly slower, for household consumers. However, prices paid by consumers do not necessarily reflect the costs of producing and transporting electricity. Some consumer groups often subsidise other consumer groups. Different parts of the value chain – from the recovery of fuels to generation and transport of electricity – are also often subsidised in one way or another, or are not fully cost reflective for other reasons. Electricity rates and taxes are often related in non-transparent ways. Changes in fuel costs and environmental regulation, affect final costs of supplying electricity and seems to be important drivers for recent increases in electricity tariffs in many particularly European IEA member countries, but are not directly related to the effects of electricity market liberalisation. In addition, investment decisions made within a vertically integrated industry influence electricity costs for a long time, hence the effects of past investment decisions

12

EXECUTIVE SUMMARY

will be reflected in retail prices for several years to come. All in all, these factors make electricity retail prices paid by end-users complex to interpret. In reality, retail prices are poor indicators of whether performance development is positive in the electricity industry. Examining performance in various specific segments of the value change paints a clearer picture. For example, many countries mothballed unused generation assets and data show that they now use existing plants more efficiently. At the same time, fundamental changes in the use of transmission assets has created more dynamic and enhanced usage, often resulting from increased trade across jurisdictions. Other indicators show marked increases in labour productivity. A recent study by the Organisation for Economic Co-operation and Development (OECD) explores the benefits of liberalising product markets and reducing barriers to international trade and investment across several regulated sectors. It singles out electricity as one of the sectors with the greatest potential for improvement. The results of the analysis assess the total annual benefits across all sectors to be 1% to 3% of GDP in the United States and 2% to 3.5% of GDP in the European Union. The study assesses only the static benefits from increased trade and better allocation of resources, but recognises that dynamic gains from increased innovation may be even greater. According to traditional performance indicators, electricity market liberalisation is already delivering significant benefits. But it is perhaps even more important to consider how a liberalised electricity sector is better equipped to meet challenges and exploit opportunities of a future more diverse and flexible portfolio of technologies. Combined Cycle Gas Turbines is a preferred generation technology option in many circumstances which is re-enforced in the new market framework where focus on operational and financial flexibility has increased. Uncertainty, which creates financial risk, is seen in a new light; today’s market players show a preference for less capital-intensive and smaller units. Technological developments are also creating opportunities for greater consumer involvement in decisions that determine the system. Another driver for change in the technology mix comes from government support of a wide range of technologies – many for their environmental merits and some of which are already being deployed at high rates. Nuclear power is re-surfacing as a technology that warrants serious consideration in some countries. All in all, a more diverse electricity system is beginning to emerge. Correspondingly, management of this system must accommodate many players, ranging from very small generation units and demand resources to very large nuclear power 13

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

plants. Thanks primarily to transparent price signals, liberalised markets are creating a more level playing field and allowing for the necessary coordination of many diverse market players. As liberalised markets begin to mature, it becomes more obvious that a centrally planned and vertically integrated approach is less appropriate for such a diverse system and is, in fact, likely to be a barrier to the innovation required to meet the future need. For the moment, it is critical to avoid being overly short sighted. Liberalisation is expected to bring large economic benefits for consumers and societies in the long term and evidence so far indicates that markets can deliver these benefits. But in the short term certain groups may not realise immediate benefits or may even experience losses. Vertically integrated utilities are likely to feel threatened by the requirement to unbundle. Consumer groups that previously benefited from subsidised electricity tariffs (at the expense of other consumers) may perceive liberalisation as a loss as cross-subsidies are unwound. Certain segments of the utilities’ workforce will feel threatened when open competition demands higher efficiency and increased labour productivity. Without question, one of the most critical policy challenges facing decision makers is the management of social and equity issues in distributing the benefits of electricity market liberalisation.

Key Message Electricity market liberalisation has delivered considerable economic benefits. Under pressure from competition, assets in the electricity sector are used more efficiently, thereby bringing real, long-term benefits to consumers. Liberalisation to introduce competition is, however, a long process rather than an event: it requires on-going government commitment to resolve challenges when vested interests and cross-subsidies are unwound.

Government has a Critical but Fundamentally Changed Role........................................ Regardless of the approach to liberalisation, the process requires strong government involvement. In fact, the level of on-going political commitment invested significantly influences the outcome. In the absence of clear signs of 14

EXECUTIVE SUMMARY

commitment, regulatory uncertainty may well become self-fulfilling and undermine a positive outcome. From time to time, all electricity systems will experience a crisis. Such crises have become important tests of the robustness of liberalised electricity markets and, perhaps even more importantly, of the robustness of the political framework backing the liberalisation process. At difficult junctures in market development, signals of strong political commitment – often expressed by not intervening – can lead to the necessary market responses. Effective markets are fuelled by competition. Thus, one of government’s most decisive roles is to establish a framework that allows for the development of effective competition. Liberalisation creates benefits by introducing incentives for higher efficiency and more innovation – with competition as the driver. The first step required to introduce competition is to break down the monopolies that exist in traditional vertically integrated utilities. It is necessary to separate network activities from all other activities, either through legal unbundling of the network entities or, more effectively, through true ownership unbundling. The key is to introduce competition in as many parts of the value chain as possible – from generation to consumption. Remaining natural monopolies (e.g. networks and system operation) should be subject to continued and improved economic regulation. Unbundling effectively breaks up the centralised decision-making process found in vertically integrated utilities, replacing it with a decentralised process where market players make decisions within markets. This can only work smoothly when markets are ‘effective’, but effective markets do not develop automatically. Creating a level playing field and developing effective, competitive marketplaces requires establishing detailed market rules, design and regulation. Within the on-going liberalisation processes, the level of government involvement through detailed legislation and rule-making has varied. But it is evident that governments are critical to establishing a framework with the necessary incentives. At the same time, independent regulators are one of the critical bodies within this framework; their role in overseeing compliance with legislation and ensuring fair and efficient economic regulation of networks is fundamental to successful market development. Real-time system operation is an aspect of the electricity sector that is maintained as a natural monopoly and, thus, should be unbundled from other competitive segments of the value chain. Market rules, design and regulation aim to direct all actions transparently, but many subtleties remain in secure, day-to-day system operation. It is inevitable that system operators will preserve certain discretionary powers, regardless of careful efforts to

15

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

regulate grid access. Their independence is particularly critical to the creation and further development of well-functioning and robust markets. In the new decentralised industry structure, transparency is a prerequisite for developing competitive liberalised electricity markets. Competitive market players do not automatically (or voluntarily) collect and publish fundamental market data and statistics. Therefore, it is important to redefine responsibility for this necessary task in liberalised markets. Increased transparency is a proven, strong instrument to ensure continuous development towards more effective markets. In fact, transparency adds to the benefits of liberalisation in its own right, by improving the decision-making framework for all actors – policy makers, industry and consumers alike. But a formal framework that allows for competition and creates a level playing field is not enough. Competition will flourish only if multiple players compete in the market. Governments and regulators have managed to enhance competition through various means, but a high level of market concentration remains a serious concern in several markets. Effective markets and transparency have been vital to easing access for new-comers. In addition, extending markets across countries and regions helps enable the “import of competition”; this is particularly important in smaller jurisdictions in which the need for consolidation limits the number of market players that can operate efficiently. To date, achievements are more limited in ex post regulation of competition. It is illegal to exercise market power, but it often remains difficult to prove such behaviour. In some cases, dealing with market power abuse has been further complicated when the largest companies are regarded as national champions or provide substantial revenue streams to their public owners. Some claim that market failures are inherent across the value chain in electricity markets requiring government intervention. But, upon closer scrutiny, many alleged failures turn out rather to be the result of regulatory failures. In the event of real market failures – as might arise from concerns about reliability of supply and the environmental impacts of electricity production – governments may be called upon to intervene in more active ways. Analysing reliability of supply in unbundled electricity sector has also called for an “unbundling” of the concept of reliability of supply into its relevant parts of the value chain. Concern has been voiced about secure supply of fuel for power generation, adequacy of investment in generation and network assets, and the security of real-time system operation. Considerations concerning market failures involving the secure supply of energy or fuel for electricity generation are an issue with geo-political 16

EXECUTIVE SUMMARY

dimensions. Increased use of natural gas as a fuel for power generation raises the issue of IEA member countries’ dependence on natural gas supplies from non-member countries – and stresses the importance of developing, in parallel, a competitive market for natural gas. Thus far, the levels of security of fuel supply ensured by commercial market players have not undermined the performance liberalised electricity markets. Also, so far, the efficiency of liberalised markets has sufficiently incentivised adequate generation capacity. Despite efforts to integrate some network investment decisions in the competitive market framework, networks remain, more or less, regulated monopolies. Adequate network assets must be incentivised through a framework of economic regulation. Finally, when it comes to secure real-time system operation, markets so far have failed to provide a complete framework of incentives without jeopardising system security. Government intervention is necessary and this has been carried out (rather effectively) through the establishment of truly independent system operators. A recent IEA publication explores the main requirements and challenges for effective regulation that creates the necessary levels of transmission system security.1 The environmental effects of electricity generation are not addressed by normal incentives in competitive markets. Environmental benefits are classical public goods and liberalised electricity markets will not adequately account for their value – or for the cost of their potential loss. Policy intervention is needed to ensure they are properly taken into account. Policies motivated by environmental and climate change concerns are already having serious impacts on liberalised electricity markets, as was intended. Many environmental policies are, however, potentially distortive beyond the initial intent, particularly when looking across internal markets within the context of international competition. Direct financial support for particular technologies, or non-transparent barriers that block development of others, can lead to inefficiencies and distort competition. This adds uncertainty to the investment decision process and ultimately poses a threat to the system. In several liberalised electricity markets, the preferred option to address this issue is implementation of cap-and-trade policies. This approach transforms the political goal into an obligation imposed upon market players. Market players are then left to fulfil the obligations in ways that they find optimal, including trading the obligations amongst themselves and finding alternative technology solutions. 1. IEA (2005), Learning from the Blackouts: Transmission System Security in Competitive Electricity Markets.

17

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

For example, carbon dioxide (CO2) emission trading was introduced in Europe to address climate change. Creating a market for CO2 emission permits creates a price for CO2 emissions, thereby making it possible to internalise the environmental cost within the total cost of producing electricity. Following similar principles, support for renewable energy has changed from direct financial support for specific technologies to market-based certificate systems backed by obligations. Fulfilling climate change policies through cap-andtrade instruments seems to be the least distortive option. In addition, this approach adds transparency that enhances the quality of decisions made by policy makers and industry.

Key Message Establishing truly independent and committed regulators and system operators precedes a competitive framework. Liberalisation still requires strong and committed government involvement, but in fundamentally changed roles. Vertically integrated utilities must be unbundled. Governments, independent regulators and independent system operators must collaborate to establish rules, market design and regulation that create a competitive market place and support its further development. Continued government commitment and signs of support is crucial, particularly at the difficult junctures every market will, inevitably, face.

Price Signals are the Glue .................................................... In the process of unbundling utilities to introduce competition, vertical integration has been replaced with markets comprising multiple players. In this new framework, price signals direct decisions in the marketplace. Efficient decisions depend on correct signals, i.e. price signals that reflect the real costs, benefits and values of producing, transporting and consuming electricity. Electricity has a value to the consumer only if it is supplied at the right place, at the right time, in the right volume and at an acceptable quality. The locational aspect of electricity pricing is the most controversial and complex issue in efficient pricing. Principles that establish a price for each node in a system are the ideal reference because they value electricity based on where it is generated and delivered, and some markets come close to achieving this. 18

EXECUTIVE SUMMARY

However, there are important trade-offs to consider when choosing pricing principles that could justify a less fine-tuned, zone-based system, where a price is established for several nodes that are rarely congested. Even though there are important trade-offs, the main controversy often relates more to social equity and distribution rather than specific pros and cons of market functioning and system operation. Nodal pricing evolved as a necessity in highly meshed networks where transmission lines are criss-crossing the electricity system (e.g. North America); zonal pricing is accepted as a good approximation in more radial networks, where the structure of congestion is less complex (e.g. Australia). Higher transaction costs and the greater complexity of nodal pricing are often used to argue for pricing principles that are less reflective of location. In reality, evidence shows that obvious congestion points have often not been priced appropriately. The highly meshed network in continental Europe is currently developing into a zonal market, often with countries constituting entire zones, thereby potentially bluring price signals and inhibiting efficiency. Open trade across jurisdictions is one of the classical merits of liberalised and competitive markets. It enables exploitation of comparative advantages – at mutual economic benefit for all regions involved. Electricity generation and transport include many factors related to resource endowments, geographical characteristics and regional skills. They are both very capital-intensive businesses that can realise significant gains by optimising asset use across as large an area as possible. This is particularly true for smaller jurisdictions. But trade across jurisdictions relies on co-operation amongst system operators. Therefore, independence and appropriate incentives of system operators are critical in the development of cross-border trade. Designing an appropriate market model is a matter of creating trading arrangements that fit to the specific circumstances of each electricity system while also addressing broader key issues of unbundling, third-party access, cost-reflective pricing and transaction costs. There is not one single winning market model - no one-size-fits-all – in each situation important trade-offs must be made. However, there is one common feature of all successful markets: some sort of formal price quotation, conceived through formal market design. Cost-reflective pricing principles are an aspect that must be fostered through regulation and active market design and will not develop without welldesigned market rules. Electricity consumption and supply are inherently volatile. But the volatility is an inseparable characteristic of the service and is not related to the organisation of the sector. Liberalised electricity markets create a more 19

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

transparent framework, allowing for cost-reflective pricing reflecting this volatility. In some instances, government interventions to suppress volatility and cap prices below what can be justified by economic reasons have blurred price signals and slowed market responses. Price volatility creates risks for market players, including generators and consumers. Risks are the result of uncertainty, and there is considerable uncertainty connected with many of the fundamental factors that determine electricity generation, transport and consumption. In the previous model of a vertically integrated and regulated sector, all costs – and, therefore, all risks – could be passed on to consumers. Liberalised markets make risks more transparent and, more importantly, re-allocate these risks to the decision makers themselves. In liberalised electricity markets, business risks can be effectively managed through contracts. Generators, retail suppliers and consumers can agree on prices, volumes, times and other conditions that create the desired certainty within the framework of the contract. In fact, liquid and effective markets for financial contracts improve competition by enabling sophisticated risk management. This, in turn, eases market access for new and smaller market players and contributes to ensuring that market power is not exercised. Most markets provide a framework for a liquid market in the day-ahead and realtime segments through market rules and design. In some markets, relatively liquid and effective financial markets for longer-duration contracts are developing, but the evolution of these markets remains a major concern for the creation of robust electricity markets.

Key Message A framework for efficient and cost-reflective prices is established through regulation and market design. Prices that reflect the inherent volatility of electricity are critical in creating a framework for efficient decisions regarding operation and investment in liberalised markets. Electricity market design, as conceived and implemented by governments, independent regulators and independent system operators, can create prices that reflect real costs through mechanisms that incorporate time, volume and location. Blurring price signals with price caps or insufficient locational signals slows market responses in short-term operation and longerterm investment. It is therefore potentially risky for reliability, particularly in the longer term.

20

EXECUTIVE SUMMARY

Empowering the Consumer.................................................. Vertically integrated utilities naturally focus on the supply side of the electricity sector, concentrating on the two pillars of electricity generation and transport. Until now, consumers paid the bill, and no infrastructure was in place to involve them in decision-making processes. Liberalised electricity markets introduce a third pillar that allows consumers to become active participants. Effective markets allow consumers to exercise their right to shift supplier, thereby enhancing competition for better services and increased innovation. Perhaps more importantly, consumer response to prices adds real resources to the system, potentially saving expensive generation or transmission investment and improving reliability. Finally, improved transparency from cost-reflective prices provides clearer incentives for more efficient energy use. This new third pillar is a product of the recent liberalisation process and, unsurprisingly, it has developed more slowly than the other two pillars. While the framework for consumer participation now exists, many of the detailed structures needed to facilitate ease of participation must still be further developed. A first building block to empower the consumer to participate is to create the necessary competitive pressure. Such pressure creates the incentives needed for retail companies to bring the opportunities of a competitive wholesale market to the doorsteps of consumers. Retail competition is based on the same principles as wholesale competition, with regulated access to the grid being the key factor. Unbundling of competitive retail activities from network activities is an important aspect of the process, but in most cases this phase of liberalisation has been less comprehensive than in transmission and system operation. Regulated access is provided by constructing systems and formal rules for consumer switching, but many markets still have small, but possibly decisive, barriers to switching – or still offer advantages to incumbent semi-integrated retail and network businesses. In all competitive markets, larger industrial consumers have switched in great numbers. The experience for smaller commercial and residential consumers is more varied, ranging from high switching rates in some markets to disappointingly low rates in others. In jurisdictions with liquid financial markets, more sophisticated retail products have been developed to better serve the needs of consumers who want to take an active role in managing risks. However, overall product innovation and development has been slow and sporadic. Anticipating that many small commercial and residential consumers will not switch retailers, some states and countries opted to introduce a regulated 21

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

default tariff or contract to protect consumers. While the justification for such regulated contracts is clear in a political context, they also often create distortions that may serve consumers poorly in the end. Establishing competitive retail markets that provide easy access to switching between competing retailers remains a challenge. Another effect of the somewhat slow development of competitive and innovative retail markets and the still often supply focused market design is the failure to bring wholesale market prices to the doorsteps of consumers. So far, there has been only limited opportunity for consumers to create benefits by shifting load as a price response. Considering that electricity is consumed by millions of different consumers for millions of different purposes, consumers are undoubtedly, in principle, willing to shift demand by varying degrees as a response to different prices. Demand is price-elastic: the challenge is to lower transaction costs sufficiently to justify participation for consumers who stand to realise the largest potential benefits. At present, a large barrier is in the hands of retail companies that must bring wholesale prices to the consumers’ doorstep. That being said, there must also be something to respond to: consumers cannot be expected to respond before prices rise sufficiently to off-set transaction costs. The largest consumers, who already have remotely read interval meters, are likely to be the first to see the benefits of shifting demand in response to price. Finding a way to take the wholesale price to the doorsteps of smaller commercial and residential consumers is, however, fraught with a technical and economic barrier given the absence of necessary metering equipment. Lack of demand participation remains one of the most serious challenges in liberalising electricity markets. The barriers are numerous. Creating easy and effective systems to manage retail switching is challenging. For small residential consumers, the infrastructure to enable switching is relatively costly compared to the potential benefits. In addition, it has been difficult to remove all distortions from semi-integrated networks and retail companies. Where governments show a willingness to intervene through price caps and other means, this also serves as a barrier to demand participation. Finally, lack of liquid financial markets makes it difficult to create the necessary innovative products. However, early evidence shows that consumers do switch suppliers and do respond to price when the conditions are sufficiently good. In fact, remarkably little demand response to price is necessary to significantly improve the performance of electricity markets, enhance system security and substantially reduce volatility and electricity prices for all consumers. 22

EXECUTIVE SUMMARY

Key Message Removing barriers to retail-switching and to active demand participation empowers consumers. Liberalised electricity markets provide a framework that can empower the consumer to exercise freedom of choice, to realise benefits from active response to prices and to consume energy more efficiently. Effective retail competition is a prerequisite to realising these benefits. A formal system must be in place to enable quick and easy retailer switching. Demand resources must also have easy access to wholesale, balancing and reserve markets. Opportunities to install metering equipment and infrastructure that serve the needs of different consumer groups should be considered, with due attention paid to costeffectiveness.

Efficient Incentives for Investment are Critical................... A substantial share of the electricity consumer’s bill goes toward financing generation and network assets. The opportunity to improve investment decisions is a significant potential benefit of market liberalisation. The ability of electricity markets to provide sufficient incentives for timely and efficient investment in generation plants continues to be one of the most debated aspects of market design. Many investment projects require long lead times and have an economic lifetime of several decades. The transitional phase of market development is characterised by uncertainty that may undermine the investment climate – and ultimately the successful transition to a competitive market. Investments in power generation are one of the big tests for the development of robust markets. Liberalised markets create a new investment paradigm in which decisions are taken under competitive pressure. When risks are shifted from consumers to decision makers, capital-intensive technologies with long construction times are viewed with greater scepticism – even if marginal costs are low. In the new competitive environment where risks are transparent, market players prefer technologies with short leadtimes that can be built in small incremental steps. Competition also pushes investment decisions to the last minute, which saves resources but can also put policy makers under pressure to intervene in a transitional phase (i.e. before the process has proven to be robust).

23

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

In situations where the supply and demand balance is tight, demand response to price can constitute the necessary buffer of last resources and add muchneeded flexibility. To date, a certain level of active demand participation has been critical in re-affirming the robustness of markets; conversely, lack of demand participation has laid the groundwork for very high price spikes needed to trigger investment. When governments have refrained from intervention and let prices reflect real costs, markets have come through – they have not failed to provide incentives for a response through investment in new generation capacity. In this context, so-called “energy only” (or, more correctly, “one price only”) markets, in which the wholesale electricity price provides remuneration for both variable and fixed costs, have performed well. Some markets have not shown enough confidence to rely on the delicate balance inherent in this new investment paradigm. These markets assume that consumers are not willing to participate and thus find that protective price caps are necessary as a consequence. However, with the barrier of a price cap, extra incentives must be added to prompt timely and adequate investment. These extra capacity measures have been implemented in various forms and have incentivised new investment. But they have also been prone to market manipulation. Another drawback is that capacity measures force decisions regarding the overall need for new generation capacity back into a centralised decision-making process. Investments in networks are, by and large, still being made within regulated frameworks. Distribution networks are still regulated, even though liberalisation reinforced the focus on cost-cutting in the whole value chain, including networks. The introduction of new economic regulation models prompted cost cuts but, in many countries, the focus is now shifting to quality aspects. Low investment levels and quality problems have directed the focus towards the necessary regulated rates of return to award investment and other incentives tied to quality. The development of regulated transmission networks is similar to that of distribution networks. Locational pricing in wholesale markets has, however, added crucial analytical input to enable more qualified investment decisions in transmission networks. In effect, transmission networks are an alternative to generation plants. Thus, in principle, transmission investments could rely entirely on locational pricing and commercial, competitive enterprise. In reality, the business model for merchant lines has proven to be fragile. Very few merchant lines are currently financed by purely commercial means, but locational pricing has still added substantial transparency to the process of making investment decisions in transmission. For example, several markets are developing information systems that enable a more co-ordinated interaction between decisions on regulated transmission investments and decisions on investments in generation plants. 24

EXECUTIVE SUMMARY

It is important to design markets and create regulatory frameworks that provide sufficient remuneration and incentives for efficient investment. But none of this makes any difference if investors cannot get permission to build. The absence of transparent and smooth approval procedures – whether to use a particular technology or to site a new generation plant or network at a particular location – continues to be a serious barrier to investment in most markets. This is not related to the liberalisation of electricity markets; rather, cost-reflective locational prices make the consequences of related environmental policies and the so-called “not in my back yard” (NIMBY) syndrome more transparent.

Key Message Minimising regulatory uncertainty is key to creating a framework for timely and adequate investment. Some amount of regulatory uncertainty is a fact in a changing world, but governments can take a number of measures to credibly minimise this uncertainty for investors. Investors in new generation capacity, particularly small market players, need access to market information, a well-established marketplace and regulated access to this marketplace. An investor will also require a transparent and clear procedure for applying for new investments, in terms of both siting and choice of technology. Any signal of government willingness to intervene in the market, including possible future support and possible price capping, will add uncertainty and deter investment. Finally, regulated network investment requires adequate regulated rates of return to maintain an acceptable quality of service in the long run.

25

INTRODUCTION

INTRODUCTION For more than a decade, market liberalisation has been a central policy path in the electricity supply industry in several IEA member countries. The movement to privatise and introduce competition in traditionally statecontrolled and regulated sectors came to the electricity sector at the end of the 1980s and early 1990s, with several IEA member countries pioneering the way with considerable success. Since then, virtually all IEA countries have initiated liberalisation processes. However, the path towards competitive electricity markets – the goal of which is to deliver real economic benefits – has not been clear. In many countries, policy makers, practitioners and academic experts continue to debate several key issues while other stakeholders voice scepticism and concern. Sceptics point to the crisis and market breakdown that hit California in 2001, only a few years after the new market was launched amidst much publicity. The crisis led to some very visible consequences for consumers, who were disconnected from the grid in rotating blackouts, and for electric utilities, which suffered major financial distress and, in some cases, were forced to file for bankruptcy. The collapse of the Ontario market shortly after its launch in 2001 is also singled out as an example of failed market reform. Another spectacular case is, of course, the bankruptcy of Enron. By being at the forefront of market analysis, trading and risk management, Enron mastered the emerging global electricity markets and successfully competed to become one of the largest energy companies in the world. Its foundation was, however, undermined by fraud, and alleged ruthless abuse of market power by the largest generators in California, including Enron, may have contributed to the collapse of that market. In the eyes of the general public, the large blackouts that hit North America, Italy and Scandinavia in 2003 are perhaps also regarded as examples of the failure of electricity market liberalisation. But extensive expert opinion and diverse research material now point to the root causes of those events, suggesting that the Californian crisis should be attributed to a wide range of factors – including both unfortunate coincidences and other more fundamental problems. For example, the “not in my back yard” (NIMBY) syndrome obstructed new construction work that was needed to meet demand. Much of the literature shows broad agreement on one point: the framework of the California market had important regulatory failures – on some very critical accounts. In essence, when policy makers negotiated the final market design, the regulatory framework failed to 27

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

properly account for some key principles linked to the nature of competition and economic incentives. The criminal acts for which Enron and other utilities now face charges only accentuated an already existing crisis. The alleged connection between the large blackouts in 2003 and electricity market liberalisation has also been explored extensively. Official investigations of the blackouts do not blame market liberalisation. The IEA points out in a new study that liberalisation has fundamentally changed the usage of transmission systems, and that the management of transmission systems had not properly adjusted to the new regulatory and market framework in the electricity sector. The study argues that, with appropriate adjustments in place, liberalised markets can create a framework that actually improves system security through increased co-operation amongst jurisdictions2. While a great deal of public attention has focused on these remarkable failures, several liberalised electricity markets operated with considerable success. In other IEA member countries, the failures put a halt to liberalisation processes and provoked second thoughts. But overall, liberalisation continues at varying speeds around the globe. This book examines the experiences of liberalised markets that have been operational for a number of years. None of these markets are perfect and all will continue to develop and evolve. But current results are starting to indicate the achievement of a robust framework. The IEA published a number of books on regulatory reform and liberalisation in the electricity sector during the liberalisation period, the latest being Power Generation Investment in Liberalised Electricity Markets3 and Security of Supply in Electricity Markets4. Until now, these reports focused primarily on analysing and describing key principles for successful electricity market liberalisation. The time is now right to change the focus and examine the knowledge and experience gained in solving key issues of electricity market liberalisation, as well as to tease out some of the lessons learned and some of the real-world market outcomes. The first serious attempt to form a liberalised electricity market was launched by Chile in 1982. Today, markets in several IEA member countries have operated relatively successfully for a number of years. England and Wales pioneered the way with the first market launch in 1990. The Nordic market started in Norway in 1991, with other Nordic countries (Sweden, Finland and Denmark) joining during the second half of the 1990s. In Australia, markets in Victoria and New South Wales began operating in 1994; the Australian 2. IEA, 2005c. 3. IEA, 2003a. 4. IEA, 2002b.

28

INTRODUCTION

National Electricity Market (NEM) followed in 1998. New Zealand liberalised around the same time, officially launching in 1996. In North America, various markets in the Northeast of the United States started operating in the late 1990s: PJM Interconnection in Pennsylvania, New Jersey, Maryland (PJM); the New England market; and the New York market. The California market opened in 1998. Texas in the US and Alberta in Canada opened their markets slightly later, in 2001. Additional markets are now opening across the European Union. Spain has operated a formal market based on regulated third-party access since 1998, as has the Netherlands. Many European countries report significant and growing trade over the past several years, with Germany taking centre stage. This book focuses on experiences in selected markets that were chosen for two reasons: a) being continually operational for the longest period of time and b) being representative of the main approaches to market organisation and design. The primary examples that meet these criteria are the markets in England & Wales, the Nordic countries, Australia and PJM. Additional lessons are also presented from other markets and overall conclusions are backed by the full range of experiences. Brief descriptions of the focus markets can be found in the Annexes of this book. However, the intent is not to describe certain countries, regions or market models, but rather to highlight different ways of addressing key issues. Although this book covers key issues for successful electricity market liberalisation, these issues are not presented in any order of priority. All items are crucial but there is a certain chronology not to be interpreted as emphasis. First, it is important to explore the main rationale for embarking upon a liberalisation process by assessing potential and experienced benefits. The second challenge relates to creating competition, the real fuel for successful liberalisation, and to understanding the roles of various stakeholders, including government. Without question, liberalisation is a fundamental change for the sector, particularly in terms of replacing the vertically integrated utilities with unbundled markets players and shifting decisions from a central body towards individual market players whose activities must be directed by transparent and cost-reflective price signals. Inherent in this transformation is the need to manage risk within this new competitive framework, which also puts the consumer in a new position. The next sections address issues associated with investment on two levels in terms of new power generation capacity (most markets face an initial phase of overcapacity) and the need for investments in networks, which seems to come sharply into focus after an initial phase in which the primary objective is to cut costs. And, finally, the book discusses the limits of liberalised electricity markets, particularly in the context of the need to ensure secure system operation and to account for environmental externalities. 29

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS Traditionally, electricity sectors have developed and operated within strictly regulated frameworks in which most or all activities – from generation to transport to distribution – are handled within vertically integrated utilities. Often, there is great government involvement, both in ownership and in determining the framework. The electricity sector was typically a centrally planned activity within which needs were assessed and fulfilled by electricity system planners and all the costs were passed on to consumers. In the late 1980s and early 1990s, a growing trend to liberalise traditionally public or regulated activities reached the electricity sector. In many countries, privatising state-owned companies is now an integral part of economic development. Within the electricity sector, the timing of the trend was motivated by inefficiencies in the sector itself and, to a certain extent, also by changing perceptions of the roles of public ownership, incentives, competition and markets. Technical development of generation technologies and – in some jurisdictions – funding problems accentuated the inefficiencies and the need for reform. Subsequent technical development enabled the management of increasingly complex systems; systems that now contain a range of the mechanisms required to support effective liberalisation of the electricity sector. In short, several factors drove the trend: poor performance of the existing systems, the growth of a “market ideology” and the declining costs of changing the system. The interaction and tension between these motives continues to drive much of the debate about benefits and costs of liberalisation in many countries around the world. Inefficiencies in the traditional, vertically integrated utilities differed markedly from country to country, but substantial overcapacity in generation seemed to be a common characteristic. This was even more evident when electricity demand growth slowed in many IEA member countries during the 1980s and 1990s. At the time, all costs – and therefore all risks – were borne by the electricity consumer, but the utilities themselves made decisions regarding investment. Hence, incentives to balance risks, costs and benefits were effectively separated from the decision makers and the decision-making process. This was particularly important when making investment decisions in terms of choice of technology, timing, capacity and location. The costs resulting from such investment decisions are, by far, the greatest share of the total costs of electricity supply. Thus, the potential benefits from improving the decision-making process for investments are also large. 31

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Improving decision-making processes for investment was important but improving actual operation also provided a rationale for liberalisation, specifically in terms of asset utilisation, fuel efficiency and labour productivity. Lack of involvement on the demand side proved to be another significant source of inefficiencies in the vertically integrated and planned sector. A constant dilemma and almost inherent characteristic of the centrally planned approach is a disproportionately high focus on supply. Customers are often thought of as “connection points” or “outlets”. There is no effective infrastructure in place to directly involve consumers in critical decisions that determine system design. Overcapacity and very limited product development are natural consequences of this environment. Some vertically integrated utilities developed “demand side management” (DSM) programmes, but such efforts did not offer the same dynamic interaction that is inherent in real markets. Freedom of choice of supplier adds value in its own right but, more importantly, it is also one of the main drivers for innovation and other dynamic improvements. Finally, a focus on supply within a specific jurisdiction and guaranteed passthrough of costs also remove clear incentives for effective and dynamic trade and co-operation across national and regional borders. For smaller jurisdictions, increased co-operation in a more integrated market provides strong potential for efficiency improvements; this is one of the most important sources of benefits from liberalisation.

Indicators of Success ............................................................ A natural place to begin assessing the success of electricity market liberalisation is to look at the prices actually paid by electricity consumers, including both industrial and residential customers (Figure 1). In all of the countries included in Figure 1, there is a clear trend showing that, after market liberalisation, electricity prices fell for industrial consumers. In the United States, this downward trend began in the early 1980s, coinciding with the emergence of independent power producers (IPPs) as a first step to introduce market competition. Development of US prices as an indicator of the success of electricity market liberalisation is blurred by the fact that not all states have liberalised. Pennsylvania is also included in Figure 1 to illustrate the development in one of the PJM states. Prices in Pennsylvania are above but converging to the US average. For household consumers the trend is less straightforward. There is a clear falling trend in the United Kingdom. In Australia, prices seem relatively stable after 1998, however, households are

32

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

Figure 1

Electricity prices paid by end-consumers: Fixed 2004 prices, excluding taxes USD/kWh 0,16 Industry 0,14 0,12 0,10 0,08 0,06 0,04 0,02

Australia

Finland

Denmark

Norway Sweden

20 02 20 04

19 98 20 00

19 96

19 94

19 92

19 90

19 88

19 86

19 84

19 82

19 80

19 78

0,00

United Kingdom United States Pennsylvania

USD/kWh 0,25 Households 0,20

0,15

0,10

0,05

20 02 20 04

19 98 20 00

19 96

19 94

19 92

19 90

19 88

19 86

19 84

19 82

19 80

19 78

0,00

Source: IEA

33

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

only contestable (i.e. free to choose a supplier) in a few Australian states and this contestability was put in place only in 2001. It is interesting to note that while there is a large convergence of wholesale prices amongst the states and territories within Australia’s National Electricity Market (NEM), such a trend is less evident for retail prices. In all Nordic countries, prices surged from 2002 to 2003 following a huge increase in wholesale prices as a consequence of a severe drought in the Nordic region. Retail prices are currently increasing in many, particularly European, IEA member countries. Increasing fuel costs and costs for CO2 emission permits in Europe seems to be important drivers but it also puts emphasis on the importance of a truly competitive framework during such periods with marked changes in market fundamentals. Comparing prices paid by electricity consumers before and after liberalisation may seem to be a very straight-forward indicator and, indeed, many countries promised that liberalisation would deliver decreasing prices. However, prices paid by consumers are not a simple reflection of the costs of producing and transporting electricity products. Traditionally, some consumer groups subsidised others. In addition, different parts in the value chain – from fuel recovery to generation and transport – are often subsidised in one way or another, or are not fully cost-reflective for other reasons. For example, in many countries the domestic coal industry has a long history of subsidisation. Substantial government support is also invested in development of various generation technologies (e.g. the nuclear energy industry). In countries with government-owned utilities, taxpayers have borne some of the risks and, therefore, some of the costs. Transmission and distribution networks contribute important components of total costs, but these result from developments in economic regulation rather than competition in liberalised markets. Another significant cost component comes from the fuel costs, such as coal and gas. Again, developments in fuel costs are not directly related to the performance of competition in liberalised electricity markets, even if currently increasing fuel costs stresses the importance of well-functioning and competitive markets for fuels. Perhaps most importantly, investment decisions made within a vertically integrated industry will influence costs for many years. Prices paid today are partly influenced by investment decisions made during the last several decades. If the electricity sector develops in new directions, driven by the re-allocation of risks and by other energy policies, consequences for prices will be clear only after one full business cycle passes. In the short term, it is more relevant to assess performance indicators of specific parts of the value chain, both in terms of the capital and labour resources required and the quality of service achieved. 34

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

Electricity generation is the most costly element of the value chain. In markets characterised by overcapacity and slow demand growth, it is common for markets that now face competition to withdraw some of the excess capacity. Capacity is either de-commissioned entirely or mothballed so that it can be brought into operation again at a later date. In Sweden, the ratio of installed capacity to peak demand fell from 35% in the early 1990s to 20% at the beginning of 2000. An immediate implication of liberalisation is that invested capital is put to better use. In New South Wales, part of the liberalised Australian market, the existing generation capacity (mainly coal fired) produced almost 12% more electricity in 2001 than in 1997, the year before the market launch. Other indicators, such as fuel efficiency, suggest that generation plants are operated more efficiently in liberalised markets. Fuel represents the main cost component in gas-fired generation plants and about one-third of the fuel costs for coal-fired generation plants (the fuel component is less important to nuclear energy costs; the investment component dominates). Fuel efficiency, in terms of the amount of fuel used per kWh produced, plays an important role in the total economic efficiency of a power plant. The fuel efficiency depends on how close a plant comes to operating at its optimum. Recent research examines the effects of market-based incentives on fuel efficiency in various US states, comparing data from plants where competition changed incentives against data from plants that maintained their status quo (i.e. no such restructuring initiative). The results show that fuel efficiency improved by 2% under competitive conditions. The study also suggests that private or public ownership did not play a role; built-in incentives seem to be the factor that matters most.5 Transmission assets are also being put to more dynamic and enhanced use, in part because liberalised electricity markets place greater focus on trade and exchange between jurisdictions. Smaller countries in particular reap important benefits of cross-border trade and co-operation. In addition, trade across country and regional borders improves utilisation of interconnection capacity and often even prompts investments to increase this capacity. Electricity trade between countries and regions has increased rapidly since the 1960s; within some specific regions, such as the Nordic countries, the rate of growth has been even stronger since liberalisation processes began. But perhaps more important is the fact that liberalisation brought fundamental change to the character of trade. Development of trade between Nordic countries is a good example. Norway liberalised its electricity market in

5. Bushnell & Wolfram, 2005.

35

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

1992; Sweden liberalised and joined the Nordic power exchange (Nord Pool) in 1996 and western Denmark became a partner in July 1999. An examination of flow over interconnectors between western Denmark and these two Nordic neighbours reveals dramatic changes in hourly energy flow from January 1995 to January 2000 (Figure 2). Figure 2

Flow over interconnectors between western Denmark and Norway, and western Denmark and Sweden MWh 800 Denmark West – Sweden 600 400 200 0 –200 –400 –600 January 1995

–800

January 2000

MWh 1200 Denmark West – Norway 900 600 300 0 –300 –600 –900 January 1995

–1200

January 2000

Positive numbers are import and negative numbers are export; the shaded area marks the rated capacity of the interconnectors. Source: Energinet.dk

36

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

Evidence from western Denmark suggests that market players used interconnections more extensively in 2000 than in 1995 (Figure 2). Many factors influenced the size of the flow in both time periods, thus selected months do not necessarily show the full picture and cannot necessarily be attributed to the effect of liberalisation and competition. What is important is that the figures clearly illustrate change in the dynamics of use. In 1995, contracts directing the flow seemed more regular and, hence, more predictable – even several days, weeks and perhaps months in advance. In 2000, it is more apparent that flow can change significantly from one hour to the next illustrating that transmission assets are put to more dynamic and flexible use. Efficiency across the region is improving as a consequence, but it also stresses some new challenges of managing the transmission system securely. Use of labour is another important productivity indicator. In this case, it is useful to compare trends in employment development. Electricity, gas and water sectors (Figure 3) is the sector category often used in standard statistical data compilations (data on Norway only include the electricity sector). The developments depicted may not all come from the electricity sector and some of the effects can result from out-sourcing various functions rather than representing a real reduction in the labour force. Figure 3

Index

Employment in electricity, gas and water (Index: 1990=100)

120 100 80 60 40 20

Australia

Finland

Denmark

Norway

20 02

20 01

20 00

19 99

19 98

19 97

19 96

19 95

19 94

19 93

19 92

19 91

19 90

0

Sweden United Kingdom United States

Source: OECD, Statistics Norway, Statistics Sweden, and Statistics Denmark 37

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

The data show a clear decreasing trend in the work force in all depicted countries. In the United Kingdom, the trend began right from the outset of the market opening in 1990 and is already evident in 1989 following adoption of the electricity act outlining market reforms (during previous years, employment was relatively constant). In Australia, employment decreases started in 1992, one year after the decision to reform the market and introduce competition was taken. In Norway, Sweden and Denmark, the decrease in employment really only started to pick up from 1999. This may reflect the fact that Finland entered the internal Nordic market in 1998 and Denmark entered in two steps in 1999 and 2000, thereby intensifying competition at the end of the 1990s. The trend underlines the important role played by the internal market for competition in Nordic countries. In Finland and the United States, the development seems to have been relatively smooth. In the United States, this could reflect the fact that the first step toward competition was introduced already in 1978 and that full-scale liberalisation was introduced in different states at varying points in time. This is supported by research suggesting that productivity of the US power industry was among the best in the world prior to liberalisation.6 A recent study on productivity in EU-15 member countries, conducted by the European Commission, concludes that labour productivity in electricity, gas and water sectors increased by 5.7% per annum between 1995 and 2001, up from 3.6% in the previous five-year period.7 All in all, it has taken fewer individuals to successfully generate, transport and sell more and more electricity – at a time when electricity output was rising, in some cases significantly. This increased labour productivity has enhanced overall efficiency in the sector. But there have also been important social costs in the wake of this development. Competition also improves transparency adding significant value in its own right. One side effect of the traditional vertically integrated model is that, in many countries, the electricity sector developed into a sector that serves many policy goals. As long as all costs could be passed on to consumers, collection of electricity tariffs often slowly evolved into thinly disguised tax collection. Revenues were used not only to cover the cost of generating and transporting electricity, but also to redistribute amongst different consumer groups and subsidise various activities. The net result is that the real costs of various policies become difficult to track, which may undermine economic efficiency – or at least obscure the efficiency of policy options. The parts of the sector now facing 6. Joskow, 2003. 7. European Commission, 2004b.

38

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

competition no longer have a guarantee that they can pass on all costs to consumers. In the new market, cross-subsidisation between consumer groups becomes difficult to manage and subsidies for other activities must be managed within the regulated network business. The experience from liberalised markets is that cross-subsidisation schemes become increasingly transparent and are often forced to end after a transition period. The costs of subsidies to other activities, such as support for renewable energy, also become more transparent. Information derived from increased transparency adds value in its own right; in many cases, it can lead to more effective energy policy adjustments. It is one thing to produce a product with fewer resources – and quite another thing to do so without losing quality. One way of measuring quality is to monitor the amount of electrical energy never delivered to consumers. But quality of service, in terms of number and duration of interruptions, has rarely been reported systematically according to consistent definitions. The United Kingdom began systematically collecting such data in 2001, coinciding with the introduction of a quality of service incentive scheme for distribution network operators.8 Initial figures show that quality of service – in terms of number of interruptions – improved 7% from 2002 to 2004. In Australia, legitimate concerns about quality of service arose when consumers – in some states, in some years – experienced more interruptions after liberalisation.9 However, there are many reasons to exercise caution in linking a quality indicator with the performance of liberalised markets. Final quality of service often depends on local networks: most instances of consumers being disconnected from the grid are due to mishaps in local distribution networks. Liberalisation has not changed local network activities. Networks are still regulated as in the past. Thus, quality of service is not really related to liberalisation except that liberalisation often prompts a greater focus on cost cutting in the economic regulation of local networks – which may ultimately lead to a lower quality of service. Finding ways to strike a balance between the short-term emphasis on cost-cutting without undermining quality in the longer run is a specific challenge for efficient, incentive-based economic regulation. Moreover, it is important to keep in mind that the effects of poor maintenance and replacement of networks will become evident only after a long time lag. Many markets are showing renewed interest in introducing specific quality parameters in their models for network regulation. For example, the United Kingdom, Norway and the Netherlands already introduced such elements in their regulation policies. 8. OFGEM, 2004. 9. ESAA, 2002 & 2004.

39

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Various studies seek to assess the overall benefits of liberalisation. In 1997, one group of researchers studied the economic benefits of liberalisation, and the distribution of those benefits, in the England & Wales market.10 They concluded that the total benefits correspond to some 5% of the total final electricity bill to consumers. The main source of the benefits was the shift of fuel from coal to natural gas. They also point out that the costs of restructuring were a significant element of the equation, which becomes particularly relevant when considering the costs and benefits of fundamentally changing a system as it happened in the England and Wales market in 2001. Several studies in Australia asses the benefits realised by the NEM. Securing Australia’s Energy Future, a white paper produced in 2004 by the Australian government, assesses the benefits from the NEM at AUD 1.5 billion annually.11 In a major federal review of NEM, conducted in 2002, the Council of Australian Government’s Energy Market Review suggested various changes and assessed that such changes would contribute additional benefits valued at some AUD 2 billion annually for the 2005-10 period.12 The Centre for the Advancement of Energy Markets (CAEM), a North American interest group, recently assessed the benefits of the internal PJM market, using cost reductions and efficiency improvements realised in the 1997 to 2002 period. In 2002, these benefits had a value of USD 3.2 billion, corresponding to roughly 15% of the electricity bill for that year. CAEM calculates the net present value of similar future benefits at USD 28.5 billion. In addition, benefits will be realised from improvements in dynamic efficiency that may arise over a longer time scale, resulting from new approaches to investment decisions, among other things.13 A recent OECD working paper summarised results of its work, The benefits of liberalising product markets and reducing barriers to international trade and investment: The case of the United States and the European Union.14 The research used econometric and general equilibrium analysis to assess the annual benefits in terms of GDP gains per capita. Investigators assessed only static benefits from increased trade and better allocation of resources, even though it is recognised that dynamic gains from increased innovation may add significant benefits as well. The analysis covers several regulated sectors, including the electricity sector. In fact, the electricity sector was singled out as

10. 11. 12. 13. 14.

Newbery & Pollit, 1997. Australian Government, 2004. Council of Australian Government Energy Market Review, 2002. CAEM, 2003. OECD, 2005.

40

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

the one in which the greatest government-imposed restrictions on trade and investment prevail – and, hence, the sector with the greatest potential for improvement. The paper uses Australia as the benchmark country for best practices in electricity. The total static benefits across all sectors are assessed to be 1% to 3% of GDP in the United States and 2% to 3.5% of GDP in the European Union, depending on the analytical approach applied. There is value in comparing the performance of vertically integrated and regulated utilities with the performance of competitive electricity markets as electricity sectors operate today. It is perhaps more relevant, however, to assess the performance of liberalised electricity markets in the context of electricity sectors as they are expected to look in the future. The level of preparedness for future needs is not captured in the static indicators discussed above. It seems likely that future electricity systems will change character significantly. Liberalised markets place greater focus on risks and capital costs. As a result, there is increased demand for less capital-intensive and more flexible generation technologies that can be built in smaller, incremental steps. The rapid development of the combined cycle gas turbine (CCGT) technology fits these new needs well and is a good example of how the sector is changing. At the same time, there is an increased array of energy policies designed to address environment and climate change challenges, which is already reinforcing a focus on smaller and more distributed technologies, such as wind power, and other distributed generation.15 At the other end of the scale, there seems to be renewed interest in very large nuclear power plants, driven particularly by the future costs of greenhouse gas emissions. On the demand side, new technology and the new market framework enable consumers to take part in the decision making process. All in all, it seems likely that future electricity systems will be more diverse in terms of technology and unit size, and will include many small and distributed resources. A future electricity system composed of a more diverse technology portfolio and exploiting the benefits of increased co-operation and trade calls for a very dynamic structure. The organisation of markets and system operation must give all resources an opportunity to interact with the electricity system in ways that ensure the best outcomes for consumers and owners of generation. This requires that all costs and benefits associated with each technology are allocated appropriately to create the correct incentives for a level playing field. It is likely that such a challenge for the future is best met by a market organisation based on incentives through competition. Allowing for interaction

15. The issue of distributed generation in liberalised markets is discussed thoroughly in an IEA study, IEA, 2002a.

41

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

amongst great numbers of market players. In a traditional vertically integrated and planned system, decisions are made centrally and any interaction from other players is possible only at the discretion of the utility itself.

Market Liberalisation is a Process rather than an Event ............................................................. Prospects for establishing market institutions that effectively serve their purpose depend largely on the point of departure within each country. In Australia, there were no institutions upon which to build the National Electricity Market (NEM); it was necessary to create a new national transmission system operator, NEMMCO, as well as all the other required market institutions. The launches of markets in New South Wales and Victoria in 1994 were important steps and final market institutions benefited from these first experiences. Still, eight years passed between the final decision to liberalise in 1991 and the market opening in 1998. The Californian market was also built upon entirely new institutions. In the north eastern United States, co-operation within PJM was already well established prior to the market launch. In Norway, the market organisation was based on co-operation amongst Norwegian utilities, a mechanism that had been in operation since 1971. The England & Wales electricity pool utilised management systems already in place within the public utility, the Central Electricity Generating Board. On the whole, the Norwegian, PJM and England & Wales markets opened with much less time needed for preparation. All of these markets underwent significant changes as they found themselves facing the challenges of new problems and new needs. In England & Wales, the entire market was fundamentally changed in 2001, eleven years after its initial launch. Australia undertook a major market review in 2002, which also led to a series of changes but no fundamental re-design. In the Nordic and PJM markets, problems and needs continue to arise at a constant pace, often stemming from issues related to co-operation between jurisdictions. There is an important lesson from this: the success of liberalisation does not depend on finding the perfect model from day one. Changes should be expected and appropriately managed. However, the costs of those changes, both for the market institutions themselves and for market players, become an important factor in further market development. Increased market transparency is now recognised as a very strong instrument to ensure continuous market development towards more competition, improved market design and more effective markets. Prices that are publicly – and easily 42

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

– available immediately reveal indications of market power abuse. Transparency in managing transmission systems congestion often discloses deeper problems; transparency in real-time system operation improves understanding of the real costs of that service. With good basic market structures that ensure independent system operation and independent governance, pressures arising from transparency can lead to changes that improve market performance. Analysing developments in the longest-running liberalised markets reveals various phases that are likely to be part of the liberalisation process. The first phase is preparatory: hard and careful political negotiations, legislation and market design are inevitable and even desirable. Technical infrastructure and management systems must be developed; institutions to operate them must be established. This phase can be shorter if structures that can be built upon already exist. A next phase is the development and maturing of the market. During the first years following market launch, some markets experienced problems with abuse of market power, which had to be addressed through changes to market design and legal action. Development of competition from new entries in the market is also common. And most markets also face some kind of crisis that ends up serving as a test of the robustness of the market. The consumer’s role in determining the fundamental structure of the sector is being reinforced – slowly but steadily. To date, no market can truly say it has passed beyond this transitional phase. It is likely that the next phase will be characterised by a continuing cycle of development, determined by the portfolio of generation plants. This might be called the ‘investment’ or ‘business’ cycle, during which markets may experience periods with excess capacity and periods with a tighter balance of supply and demand. It is also likely that there will be continuous pressure on market rules until a full business cycle passes (i.e. until old plants are closed and replaced with new ones). In reality, a full liberalisation process – from market launch to a robust and relatively stable market – will probably last at least one to two decades, or perhaps even as long as the economic lifetime of existing assets. From the preceding discussion of liberalisation experiences and market indicators, it is clear that it is meaningless to think of electricity market liberalisation as an event. It is difficult – impossible, even – to take a snapshot and use that as proof of success or failure. It is also clear that the level of political involvement required does not end with the decision to liberalise or with the completion of the legislative work to establish the market framework. Monitoring, oversight and decision-making by governments and regulators will continue to be a necessary aspect in determining future development of the electricity sector. As they have in the past, market players will anticipate this 43

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

factor and will react to the political, economic and regulatory risks it creates. Several examples show that at difficult junctures in market development a strong signal of political commitment can create the necessary market response. In contrast, other examples demonstrate how small amounts of political intervention, which was only intended to be transitional, undermined trust in the long-term political commitment to market operation. One case in point is the temporary price cap on retail prices introduced in Ontario, which aimed to shield consumers from high prices caused by a tight balance between supply and demand. The intervention added uncertainty, which deferred investments and tightened the balance even further. As a consequence, the necessary market response of adding new generation – which likely would have solved the problem – did not take place.16 Signals of willingness to intervene easily become self-fulfilling.

Distributing the Benefits of Liberalisation .......................... Liberalisation of electricity markets is much more than a matter of just finding the right model. Most of the really difficult struggles in the political negotiations associated with the liberalisation process stem from dealing with the tensions that arise amongst different interest groups. Many of the inefficiencies in traditionally vertically integrated systems accrue from benefits (rents) or subsidies typically received by these different groups. Such rents create losses to societies, but the recipients of rents and subsidies unsurprisingly regard the potential loss of the rent as a real loss. When benefits are spread across large groups and losses are concentrated on small but powerful groups, the political landscape for negotiations becomes even more difficult. This landscape is different from country to country, depending on the point of departure and on the fundamental characteristics of each electricity system. In the short term, the most obvious “losers” are those companies suddenly facing strong competitive pressure. This was not a big problem in countries where most utilities were state owned at the time of liberalisation. In England & Wales and in Australia, the competitive framework was relatively well known at the time utilities were privatised. The Nordic countries, New South Wales and Queensland, in Australia, still have large shares of public ownership. In many of the markets that are progressing at a slower pace, reluctance by incumbent utilities to effectively un-bundle system operations is an important barrier to development. In the United States, Spain and Denmark, efforts to recuperate 16. This example is described and discussed more thoroughly in an IEA publication on “Power Generation Investment in Electricity Markets. IEA, 2003a.

44

1 ELECTRICITY MARKET LIBERALISATION HAS DELIVERED LONG-TERM BENEFITS

so-called “stranded” costs also played important roles. Companies argued that they had made investment decisions based on a regulated framework – decisions that will not be profitable in a competitive framework and for which they need to be compensated. In PJM, recovery of stranded costs was one of the key reasons for introducing a “capacity obligation” that enables incumbent generators to secure remuneration for all generation assets built prior to liberalisation (discussed further in chapter 5). Spain introduced a complicated compensation system, the Compensation for Competition Tariff, which links compensation to the price level in the market – and has effectively distorted market functioning.17 Denmark compensated incumbent utilities through a lump sum paid over 10 years, funded directly by electricity consumers. Various parts of the electricity sector workforce represent another powerful group that stands to lose from liberalisation – at least in the short run. To the extent that the sector has been overstaffed or that the change of the sector calls for a change in the staff, some employees stand to lose their jobs. Increasing dynamism in the sector is generally seen as profoundly threatening to most staff. As a natural consequence, labour unions and other workforce representatives rarely support the liberalisation process. In many countries, different consumer groups also opposed electricity market liberalisation. The largest industrial consumers often have the most beneficial contracts with low rates as a part of industry policy. Residential consumers sometimes also receive preferential rates in the sense that even though they pay more than other groups, it is insufficient to cover the full costs of managing small customers. As a result, it is often the medium-sized industrial, commercial and service sector consumers that realise the greatest benefits from liberalisation. Large industrial consumers have often tried to argue for prolongation of preferential long-term contracts. In Norway, this segment of the market was offered preferential long-term contracts at the outset of liberalisation; these contracts are now maturing, which has triggered pressure on policy makers for renewals. Cross-subsidies amongst different classes of consumers, depending on their willingness to pay for uninterrupted electricity, are another dimension that comes to light with unbundling and transparency. The vertically integrated sector’s strong focus on the supply side may have been a preferable outcome for consumers who place very high value on uninterrupted and high quality electricity service. Often, for these consumers uninterrupted supply is critically important to their operations. In a vertically integrated system, such supply is

17. IEA, 2005b.

45

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

secured and the costs are shared by all consumers. In a liberalised market, consumers have the possibility to choose to consume and, therefore, also the possibility to choose not to consume if electricity prices rise above the value of the service it brings to them. This creates a situation in which consumers with the very high marginal value of electricity supply are left on their own to foot the bill. To date, these cross-subsidies have created only minor pressures but more firm reactions may be expected in the future, when the volatility of electricity prices increases. In some countries in which electricity generation technology benefits from indigenous natural resources (e.g. hydro or cheap fuel), a large part of the benefits from liberalisation derives from better exploitation of this comparative advantage in an open market and may include re-distribution from consumers in general to owners of generation plants. Liberalisation and open markets enable local generators to sell electricity to a larger market and enable a larger market to benefit from a superior technology. This improves the economic outcome for the larger market but the benefits in the local market may be concentrated on the electricity generator. In Norway, preferential contracts offered in the past to the largest electricity consumers were based on the availability of cheap hydro resources. With the opening of markets, these resources can now instead be sold at higher prices in neighbouring markets to the net benefit of the Norwegian economy as a whole and to the benefit of Norwegian generators - but maybe at the loss of competitiveness for large electricity-intensive industries. Many groups have important interests at stake in the liberalisation process. Pressure from these groups is a natural part of the political landscape and must be incorporated in the final package of solutions. However, if solutions to address social equity and re-distribution are allowed to distort the efficient functioning of a competitive market, a final positive outcome with net benefit for the economy may be jeopardised.

46

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS The goal of electricity market liberalisation is to create benefits by introducing incentives for higher efficiency and more innovation. Effective incentives are created by introducing competition between market players. Competition exposes market players to the risk of losing market share, or even going bankrupt, if they are not sufficiently efficient and innovative. But it also provides rewards for taking risks and performing better than one’s competitors. Failure to introduce effective competition can undermine the benefits of liberalisation in terms of lack of efficiency improvements and perhaps even deteriorating efficiency. In addition, abuse of market power has unacceptable consequences for financial distribution and equity. The philosophy of full electricity market liberalisation is to introduce competition and choice in as many parts of the value chain as possible – from generation to consumption of electricity. Various countries have applied different steps to progress through the different phases of the liberalisation process. In some countries, a first natural step was to open opportunities for competition between independent power producers (IPPs) and incumbent utilities. The United States Public Utilities Regulatory Policy Act of 1978 (PURPA) is one such example. Other countries introduced the concept of contestability in various steps – that is, giving electricity consumers the freedom to choose their supplier in different steps over time, depending on their level of electricity consumption. Competition can be introduced in most parts of the value chain, with a few exceptions. Local networks are natural monopolies and even though transmission networks could, in principle, be built in competition with generation, they are also still largely regulated monopolies. Transmission and distribution networks thereby remain, for the most part, regulated activities in which efficiency in terms of both investment and operation depends on the quality of economic regulation and the incentives it provides. The traditional structure of the electricity utility is a vertically integrated entity, which is often state owned. Competition is introduced by breaking up that structure into its constituent parts and care must be taken to identify those sections that can actively participate in competition and those activities that should remain regulated. As the unbundling takes place, the communication and co-ordination within previously vertically integrated companies must be replaced with other structures to enable and improve the necessary incentivedriven co-ordination that leads to efficient outcomes. The breaking up of 47

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

vertically integrated utilities and the establishment of new structures requires detailed structures for governance and solid legislative backing. Finally, when vertically integrated companies have been replaced with competitive markets, the level of competition must be monitored continuously.

Competitive Markets Replace Vertically Integrated Utilities................................................................ The first requisite to introduce competition is to separate or unbundle the natural monopoly network activities from all other activities. If an incumbent generator or a retail company maintains control over an affiliated network and can exclude or limit access to that network by competing generators or retailers, the network monopoly can be extended to an effective monopoly in the whole value chain. For this reason, transmission and distribution networks must be operated independently of generation and retail (Figure 4). While it might seem possible to slowly tinker with a vertically integrated industry in small steps, in fact, the very first step towards building a competitive market is the big step of deciding effectively to unbundled. Figure 4

A new model for the electricity sector

G

er a en

Coordination through vertically integrated monopolies Re ta il

48

n tio

sion smis Tran

Distribu tio

n

U N B U N D L I N G

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

On the transmission level, the main challenge is to ensure that all generators have equal opportunity to feed into the transmission grid and all consumers share the same ability to extract electricity. Because electricity cannot be stored economically, maintaining quality of service require balancing of consumption and generation in every instant. At present, there is no cost-effective technology available to achieve this automatically. The alternative is to establish a system operator that operates the transmission system and thereby balances supply and demand in the whole electricity system. Centralised system operation is a necessary natural monopoly, which must also be fully independent of generation, trade and retail. A central body responsible for system operation will manage the interface between the market and the actual physical outcomes. In fact, effective operation of the system must provide an indispensable part of the financial incentives in a liberalised market. By ensuring that the lights are kept on through balancing supply and demand in real time, a system operator delivers something resembling a public good. But the delivery of this public good relies on the rational actions by individual market players. They can only be expected to respond efficiently if their individual actions are committed in advance and thereafter metered to ensure fulfilment. Thus, the system operator must have full overview of generation and consumption, which can only be attained by establishing a very extensive set of rules regarding how generators, retailers, traders and consumers must interact with the electricity system. In the terminology of the European market directive, this is denoted as regulated third-party access. In other markets, such as the US and Australia, this is referred to as open access. Some of the most critical issues include rules for reporting and settlement of generation, trade and consumption schedules versus actual physical outcomes. There are, however, a vast number of subtleties implied in secure operation of an electricity system - so many, in fact, that it is impossible to prescribe rules for every necessary interaction. Therefore, system operators still maintain certain discretionary powers, regardless of careful efforts to regulate grid access. In addition, many of the detailed rules could only be developed once systems had been tested in real operation and real problems emerged. It becomes critical that system operators have the right incentives to detect these problems and contribute to solving them in a way that enhances competition rather than limits it. This has created a particular challenge in many of the countries in which the system operator is unbundled from an old vertically integrated incumbent. The first set of rules are often based on old habits and principles, which were designed in the past to meet the needs of the incumbent – and will, therefore, often favour the incumbent, at least to a certain extent. To ensure his contribution to developing a competitive market, the system operator must have incentives to work toward improved rules that 49

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

create a truly level playing field for all market players, with due consideration to the costs these changes may trigger. Co-operation between neighbouring system operators is also critical to the development of trade between neighbouring jurisdictions. Again, such a development is only likely to be realised if the involved system operators all have incentives to co-operate. Negotiations about the basic rules and principles for cross-border trade can take place between policy makers and regulators in different jurisdictions. But it is the involved system operators who must find ways to manage the actual day-to-day co-operation. Incentives to work for the establishment of an effective market with a truly level playing field and effective co-operation across jurisdictional borders are only ensured by truly independent and effectively regulated system operators. Liberalising markets have applied several approaches to the unbundling of transmission networks and system operations. While different markets have followed somewhat unique paths, each has had to address fundamental issues within two dimensions. One dimension is whether or not to keep transmission network ownership unified with system operation. The other concerns the form of unbundling in terms of ownership and level of independence. Regarding the first dimension, North America and Australia chose to form independent system operators (ISOs), but to let previous owners maintain transmission. In Europe, most countries kept transmission ownership and system operation together in the form of transmission system operators (TSOs). Both solutions have a number of pros and cons. With ISOs, in which system operation and transmission are separate, it has been easier to create truly independent system operators that cover several jurisdictions. Critically, separate transmission ownership also makes it easier to develop a competitive framework for transmission investment, especially considering that transmission investments can be alternatives to investments in generation. Perhaps most importantly, full separation – in which roles and responsibilities are settled through legislation, rules and contracts – tends to improve transparency. The challenge here is to create governing and regulatory structures that ensure the required co-operation between system operators and transmission owners to achieve secure and efficient operation. It has proven difficult to effectively allocate all responsibilities, particularly when it comes to the more detailed aspects of system operation, with congestion management being an especially challenging issue. In Europe, the argument for TSOs is based on the idea that transmission planning and ownership are integrated components of system and market operation. In this case, the challenge is to create governing and regulatory structures that provide transparency and efficiency in investment and 50

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

operation and that do not favour transmission investment over generation investment and vice versa. In general, Europe has encountered slow progress in creating robust markets that support efficient cross-border trade; this could indicate that a second challenge has not yet been met – that of creating sufficient independence and incentives for committed market development among European TSOs. In short, the pros and cons for ISOs and TSOs are that ISOs enhance transparency but perhaps at the cost of co-ordination; TSOs enable co-ordination but perhaps at the cost of transparency. Regardless of the approach used for system operation, it should be noted that transmission ownership is an important potential source for market power abuse. When transmission owners have large stakes in other parts of the value chain, incentives for investment in transmission may be distorted. In the second dimension on ownership and the level of unbundling, the approaches taken by IEA member countries are even more diverse. The range includes examples of both private and state ownership, as well as various levels of unbundling. The European Commission issues annual benchmarking reports on the development of the internal European market and the fulfilment of the European market directive.18 These reports introduce a terminology for levels of unbundling, classified according to ownership, legal separation, accounting or management. Unbundling by ownership and by legal separation are regarded by the Commission as being in accordance with the market directive.19 Table 1

Models for unbundling of transmission system operation Transmission System Operator (TSO)

Independent System Operator (ISO)

Transmission Owner (TO)

Unbundled by ownership – state owned

Denmark, Norway, Australia Sweden

New South Wales (AUS), Queensland (AUS)

Unbundled by ownership – privately owned

Great Britain, Finland

South Australia, Victoria (AUS)

Legal or functional unbundling – privately owned

Northeastern United States

Northeastern United States

18. European Commission, 2005. 19. European Union, 2003a.

51

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

There are strengths and weaknesses in each of these approaches described above (Table 1), yet they share a common element: they have all managed to establish a necessary level of independence from other individual players in the market – an independence that is rarely challenged. In the United Kingdom and Australia, as well is in PJM and Nordic markets, all system operators are now unbundled by ownership. It could be questioned whether state-owned system operators are effectively unbundled when governments still maintain ownership over large electrical utilities operating in the competitive parts of the sector, as is the case in the Nordic market. An argument for state ownership could be to ensure that the system operator’s main objective is to optimise social welfare. However, state ownership may actually make it more difficult to create clear incentives for efficiency and timely investment.20 With the establishment of independent system operators, the next challenge lies in ensuring necessary co-operation across jurisdictional borders to realise the full benefits of cross-border trade. Successful co-operation would also serve as an indicator of the level of independence of the co-operating system operators. Australian state and federal governments took the most radical approach to encouraging co-operation between system operators by forming one single, independent national system operator, NEMMCO. The PJM market has also developed according to the philosophy that market extension implies the expansion of unified system operation. System operators in the Nordic market are separate companies but have an association, Nordel, that serves as a forum to reach agreement on the rules directing trade and co-operation. Nordel has played an important role in supporting development of the Nordic market towards increased integration through harmonised rules. In Europe, system operators are organised through the European Transmission System Operators (ETSO), whose overall mission is to facilitate the development of the internal European market. Many of the operational aspects of physical cooperation in the synchronised continental European system are dealt with through the Union for the Co-ordination of the Transmission of Electricity (UCTE). In North America, the regional reliability councils that are members of the North American Reliability Council (NERC) play a similar role. When it comes to unbundling local distribution networks, the main challenge lies in ensuring independence in relation to retailers. An effective system is needed to support smooth and easy consumer switching of retail supplier, to put all retail suppliers on a level playing field. The European Commission uses the same standard for unbundling of local distribution networks as for transmission networks (i.e. ownership, legal, functional or accounting). Only a 20. Considerations regarding governance and regulation of system operation are discussed in a recent IEA study reviewing electricity reform in Russia, (IEA, 2005a).

52

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

few countries, such as New Zealand, require ownership unbundling of local networks. Proposed legislation for ownership unbundling of local networks is also being discussed in the Dutch parliament. However, legal or functional unbundling is required in the Nordic countries, the United Kingdom, PJM states and the Australian NEM states and territories. In several countries that do not require ownership unbundling, discussions continue regarding alleged cross-subsidisation and limits to regulated third-party access on a level playing field. Several management considerations provide reason to challenge the independence of local networks and, in particular, the need to manage consumer switching, historic load data and consumers who chose to keep the initial supplier. Unbundling is not enough to set the stage for a competitive market. While the general functions of the now-unbundled actors are well known, the detailed allocation of roles and responsibilities is critical to transforming an unbundled marketplace into a competitive one. The electricity system is operated through a number of action points and planning sequences over time, with due consideration to available resources, needs and constraints. These fundamental features do not change with unbundling and roles and responsibilities must be carefully distributed to ensure that all necessary actions will be taken in a timely manner. Some of the constraints, action points and planning sequences that are important in operating an electricity system are listed in Figure 5. There are several necessary planning sequences between the decision to invest in new generation, transmission capacity and fuel infrastructure and the final operation of the system and, hence, an ongoing need for co-ordination amongst stakeholders to ensure that all these steps are taken efficiently. Transparency and free flow of relevant information are critical instruments in enabling the necessary co-ordination.

Figure 5

Forecast

• Load • Generation • Transmission

• Load • Generation • Transmission

• Load • Generation • Transmission

• Load • Generation • Transmission

• Load • Generation • Transmission

Contraints

• G & T capacity • Fuel

• Fuel • Revision & maintenance

• Working hours • Computing • Co-ordination with neighbors

• Plant start-up

• Ramp rates • Reserves • Co-ordination with neighbors

• Reserves • Black start capacity

Action

Timeline for planning and operation

• Investment • Fuel contract

• Fuel purchase • Revision planning

• Day-ahead planning and optimisation

• Planning for operational reserves

• Balancing supply and demand

• Automatic regulation • Contingency management

1 year

1 day ahead

6-8 hours

15-2 years

15-1 minutes

Operation

53

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

As shown in Figure 5, the planning sequence starts several years in advance of operation. New generation plants and networks must be built to be available when needed and this must be planned as long in advance as it takes to design and build new equipment. In the shorter time frame fuel must be available for generation plants and existing equipment needs to be maintained: both activities must be arranged well in advance. Closer to the actual moment of operation, the system must start to prepare taking actual short term fundamental data into account. Electricity demand usually follows daily cycles that reflect the weather, the time of week and the season. Thus, all possible configurations of consumption and generation must be considered when analysing the transmission system’s ability to transport electricity. It is imperative to conduct a careful analysis of system security and optimisation, which takes time to compute. In addition, staff is limited and staff who must be called to work outside of normal working hours are particularly costly. Given the daily cycle in consumption and the constraints associated with computing and staff, a day-ahead planning and optimisation cycle is the natural choice for most systems. After the daily planning cycle, most decisions that will direct the bulk of the actions have been made. But many things can still change or go wrong. Resources must be available to manage any sudden changes in supply and demand. Many conventional thermal power plants take six to eight hours to start up from cold; nuclear power plants take longer and CCGTs can be operational more quickly. The start-up time often determines the last chance to respond to the need for system reserves. In minutes ahead of the actual moment of operation, supply and demand must be balanced; to safely perform this, it is necessary to consider the limits in ramping various generation technologies up and down. During actual operation the system must be prepared to respond to contingencies: to effectively deal with the unexpected, there must be reserves for frequency control and generators must have black-start capabilities. This list of action points and time sequences illustrates some key points in the planning and operation of an electricity system but it is far from exhaustive. The important fact is that these key points do not change with unbundling. Thus, all responsibilities must be allocated appropriately between the system operator and market participants; the tools for delegating responsibility include legislation, regulation and market design. The next challenge is to establish an organisational infrastructure that facilitates all the necessary actions to fulfil the responsibilities, while at the same time respecting the constraints. It is particularly important that all stakeholders know the exact responsibilities of the system operator. This is usually limited to the real-time system operation and to dissemination of information when it comes to the longer term. All other activities on the longer term are left for the competitive market, relying on prices to provide the necessary signals to act on. 54

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

Allocation of responsibilities differs somewhat from country to country. In the Nordic countries, such deviations pose challenges for optimal co-operation within the increasingly integrated market. Within its ongoing work, Nordel describes four responsibilities for TSOs that are common in the legislation in all the Nordic countries: ensure the operational security of the power system; maintain the momentary balance between demand and supply; ensure and maintain adequacy of the transmission system in the long term; and enhance efficient functioning of the electricity market.21 It is one thing to allocate responsibilities and establish rules that enable all the necessary co-ordination and actions. None of this will work efficiently, however, without also putting in place the necessary incentives for individual market players. There must be incentives for competitive market players to invest and to operate plants efficiently. There must be incentives for all market participants to report planned generation and consumption as accurately as possible and in a timely fashion. Incentives can also be used to prompt delivery of all the other services that the system operator needs. Establishing competitive markets for the full range of products necessary in electricity supply is the key measure to create an effective incentive-based framework. Organising a business by linking small companies together through contracts negotiated in open markets provides an alternative business model to operating within a large, vertically integrated company. Various economic factors often determine which of these two approaches is preferable within a given context. If large benefits can be realised by building large units and operating many units together, these economies of scale are best exploited in large companies. If the scale economies are less clear – and if the cost of establishing contracts that can mimic the structure of an integrated company are low – the market approach may be the best alternative. In short, low transaction costs make markets a possible alternative to vertically integrated companies. Keeping transaction costs as low as possible is critical to electricity market liberalisation; one way to do that is to ensure that markets are well organised and transparent. The structures for communication within vertically integrated companies must be replaced with structures that allow for the same level of communication and interaction between independent market players and the system operator. If incentives are clear and competition is strong, good market structures will ensure that all players put their best effort into carrying out their actions efficiently, and actions will lead to overall efficiency even when decisions

21. Nordel, 2005a.

55

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

are taken on this decentralised level. This is potentially a much stronger, more efficient and flexible – and, thus, more robust – framework than can be replicated in vertically integrated systems. In such traditional systems, efficiency depends on management of individual utilities and on the ability to use economic regulation to mimic incentives for efficiency. In addition, vertically integrated structures motivated by regulated incentives will adapt to changing circumstances at a pace determined by regulation, whereas a competitive framework will effectively incentivise quick and dynamic adaptation. One of the most important justifications for vertical integration is the need to manage risk. There may be benefits to gain from integrating several aspects of the market (e.g. generation, electricity retail, upstream fuel supply and natural gas retail) to manage the risk of the substantial capital commitments required. Hence, it is particularly important to replace the structures for risk management within vertically integrated companies; to create the opportunity for unbundled market players to enter into contracts, they must have the same possibility to manage risk. A liquid market for financial contracts offers a dynamic possibility for various players – generators, traders and retailers – to hedge risks at minimum transaction costs. The retail side is an area in which clear economies of scale seem to remain in many markets. A relatively large number of customers seem necessary to make a profitable business of supplying small commercial and residential consumers. This is a logical consequence as long as product differentiation in contracts with residential customers adds only little value to consumers and as long as there are barriers to switching retailer. In that case, a retail business offers only a very standardised product with low profit margins. The lack of innovation and product differentiation could, however, also be an indication of lacking competitive pressure, from which retail companies would be forced to offer more differentiated products. More effective competition in the retail segment is likely to provide incentives for increased innovation, product development and differentiation – all of which might enable smaller innovative retail companies to operate profitably. (These issues will be discussed further, particularly in Chapter 4.) Several markets, particularly the market in Great Britain, recently experienced varying degrees of vertical re-integration, i.e. generation companies have bought retail companies, or vice versa. In Australia, a retail company called Origin Energy moved to re-integrate by building new generation plants at various locations in the NEM. This re-integration raises concerns about the factors that drive such a development and the implications for competition. At present, there is no clear return towards vertical re-integration in the Nordic 56

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

countries. This may be explained by a variety of reasons but it is likely that the relatively well-developed and liquid financial market is one important factor. Some countries have taken a more voluntary approach to electricity market liberalisation, i.e. they let participants form the markets that they find necessary, without any detailed market design or formal rules established through a formal governance process backed by legislation. As all the functions of the value chain (from generation to consumption) must be maintained at all times, such an approach easily puts the incumbent in an advantageous position. Also, it is evident that unless the system operator has incentives to form and facilitate markets, the process develops very slowly. In Great Britain, Australia, PJM and the Nordic markets, system operators all hold the formal responsibility for operating real-time markets. PJM also operates a day-ahead market and the Nordic TSOs own the Nordic power exchange (Nord Pool Spot), which operates the Nordic day-ahead market. (It should be noted that Nord Pool Spot is separate from the Nordic TSOs but they interact very closely through information sharing.) Similarly, TSOs in these four markets have the formal responsibility to operate markets for reserves and ancillary services. All in all, the term system and market operation better reflects the real responsibilities and activities of these companies and the establishment of markets is the result of active regulatory initiatives. There are also examples of countries that have opted for a clear distinction between system and market operation. In Spain, the electricity exchange Operadora del Mercado Español de Electricidad (OMEL) operates the market and the system operator Red Electrica de España (REE) operates the system, even though very close interaction and cooperation between these two companies is required to make this approach work. Regulated transmission networks also play key roles in the market, even though they will not necessarily be incentivised efficiently through liberalisation. One of the challenges facing the regulatory framework is to find ways to incentivise transmission owners to maximise available transmission capacity e.g. when planning maintenance outages.

Legislative and Regulatory Framework for Effective Competition ..................................................... Active legislation, regulation and market design, established collaboratively by governments, independent regulators and independent system operators, play critical roles in the development of liberalised and competitive markets. Liberalisation requires the necessary legal framework and a targeted process, 57

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

launched by active government decisions. The intentions of government, as expressed in the legislation, then need to be implemented in a way that stakeholders can both predict and challenge. This requires a regulatory body that is independent of government. Many of the detailed market rules will directly influence the system operator’s abilities to operate the system securely. Thus, system operators often play an important role in establishing market rules, either within collaboration with the designated authority or as an important advisory capacity. The roles of different actors and the instruments they have used to create the framework for competition differ significantly from country to country. Government legislation to implement liberalisation ranges from relatively light legislation to a more detailed legislative framework. In Australia and United States, legislative development was strongly influenced by the fact that energy matters are, by-and-large, under the jurisdiction of state governments. In the United States, the Federal Energy Regulatory Commission (FERC) has jurisdiction over matters regarding trade across state borders. Apart from the 1978 federal energy act, the Public Utility Regulatory Policies Act (PURPA), which introduced the first level of competition, two additional pieces of legislation promote the development of liberalised markets: the Energy Policy Act of 1992 and the recent Energy Act of 2005. The Energy Policy Act of 1992 gave FERC the authority to order open access for wholesale transactions between utilities. FERC tried to promote further development of liberalised markets with this authority but achieving fully liberalised and competitive markets required additional state legislation. In the northeastern states, in California and in Texas, liberalisation brought forward through state legislation ordered partial or full retail access for consumers to switch supplier. In the United States, a large part of the framework for operating the electricity sector is decided by the state Public Utility Commissions and FERC. Within the PJM area, it is PJM that decides on all the actual market rules and design features that, in the end, direct electricity flow – but always within the framework of rules and regulations set by regional and federal regulators. FERC also approves various entities to act as Regional Transmission Organisations (RTOs), whose role is to provide non-discriminatory wholesale electric transmission service under one tariff for a large geographic area. PJM is one of the RTOs designated by FERC.22 In Australia, the Council of Australian Governments (COAG) took a decision in 1991 to introduce competition in the electricity sector. This led to the necessary legislative reforms in the States and Territories, which hold constitutional 22. Sally Hunt gives an excellent description of the development of the liberalisation process in the United States in Making Competition Work in Electricity (Hunt, 2002).

58

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

jurisdiction regarding unbundling and open access. Following its decision to liberalise, COAG established a supervisory body – the National Grid Management Council, which worked in a comprehensive consultative process with governments, the electricity supply industry and electricity consumers. The Council’s work led to the opening of the NEM in December 1998. The Australian Competition and Consumer Commission (ACCC; the federal body responsible for policing consumer and competition laws) authorised a new body, the National Electricity Code Administrator (NECA) to manage the electricity market code. Regulators in the country’s six states and two territories were responsible for the regulatory framework of the electricity sector. COAG published a review of the NEM in 200223, which prompted the formation of a federal energy regulator covering both electricity and gas. In 2005, two new federal bodies, the Australian Energy Market Commission and the Australian Energy Regulators, replaced NECA and state regulatory bodies (some 13 bodies, in all). Liberalisation in United Kingdom was triggered by the Electricity Act of 1989, which – in various steps – set the scene for privatisation, unbundling, thirdparty access to the grid, and retail contestability. The England & Wales market was reviewed in the late 1990s, which led to a major change in market design in 2001, also backed by the Utilities Act 2000. The British regulator, the Office of Gas and Electricity Markets (OFGEM), plays a key role in establishing market rules and developing the current market. In the original market set up, the system operator National Grid had a stronger role, building on the old structures for operating the system. The role of the regulator (then Office of Electricity Regulation - OFFER) was weaker and it was difficult for OFFER to introduce necessary changes. This was part of the motivation for the introduction of the new trading arrangements. Legislation in the Nordic countries also created the framework for unbundling, regulated third-party access and retail contestability. In the Danish case, legislation effectively implemented the European Union market directives; Norwegian, Swedish and Finnish processes were ahead of the directives and essentially in compliance with them. The distribution of roles and responsibilities varies from country to country but, in general, Nordic regulators have the formal responsibility for market rules but focus on the regulation of networks; system operators play a critical role in establishing market rules and developing market design. This role has been reinforced through the cooperation within Nordel, primarily out of a need to further enhance efficiency of the integrated Nordic market through greater harmonisation.

23. Council of Australian Governments, 2002.

59

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

New Zealand presents another interesting example of governing structures. Here legislation provided the framework for privatisation, unbundling and retail contestability through the Electricity Companies Act of 1993 and the Electricity Industry Reform Act of 1998. Subsequent development of market rules was left to a framework based on voluntary agreement, which allowed for the development of a sophisticated market place. However, a general failure to further develop the market, as was necessary in the beginning of this decade, triggered the government to establish a regulator, the Electricity Commission in 2003. This effectively put an end to the concept of market governance based on voluntary agreement. The rules and regulatory frameworks accommodate a variety of needs (Figure 5) and facilitate effective markets on three fronts: communication, market design and information. One set of very precise rules and regulations ensure smooth and effective communication between system operators, market operators, generators, retailers, consumers and traders. Additional rules concern scheduling and communication of bids. Rules related to market design and the market itself set the framework for trading and will, thus, direct all incentives. Rules on pricing principles and bidding requirements will direct all actions of virtually all players. Finally, transparency is a necessity for efficient markets, so there must be rules for disclosure of information. Market players, as well as system and market operators, must disclose the information needed to improve understanding of the fundamental conditions of the market. The Australian, PJM, British and Nordic markets all have very extensive sets of rules on all three of these key points. The nature of liberalisation is that all markets are in a state of continuous development. The actual experience of market operation provides impetus for some developments, which are designed to improve overall market functioning. In several markets, incidences of market manipulation and lack of transparency prompted changes. Other developments result from changes in the underlying electricity system, such as interaction across larger and larger areas in United States and Europe or the increase in wind power in Denmark and the United Kingdom. The inherent lesson is that governing structures of an effective and robust market must be able to manage change. Market players, investors and policy makers all want stable rules and regulation; in fact, continuous changes are necessary to further develop markets and to adapt to changes in the framework. These framework conditions require a predictable and standardised system for managing change. User groups and close interaction between system and market operators and market players are important means of ensuring this predictability. Once this framework is in place, adaptability to change becomes one of the market’s most important strengths. 60

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

There are some parts of the electricity sector that remain to be economically regulated after liberalisation, but with a reinforced focus on performance. System operation, transmission and distribution networks are more or less natural monopolies that will not be efficiently incentivised by competition. In principle, nothing has changed for these unbundled parts of the sector. They are monopolies as they were when being part of a vertically integrated framework, and their performance will, therefore, still depend on the incentives introduced by the regulatory framework. Experience from well-developed liberalised markets shows a renewed focus on efficiency, particularly of local network companies and of new models for regulation that have been developed. Great Britain was the first to introduce a new performance based method with inbuilt incentives for improved efficiency, in a regulatory system using a revenue or price cap that is allowed to increase with inflation but also requiring continuous efficiency improvements of a certain percentage (known as the CPI-X or RPI-X method). Others maintained the standard method of allowing the network owner to pass on all costs to the consumer - plus a certain rate of return (the cost-plus method). After an initial phase of high focus on cost-cutting within the regulated businesses, there is now increased focus on quality. Great Britain, Norway and Sweden reformed their regulations to include service quality and reliability as specific elements of incentive packages; the general principle is to reduce revenues if quality does not fulfil a certain benchmark. Spain introduced a system in which network companies must compensate electricity consumers suffering from lack of quality. Governing structures, in the form of legislation, rules and regulation, have been critical in establishing robust markets. However, there are also many examples that show the limits to these formal structures. In addition, the vast number of subtleties in system operation means that system operators will continue to have large discretionary powers: written rules simply cannot account for all possible situations. Regardless of the established framework, some severe situations may also create outcomes that are questioned and challenged for being unacceptable. There are several examples of electricity systems that faced extreme situations, which have often proven to be “the big tests” for markets. Extreme situations have always occurred from time to time, but probably less so in the past and with significantly less transparency. In traditional vertically integrated markets, it was easier to invest in excess generation and network capacity - and simply pass the full costs of such gold plating on to consumers. 61

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

From the launch of the Australian market in December 1998, there was a tight supply/demand balance in Victoria and South Australia. This led to very high peak prices, which raised political concerns over the functioning of the market. The South Australian government initiated an investigation to explore possibilities for intervention. The report concluded that intervention was not a good option and the government expressed its continued support for the market.24 (This case is described in detail in Chapter 5.) The Swedish electricity system faced a tight supply/demand balance during peak hours on two occasions in 2000 and 2001. During two particularly cold spells with extreme peak demand, there were fears of shortage. The situation prompted the government to request that the Swedish TSO, Svenska Kraftnät, submit a report outlining the situation and proposing solutions. Svenska Kraftnät stated that the market was intended to solve such situations and stressed the danger of government intervention. It also proposed various transitional measures that would have the least distorting effects on the market. The government followed the proposals and asked Svenska Kraftnät to implement the transitional measures.25 (These measures are described in greater detail in Chapter 5.) Another case in point is the Nordic market, which was struck by an extreme drought in the winter of 2002/03. Prices increased dramatically and the supply/demand balance was tight; fears arose of not having sufficient energy in the hydro reservoirs until they would start re-filling in the spring. The Norwegian government initiated an enquiry into the functioning of the market and presented its report to parliament, concluding that the market had handled the situation relatively well. Importantly, the Norwegian government showed continued faith in the ability of the liberalised market to deliver reliable, competitive supply.26 (This case is described in detail in Chapter 3.) In all of these cases, governments faced strong pressures to intervene because market opponents claimed the outcomes were unacceptable and ample proof of market failure. Government expression of its commitment to liberalised markets was likely critical to the healthy market responses observed in connection with these specific crises. More importantly, these signals of commitment provide credibility to the market for the future and thereby remove some of the political uncertainty that lingered in the minds of market participants. In 2002, the Ontario government intervened in the early stage of market liberalisation by capping retail prices. The intervention might appear to be of minor importance, but the signal of non-confidence it sent to the market 24. South Australia Government Electricity Taskforce, 2001. 25. Svenska Kraftnät, 2002. 26. The Ministry of Petroleum and Energy, 2003.

62

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

was critical. In fact, the Ontario Government came under strong pressure to implement a more severe intervention that essentially reversed the reform.27 These examples demonstrate the importance of signals: the absence of signals of government commitment – or perhaps presence of signals of a willingness to intervene – may be self-fulfilling and actually deter efficient market response.

Regulating Competition ....................................................... It is not enough to establish a formal framework that allows for competition through unbundling, regulated third-party access and retail competition. A second criterion for robust markets is the active participation of a number of market players, competing with each other. The actual level of competition remains a serious concern in many markets, due to high levels of market concentration. It is illegal to exercise or abuse market power and competition authorities in most countries have means to address the breaking of laws. In legal terms, abuse of market power is, generally speaking, a matter of determining that a company with a dominant position is abusing its place in the market to realise substantial and sustained profit. In electricity markets, proof of abuse of market power is difficult to obtain. Part of the problem stems from the fact that it is often difficult to define the relevant market and the relevant product in a way that will hold up in the relevant legal forum. Is the relevant market the jurisdiction or the integrated market? Was the relevant electricity product traded in each separate half or whole hour or is it traded in a yearly contract? With such problems in clearly defining the relevant market and product, it becomes even more difficult to define the notion of substantial and sustained profit. A related case came to light in Denmark in 2001. Danish market players complained to the Danish Competition Authority over alleged abuse of market power by the two dominating generation companies, Elsam and Energi E2, in 2000 and 2001. The Danish Competition Authority concluded it was not possible to prove that the two generation companies had abused their dominating positions. However, the both companies made an agreement with the Danish Competition Authority, committing to change their behaviour in the market. One element in this agreement was a commitment to act as a “market maker” that continuously provides bids to buy and offers to sell in the market within a certain price spread.28

27. IEA, 2003a. 28. Danish Competition Authority, 2003.

63

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Both competition authorities and regulators play critical roles in the development of competition in several markets. While it is difficult ex post to prove abuse of a dominating position, there are better opportunities to regulate competition ex ante in connection with mergers and acquisitions. This has been proven true in several cases within the European markets, the largest of which required approval from the European Commission. A much-publicised case was the mergers of German utilities that led to the formation of the very large utilities E.ON and RWE. These mergers were approved by the European Commission and the German Bundeskartellamt, with some important conditions including divestitures and some changes to the market principles.29 In response to recent increases in prices, rigidities in cross-border trade and high market concentrations, the Competition Directorate and the Directorate for Transport and Energy of the European Commission decided to conduct a sector inquiry into electricity market competition. The inquiry will focus on the functioning of wholesale markets and how prices are formed, including levels of market integration and the functioning of cross-border trade. They will also focus on relations between network operators and their affiliates to examine barriers to entry in the electricity market. Another instrument that has been used several times, and in various forms, is to require the sale of long-term contracts as an alternative to physical divestiture. In Europe, these contracts are called virtual power plants (VPPs); in other financial markets, similar products are known as option contracts. The buyer of the VPP obtains the right to draw electricity from a plant (or a pool of plants) according to the rules set in the contract. Contract forms vary but a common element is that the VPP is auctioned at a price that guarantees the right to draw energy at a predetermined energy price. The auctioned price corresponds to the option premium and the predetermined price corresponds to the strike price in the option contract. VPPs have been used in France, Belgium, the Netherlands and Denmark, in all cases as part of an agreement in connection with a merger or acquisition. VPPs increase liquidity in electricity markets and academic research broadly agrees that long-term contracts, such as VPPs, in electricity markets limit the possibility to abuse a dominating position. VPPs are, however, far from a real physical divestiture. The rules for all VPP arrangements stipulate that buyers must give one day’s advance notice if they wish to use the VPP. This excludes the use of the VPP from the real-time market for balancing power. VPPs are auctioned for a set period in the future: typically a month, a quarter or a year. If a dominating player has the power to raise the

29. European Commission, 2000.

64

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

wholesale market price, this will probably increase the value and the price of the VPP. It is thereby unclear how large a share of a monopoly rent a dominating player will lose in the end from this forced auction. Hence, it is unclear how effective this instrument is as a means to incentivise more competitive behaviour. The easy access for new entrants is crucial to liberalised markets – especially those that have only a few competitors and for which there is insufficient political will or power to change that through forced divestitures. A dominating market player may have incentives to delay investment in new generation capacity: withholding capacity and investment will increase prices. In contrast, new entrants enhance competition and their role in adding new generation capacity to ensure that supply can meet demand may be even more important. Smooth, timely, clear and transparent approval procedures for construction of new power plants are essential for competition and reliability of supply. Easy access for new entrants provides a particularly good opportunity to enhance competition, especially in countries where electricity demand is growing at a high rate. Spain provides a good example; competition is not yet satisfactory but new entrants are starting to gain market shares.30 If competition cannot be developed sufficiently within a market, the most robust way to boost it is to extend the market by integrating several markets. The FERC strongly promotes this philosophy in its efforts to encourage the formation of Regional Transmission Organisations across the United States. In fact, this was the key philosophy in North America during the formation of northeastern markets, such as PJM. Market integration to enhance competition is also a key element in the philosophy of Australia’s NEM. When Sweden decided to liberalise in 1996, one option considered was to establish a Swedish market that did not focus on market integration. This option was abandoned, in large part, because it was recognised that it would require breaking up of the large state-owned utility Vattenfall. The European Commission also sees market integration as the main path to a competitive European electricity market. This is underlined by the European Commission’s TEN-E project, which identifies priority projects to relieve critical congestion areas and points in European electricity and gas networks.31 Regulation of competition has another complicating dimension in many countries. In many European countries, the largest utilities are state owned or have close ties with government through other connections. The largest utilities in New South Wales and Queensland are also state owned. These companies are often recognised as national champions and often provide 30. IEA, 2005b. 31. European Commission, 2004a.

65

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

substantial revenue streams; hence these companies often create conflicts of interest for governments. In the United Kingdom, Victoria and South Australia, governments privatised their vertically integrated, state-owned utilities during the initial phase of liberalisation. Admittedly, it has been difficult for competition authorities ex post to prove abuse of a dominating position in day-to-day market trade. However, the threat of complaints and the force of publicity have been used with some success. The academic world, competition authorities and system operators have undertaken extensive modelling of market power abuse in North American, the United Kingdom, Australian and Nordic markets. Competition indices are one of the analytical tools used to screen for potential market power abuse. The Herfindahl-Hirschman Index (HHI), which sums up the squares of market shares of individual companies, is the most well-known measure of market concentration. In electricity markets, HHI has proven to be a complex tool that should be used with caution. In some circumstances, it may be most relevant to compute HHI by looking at produced energy, but in most circumstances it is more relevant to look at market shares in terms of installed capacity. It may be even more relevant to analyse the specific types of technologies owned by various market players. A large nuclear unit must run as base load for technical and economical reasons. A CCGT plant is more flexible and may be a more powerful instrument in the hands of a dominant player. The annual State of the Market Report, produced by PJM’s Market Monitoring Unit, is a very comprehensive market report that includes various HHIs as concentration measures, as well as analyses of the special circumstances in which one plant is a pivotal supplier.32 Analysing market power for the pivotal or residual supply in specific market circumstances capture many of the complex considerations important in electricity market competition. Both PJM and Nord Pool have independent market monitoring units with responsibility of monitoring and analysing trade to detect breach of rules that support market manipulation.33 The Danish TSO, Energinet.dk, issues monthly market reports that contain data and brief analyses, including comments on abnormal price developments. Nordic TSOs and regulators co-operate to model market power on a continuous basis.34 Transparency is a prerequisite for analysis and understanding of the fundamental market conditions. All the necessary information must be easily available, in a 32. PJM, 2005b. 33. Many aspects of analysing market power in electricity markets are discussed in Newbery et.al., 2004. 34. Nordic Competition Authorities, 2003.

66

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

timely fashion, for all market players and authorities. Easy access to basic market prices is a first step. These are all publicly available on the Web sites of PJM, the Australian system operator NEMMCO, the British system operator National Grid, and Nord Pool in the Nordic market. Transparency on market data in the British market is limited by the fact that the new trading arrangements, implemented in 2001, are based on purely voluntary bi-lateral markets. In the Australian and Nordic markets, all information about fundamental factors in the market must be disclosed immediately. A market player who signs a contract with Nord Pool to participate in the market is obliged to give immediate notice of all changes to the status of generation plants above 50 MW. These changes are immediately posted on Nord Pool’s Web site. Known as “Urgent Market Messages”, these notices can relate to outages, plant re-connections, changes in schedules for planned outages etc. Without this information in the public domain, it is possible for utilities owning large plants to trade on information that only the owner possesses. There are several examples in the European market that illustrate how withholding information has given a plant owner an advantage.35 In most financial markets such behaviour is called insider trading and is regarded as a serious criminal offence. Efficient outcomes in electricity markets rely on timely investment in new assets and efficient operation of these assets. Skills in building and operating plants may, however, add little value if they are not backed with an understanding of what role these assets will play in the total electricity system. Information management and analysis is an equally important part of the value chain as traditional asset management. The level of transparency in markets can be seen as the level of information flow that enables the necessary information management and analysis. In most markets the skills to manage and analyse information develop into a new value-adding business segment – known as simply “trading”, which was previously part of the vertically integrated utility although perhaps was not given the same weight. In the Nordic, British, Australian and PJM markets, traditional market players with generating assets or retail consumers have trading activities to manage these commitments and market positions. These markets also have market players who specialise in this segment, participating only in the trading role. Traders play important – and, in some cases, pivotal – roles in making markets more competitive and efficient. But they can only participate in electricity markets if they are able to manage their risks. In that sense, the presence of traders is also a good measure of the level of transparency and liquidity in the market. In theory, a fully transparent market should not leave room for traders; 35. Vilnes, 2005.

67

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

some regard this as the ideal, considering the controversy the role of traders often creates. In practice, markets have proven to be far more dynamic and ever changing for this to be realistic and traders continue to provide crucial liquidity and competitive pressure.

Box 1 . Development of competition in the England and Wales electricity market In the United Kingdom, the passage of the 1989 Electricity Act reflected the decision to unbundle and privatise the state-owned Central Electricity Generating Board (CEGB), which served England and Wales and to grant consumers above 1 MW open access to the grid. The first key steps of the restructuring took place over the next two years, with unbundling and launch of the trading arrangements in the so-called “Pool” in March 1990 and privatisation in February 1991. After unbundling, the sector in England and Wales consisted of a TSO, three generation companies and 12 regional electricity companies (RECs). The CEGB was divided into four bodies: National Power (50% of capacity), PowerGen (30%) and Nuclear Electric (20%) and the TSO, National Grid. All of the nuclear power plants were owned by Nuclear Electric. Originally, the intent was that the nuclear power plants should be part of National Power, which would then be large enough to absorb the business risks of owning and operating the nuclear stations. It turned out to be impossible to privatise National Power with these nuclear assets included in its portfolio, so they had to be withdrawn at a late stage in the process.36 From the opening of the market, Pool prices were lower than expected, primarily as a consequence of the structure of initial electricity supply contracts. Some of the factors in these contracts related to the obligation to buy domestic coal and the possibility to pass on costs to captive customers. Over the course of 1992-93 prices started to increase. In 1993 the regulator, Office of Electricity Regulation (OFFER), signed an agreement with National Power to divest 6 GW to a competitor – under the threat that their behaviour may be referred to the Monopolies and Mergers Commission, the competition authorities. In 1995, a “golden share” clause in the ownership of the 12 RECs lapsed, igniting a takeover and merger phase. In 1998 National Power and Power Gen each divested an additional 4 GW in exchange for approval to buy into the RECs now on the market. 36. Evans & Green, 2005.

68

2 COMPETITION IS THE FUEL FOR EFFECTIVE MARKETS

From 1991 to 1999, in what is known as a “dash for cash”, some 17 195 MW of CCGT was added to the grid, making up 23% of total installed capacity. By 2003, CCGT accounted for 33% of the installed capacity in Great Britain and 37% of the generation. These factors – the agreed divestitures, the takeovers and mergers, the construction of new CCGTs and additional voluntary divestitures – changed the structure of the sector completely. In 1990, the three largest generators accounted for 91.3% of total installed capacity while National Power alone owned 45.5% of the capacity. In 2000, the same three companies accounted for less than 50% of the installed capacity37; by 2004, they held only 37%. In 1997, the UK government asked OFFER to review the performance of the Pool. OFFER concluded in its review (published in July 1998) that prices did not fall as much as they should have, despite the increasing level of competition. Companies with a dominating position were still suspected of “gaming” the Pool, particularly that segment of the market relating to capacity payments.38 A decision was made to change the trading arrangements in the market to further enhance competition. The Pool was replaced by the New Electricity Trading Arrangements (NETA) in 2001. NETA was a fundamental change of market design; the key feature for enhancing competition was to change the trading arrangements from the obligatory central dispatch in the Pool to decentralised bi-lateral trading through NETA. The only formal market in NETA is the balancing market, which determines prices through an auction with discriminatory pricing rather than uniform auction prices used by the Pool. NETA’s discriminatory auction price means that accepted bids will be paid the bid price. This is fundamentally different from the pricing principles in most other electricity trading arrangements, in which prices are determined by the price in the marginal bid. Pay-as-bid pricing versus marginal pricing is discussed further in Chapter 4. Considerable theoretical and empirical research has been conducted exploring the NETA’s impact on the level of competition. One recent study asks “Why did British electricity prices fall after 1998?”39 Pool prices did decrease markedly from the end of 2000. Using econometric analysis, the study concludes that prices fell due to improved competition and capacity factors, rather than the introduction of NETA. The study strongly suggests that NETA did not change the behaviour of market participants. 37. Hunt, 2002. 38. See Annex 3 and Chapter 5 for further descriptions of the England and Wales capacity payment. 39. Evans & Green, 2005.

69

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

The most recent development in the structure of the British electricity sector is that there has been a vertical re-integration: large generating companies are acquiring retail businesses. With such a development, the focus of much research is now more on the effects NETA may have with respect to market transparency, transaction costs and liquidity. The main argument for re-integration seems to be a need for hedging in the generation business. Generation companies want a secure market for their product; in that regard a retail business becomes a so-called physical hedge. A physical hedge of that kind is very static and implies considerable transaction costs in connection with selling and buying retail businesses. Hence, the development could indicate that it is difficult to create the much more dynamic and cost-effective alternative in the form of liquid financial contract markets. Even worse, and of concern to several competition authorities, it could also indicate that retail can still be regarded as a semi-monopoly, considering that many smaller consumers do not search for a cheaper supplier on a regular basis. This creates an opportunity for vertically integrated companies to use the retail business to cross-subsidise the generation business, once again shifting the risks to retail consumers.

70

3 PRICE SIGNALS ARE THE GLUE

PRICE SIGNALS ARE THE GLUE Generators, traders and retailers come to a market place to trade and make contracts. They all have needs and obligations, which are fulfilled by committing trading partners – through contracts – to generate or consume a specific amount of electricity at a specific time in the future. Contract negotiations will spell out the conditions (time, place, volume etc.) and the compensation – the price. If the conditions and products are standardised and well defined, market players have only to agree on the price. Such a market is said to be effective; it keeps transaction costs at a minimum and price becomes the signal that directs all actions. In effect, the price becomes the glue that links the necessary decisions in an efficient manner and makes the sector work with as little friction as possible. The standardisation of electricity as a tradable product becomes the critical challenge for successful electricity market liberalisation, enabling the creation of a framework for transparent prices that signal real cost and the willingness to pay those costs. If the prices reflect the real values at stake and if incentives are right, the resulting actions will lead to efficient outcomes. Creating a framework that generates cost-reflective price signals to which market players can easily respond carries many inherent challenges. First of all, it is crucial to capture the true drivers of electricity costs and benefits, such as timing, location and volume of the delivery. The next challenge is to organise a market place in which market players and system operators can communicate prices based on these characteristics as smoothly and easily as possible, even though there will be important trade-offs between the level of sophistication of the pricing principle and the transaction costs of managing that pricing principle. Finally, in many markets important benefits are realised through trade across country and regional borders; in fact, cross-border trade may also be critical for effective competition in a specific region. Cross-border trade offers an opportunity for enhanced benefits; but it is also a constraint due to the need for co-ordination and harmonisation across borders.

Prices to Reflect the Inherently Volatile Nature of Electricity .......................................................................... By its very nature, electricity is a tricky product. It cannot be stored efficiently, yet consumers expect it to be delivered to fulfil necessary services in a timely, stable and reliable manner. In addition, consumption fluctuates from minute to minute and hour to hour and the conditions for generating electricity can 71

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

change at very short notice. Electricity consumption often changes with the weather and there is typically a daily cycle with very quick and marked variations between peak hours and off-peak hours. It is likely that these fluctuations will increase in the future, along with the increased use of air-conditioning and other electrical appliances in both residential and commercial sectors. In addition, supply and transport capability can fluctuate at short notice. Generation plants and transmission lines can face sudden problems that force them out of operation. For large generation and transmission units, one single outage can have a large impact on the balance between supply and demand. All in all, these factors make electricity inherently volatile and the volatility is not expected to decrease in the future. In fact, fluctuation of supply is likely to become more pronounced with increased shares of intermittent resources such as wind power. This volatility is an inseparable part of the characteristics of the service; it is not related to the organisation of the sector. That being said, the sector must be organised and aligned to effectively minimise and manage this inherent volatility. In order for market prices to provide incentives that lead to efficient outcomes, prices must reflect the fundamental factors of the traded product. For electricity, this implies that prices for products and services in the different parts of the value chain reflect immediate conditions in generating, transporting and consuming electricity. Thus, prices that reflect the inherent volatility of electricity should be expected to be correspondingly volatile. The boundaries of the volatility can go from very high prices to reflect a tight balance between supply and demand, to very low – or even negative prices – to reflect an excess of supply. Product prices in any liberalised and competitive market reflect considerations about marginal costs of supply and marginal benefits from demand. If competing with other suppliers, a supplier will be willing to sell his last produced product at the cost of this last marginal product – that is, at a price based on marginal costs. Extensive consideration and numerous experiments have tried to identify how this is best accounted for in the trading arrangements of liberalised electricity markets. A pricing and trading principle based on marginal costs means that the last accepted bid sets the price for the whole market. Some argue that such marginal pricing enhances the possibility of gaming between dominating players and has the potential to create windfall profits. Generators and consumers have no incentives to disclose their true marginal costs and benefits unless forced to do it through competition. One alternative considered, which seems to appeal to the philosophy used in fully regulated sectors, is to pay dispatched generators the amount they 72

3 PRICE SIGNALS ARE THE GLUE

initially bid (often referred to as “pay-as-bid” or “discriminatory” pricing). The standard critique of this approach is that generators will be forced to try to bid as high as possible, but still low enough to be dispatched. This forces generators to focus more on the marginal costs of their competitors than their own marginal costs, which is likely to prompt higher prices in the end. Most competitive electricity markets are currently based on a marginal cost pricing principle. One important exception is the British balancing mechanism, which uses a discriminatory pricing principle in which dispatched generators are paid what they have bid. It is important to consider how market prices relate to supply and demand in a competitive market with prices based on marginal bids (Figure 6). Some technologies, such as hydro and wind power, have very low short-run marginal costs; nuclear power often has relatively low marginal costs while other thermal power has higher marginal costs depending on the technology used. Short-run marginal costs, such as those that determine competitive prices (Figure 6), compensate only a fraction of the long-run marginal costs, however. For nuclear, hydro and wind power, by far the largest share of costs are investment costs. The overall profitability of these technologies is highly dependant upon prices determined by other technologies when these are needed to supply the demand. For example, a nuclear plant will only recover investment costs and profits during hours when the market price is determined by other resources with higher short-run marginal costs. Hydro power usually has a very different relationship with market clearing prices than the very simple relationships depicted in Figure 6. For hydro power owners, the cost assessment is not based on marginal cost considerations. Instead it is much more a planning problem in which the real costs are the alternative costs of using water today rather than saving it and using it at a later stage. The size of the reservoir is the constraint. The assessed “water value” determines bids rather than the short-run marginal costs. Hydro power plays a critical role in the Nordic market, with very high shares of hydro power particularly in Norway and Sweden. High prices – and even price spikes – occur when only very few generation resources in the system remain unutilised. These resources often have very high marginal costs. In addition, a generator with the last resource available has extensive market power and may consider bidding a price that is significantly higher than the short-run marginal cost. The profitability of a plant intended for peak load relies on payment in the very few hours of annual operation that compensates for both variable and fixed costs. Hence, plants intended for peak load rely on a certain scarcity rent to be profitable. 73

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Figure 6

Supply and demand in liberalised electricity markets Price

Hydro & wind

Nuclear

Demand off peak

Coal, gas & oil

Demand extreme peak

Demand peak

Demand price elasticity

Supply

MWh

If one market player owns the last remaining generation capacity in a certain situation, the price may be determined by other constraints such as imports from neighbouring jurisdictions, a price cap or consumers who are willing and able to change consumption patterns. The Australian, Nordic and British markets experienced short-lived but significant price spikes. In PJM there is a price cap of USD 1,000 /MWh, which is below the price spikes experienced in other markets. Price spikes and price caps have a crucial influence on investments, and demand participation (discussed more thoroughly in Chapter 5). Excess of supply resulting in negative prices can occur when a generator wants to avoid closing down a plant, even if its marginal costs are not covered for a short period. This can happen for thermal plants (such as nuclear and lignite fuel plants) in which start up is costly, or with wind power in which there is lack of plant control and a subsidy or some form of obligation that acts as an important part of the effective payment. Hydro plants may also be willing to operate 74

3 PRICE SIGNALS ARE THE GLUE

temporarily at negative prices, if the alternative involves bearing costs associated with spilling water. The Australian market has experienced negative prices for a few hours every year since market launch; prices in PJM market are also negative on a regular basis. In the Nordic market, negative prices are currently not allowed. However, Nord Pool announced that they will be introduced with new settlement software at the end of 2006.40 Prices have been zero for numerous hours in the western Danish trading zone, where there is a high concentration of wind power. During these hours, prices would have been negative if the practice was allowed. Electricity has a value to the consumer only if it is supplied at the right place, at the right time, in the right volume and at acceptable quality. The importance of these aspects is all directly related to the fact that electricity cannot be economically stored for later use. Quality is a standardised aspect of the service, ensured by the system operator. So far, quality is not an aspect that can be traded in the same way that the three other aspects are priced and traded. Volume and timing aspects are relatively straightforward. Ideally, electricity should be priced and traded for every Wh for every second. This is what matters to the consumer. On the other hand, only the largest consumers can be expected to show interest in participating in the wholesale market, so a certain degree of aggregation is practical for the wholesale market even if electricity consumption is measured by the Wh. Wholesale electricity prices are quoted by the MWh and minimum bids in wholesale markets are one or ten MWh. A certain aggregation is also necessary and practical when it comes to the aspect of time. In the second-by-second time span, system operators need a certain amount of capacity for automatic regulation. This generation capacity adjusts production according to a technical signal rather than an economic signal. Automatic regulation is only for the “fine tuning” of the system and results in a negligible share of total energy supply. So-called “real-time” markets are used to balance the system according to bids and offers in merit order. Balancing according to this normal economic dispatch makes it possible to meet variations between scheduled/actual generation and consumption, with a reaction time of five to 15 minutes. In most cases, energy delivered in real-time market is still priced and measured per MWh. Half-hourly or hourly pricing captures the bulk of the volatility in electricity generation, transport and consumption. The Australian and British markets are based on half-hourly prices; the Nordic and PJM markets use hourly prices. Longer term standardised contracts are also common, with delivery referring to a future

40. Nord Pool, 2005.

75

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

day, week, month, quarter, season or year. These standardised products use the half-hourly or hourly prices as reference. In short, considerations about timing and volume standards are trade-offs between capturing the fundamental factors that drive costs and benefits and the transaction costs of managing the standardised product. In some markets, final settlement with market players includes an extra penalty for imbalances. The system operator must consider the final, aggregate, net deviation from the scheduled generation and consumption. Most market players will have deviations that contribute negatively to the final net-deviation, but there will also be a few players whose deviations contribute positively to total system balance. In Australia, PJM and Norway, those players that cause the net imbalance pay to those that relieve it, regardless of whether they relieve it by chance or because they were called upon by the system operator in the real-time market. In the British, Swedish, Finnish and Danish markets, those that contribute negatively to overall imbalance pay but there is no reward for those that contributed positively by chance. This dual-pricing system aims to provide extra incentive to reduce imbalances, assuming that market players will not rationally weigh costs of imbalances with the costs of improving forecasting. One side effect is that it also creates extra incentive to be large and thereby reduce imbalances through aggregation. Aggregation does not contribute to reducing system imbalance per se, hence the effectiveness of the extra incentive for improved forecasting is questionable. Moreover, it undermines liquidity in this market segment and does become a threat to competition.

Locational Pricing ................................................................. The locational aspect of electricity pricing is the most controversial issue in establishing and developing liberalised electricity markets. As in all other markets, location is linked with transport service and transport costs money. In electricity, transport costs are significant. IEA estimates that, over the period 2003-30, some 13% of investments in the electricity sector within OECD member countries will be in transmission and 32% in distribution networks. The remaining 55% will be in new generation capacity.41 On top of the significant investment costs, the issue is further complicated by the fact that resistance in electricity networks creates losses, which also adds transportation costs. Electricity follows the path of least resistance ignoring any flow path that may have been intended with contracts. On any given transmission line, the 41. IEA, 2004b.

76

3 PRICE SIGNALS ARE THE GLUE

resistance and losses from transmission increase with the load. None of these relations are linear or constant, which makes optimising dispatch in economic terms highly dynamic and complex. The level of complexity is lessened somewhat in networks that connect single lines in a radial system to support generation and load centres. In contrast, complexity increases in meshed networks in which transmission lines criss-cross the transmission system, thereby creating several alternative flow paths. As an example of such “loop flows”, a contract between German and French market players will usually partially be delivered physically via Belgium. Prices that reflect the real costs of generating, transporting and consuming electricity must properly account for the costs of networks and network losses. Assuming that generation and load are evenly distributed across a radial transmission system with no congestion points, any generator would incur the same transmission loss. Thus, the optimal dispatch in such a system will always be the generators with the lowest marginal costs. In reality, no electricity system looks or behaves like this, but loss of economic efficiency can be minimised in systems with a radial structure. In a highly meshed network, however, the notion of efficiency and optimality may lead to fundamentally different results. Dispatch of the generator with the lowest marginal costs may, at a certain point, trigger grid losses that out-weigh its competitive advantage, thereby making a generator with higher marginal costs a cheaper alternative for the entire system. Dispatch of a generator with the lowest marginal costs may also lead to congestion somewhere else in the network, blocking access for other relatively cheap generators. If these considerations are not properly taken into account, it may force a very expensive generator into the dispatch, making the final outcome more expensive and less efficient. Ideally, electricity is priced for every location in the grid with due consideration to both transmission capacity and incurred grid losses. The pricing principle which is closest to this ideal is the so-called “nodal pricing principle”, which is applied, for example, in New Zealand and PJM. The philosophy of nodal pricing is that every transformer station in the transmission grid is a node and each node is priced, taking into account transmission capacity and grid losses (determined using computed loss factors). Nodal pricing properly prices all flows and constraints, including loop flows; thus it ideally creates full transparency and efficient incentives for investment and operation. However, many arguments are used for not applying nodal pricing, particularly that it would atomise liquidity in small shares, enhance market power, and increase transaction costs. It also raises issues of equity and distribution of wealth across regions and between consumers and generators often being the determining factor, particularly for policy makers. 77

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

The alternative approach is to apply pricing principles that ignore the locational aspect of electricity markets, or that address it only on a zonal basis. Zonal pricing aims to identify the main congestion points or areas in the transmission grid and then use them to form a group of nodes with uniform pricing. The Australian and Nordic markets could be described as zonal pricing markets. Each Australian state in the NEM forms a zone, as does the Snowy Mountain hydro plants region. The Australian system operator uses a nodal basis to account for network losses, computing loss factors for different reference points in the network. Dispatched generators are responsible for loss compensation. In the Nordic market, Norway is divided into three zones, Denmark is divided into two zones, and Sweden and Finland constitute one zone each. In this market, the network owner handles network losses; it is the owner’s responsibility to purchase the losses in the market and thereby the loss is appropriately reflected in the zonal price as a part of the demand. In the British market uniform balancing charges are applied for the entire system, even after the inclusion of Scotland in 2005. The pricing principle in the British market does not include any transparent locational signal. Network tariffs are another means of addressing locational aspects of pricing electricity. Tariffs for using the network can reflect the network costs associated with consumption or generation at a certain place in the grid. Such signals will normally be established on an annual basis, which is far from the dynamic, hour-by-hour pricing typically used in nodal or zonal pricing schemes. Sweden uses such locational signals in network tariffs. A more powerful locational signal can be sent along with the tariff for grid connection. For example, PJM uses so-called “deep network charges” in which the tariff for grid connection reflects many of the costs and benefits to the whole network with connection of a specific asset. Such deep network charges send very powerful locational signals, but it may be difficult to identify and allocate costs effectively and to manage them over time – e.g., when new generation or consumption takes advantage of existing networks. All markets experience congestion or other constraints that limit the possibility of operating the system according to normal merit order dispatch. Nodal pricing should reduce these situations but experience from PJM shows that there will be situations that force the system operator to dispatch out of merit order. In some areas, generators must operate to ensure system security, regardless of their position in the merit order. Zonal pricing schemes also apply out-of-merit order dispatch to relieve internal bottlenecks. If there is congestion within a zone, by definition it cannot be priced transparently nor can it be taken into account in settling market prices. As a result, the system operator may be forced to dispatch a more expensive generator on one side 78

3 PRICE SIGNALS ARE THE GLUE

of the constraint and remove a cheaper generator at the other end. Such a decision comes at a cost for the system operator: both generators will want compensation for the out-of-merit order dispatch (this type of activity is also called “counter trading” or “re-dispatch” in some markets). Often the system operator can use alternative means to manage congestion within zones. Instead of costly dispatch out of merit order, it may be possible to relieve internal congestion by limiting the available transmission capacity between zones and jurisdictions. This saves costs for the system operator but undermines the benefits of trade between zones and jurisdictions. Transmission capacity available for trade is a critical factor when incorporating the locational aspects into efficient electricity pricing. System operators might limit the transmission capacity available for trade at levels below the actual physical thermal capacity of the line. Often, it is a question of system security concerns: system operators continually model and analyse the system, assessing system security against a pre-determined set of criteria. When system analysis points to breaches of security standards, one option is to keep transmission capacity available for responses to contingencies. To a great extent, system operators use methods for analysis of system security similar to those applied prior to liberalisation. These methods are often overly conservative, they are rarely based on probabilities for critical events and they rarely exploit the information on costs and benefits that the competitive market framework can contribute. System operators are now under pressure to better align practices with the new market framework and to ensure maximisation of available transmission capacity. A recent IEA study discusses more thoroughly the issue of system security in liberalised markets.42 There is also increased pressure for better transparency in the process of determining available transmission capacity and, thereby, improved predictability for market players. The pressure for higher transparency is even greater when there is zonal or no locational pricing and the independence of system operators can be challenged. In the European market, the European Federation of Energy Traders (EFET) calls for greater transparency in the management of transmission capacity available for cross-border trade.43 A number of examples illustrate the fundamental nature of the locational aspect within electricity pricing while also demonstrating why it remains controversial and complex. Nodal pricing is the ideal reference. In fact, if the chosen option for the locational aspect is different from the principle of nodal pricing, price signals will be distorted to a certain degree. The level of 42. IEA 2005c. 43. EFET, 2004.

79

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

distortion depends on many factors but, in general, it will increase with the number of congestion points left un-priced. Often the options applied seem to be more influenced by political considerations – in terms of social equity and distribution – rather than the pros and cons related to market functioning and system operation. Non-transparent management of interconnections is a serious barrier to trade and competition. Shifting bottlenecks to jurisdictional borders and non-transparent management of transmission capacity restrict competition, and thereby automatically challenges the independence of the system operator, particularly when the system operator has close ties with the incumbent “national champion”. When the electricity market in Texas first opened (August 2001), there was only one control area or zone for the entire market. The agreed upon protocol was to manage congestion through dispatch out of merit order and to use a tariff up-lift to spread the cost amongst users. The target cost frame was some USD 100 million annually. In fact, during the first month of operation, congestion management cost USD 137 million. A decision was taken to divide the market into five zones (effective March 2002). Since then, expenditures on congestion management within these zones have remained significant at USD 200 million to USD 300 million annually. PJM market was also initially launched as a single price market but high costs for congestion management triggered development towards a nodal pricing principle after the first year of operation. Sweden is an expansive country with high concentration of load in the south and high concentration of hydro power plants in the north. It constitutes a single zone in the Nordic market. The TSO Svenska Kraftnät manages congestion within Sweden, typically through out-of-merit order dispatch or by limiting the transmission capacity available for trade across borders. This triggers pressure, particularly from Norway and Denmark, to introduce locational signals instead of limiting cross-border capacity, which they regard as a discriminatory practice. In 2003, the Association of Energy Retail Consumers (an association for Denmark’s largest industrial consumers) sent a complaint to the European Commission.44 It argued that Svenska Kraftnät discriminated against Danish electricity consumers by limiting the transmission capacity available for trade between Denmark and Sweden – with only 24 hours notice – due to internal Swedish bottlenecks. This practice allows consumers in the south of Sweden to access cheap hydro power from

44. FSE, 2003.

80

3 PRICE SIGNALS ARE THE GLUE

the north of Sweden and Norway at the expense of Danish electricity consumers. The debate is still on-going. In 2004, the Swedish Energy Markets Inspectorate (the Swedish regulator) published a report showing that all Nordic TSO apply the practice of shifting internal congestions to the country borders to a certain extent. A few intra-state congestions points exist in the transmission systems within the Australian market. Management of the Tarong constraint (southern Queensland) has generated much debate, particularly at times when congestions have been effectively moved to the border between Queensland and New South Wales. Even with nodal pricing, which aims to identify and price all possible constraints, creating efficient and transparent price signals remains problematic. In PJM, it is also sometimes necessary to opt for out-of-merit order dispatch for system security reasons, usually in locations where transmission networks are very congested. Such out-of-merit order dispatches incur a cost for PJM, which is spread across all users. However, the resulting locational marginal price (LMP) in the affected location is not set by the plant that was asked to run out of merit order. Such practice is one of the barriers to establishing price signals that provide efficient incentives for operation and investment.45 A recent study modelled the costs of not using locational pricing in the England & Wales market. When this electricity market is modelled with 13 nodes, rather than the current large single region, the study showed that pricing with 13 zones can raise total social welfare by what corresponds to 1.5% of total generator revenues. Moreover, the zonal approach would reduce vulnerability to market power and create better investment signals. However, it may also lead to regional gains and losses that may be politically sensitive.46 All locational pricing gives rise to revenues. Two neighbouring areas with different prices will see a net-benefit from trade. In general, these benefits will be on the consumer side in the high price area and on the generating side in the low price area. But distribution of the benefits will also depend on how the trade is organised. A first pre-requisite for realising the benefits is the act of linking the two areas with a transmission line (Figure 7). The owner and manager of this transmission line thereby becomes the “gate keeper” and acquires the potential power to collect the benefits.

45. Joskow, 2004. 46. Green, 2004.

81

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Figure 7

Electricity trade between two areas Low price area

High price area

Price

Price Supply

Supply

Demand Demand MW

MW

Transmission capacity

Transmission capacity linking two price areas or two nodes has an inherent value. The actor who controls flow over the transmission line will be able to collect what corresponds to the price difference between the two areas multiplied by the flow on the line. This stream of revenues goes by different names in different markets. In North America, and in theoretical literature, access to this stream of revenues is called a financial transmission right (FTR); in Australia it is known as the settlement residues while the Nordic region refers to congestion rents. Access to FTRs can provide two key opportunities. Market players can use it as a hedge to manage risks associated with price differences between areas and it can provide a price signal for efficient investment in new transmission capacity. How these possibilities are exploited depends on the management of the FTRs. PJM initially allocates FTRs to transmission owners, holders of old contracts and builders of new transmission capacity. The “owners” can then use the FTRs to hedge bi-lateral trade and self-dispatch or for trading. PJM conducts monthly and yearly FTR auctions. In 2003, PJM introduced a new tool for allocating the initial FTR according to earned rights, i.e. instead of allocating

82

3 PRICE SIGNALS ARE THE GLUE

the FTR that is specifically linked to a certain contract path, PJM allocates the annual revenue rights (ARRs) (which corresponds to the value FTRs sold for in the annual auction). An ARR holder can choose to transform it directly to corresponding FTRs if he wants to back a specific transaction or he can keep the ARR and perhaps buy other FTRs that better meet his specific needs. FTRs at PJM are firm, meaning that PJM guarantees the allocated capacity. If it turns out that the committed transmission capacity is not physically available, PJM must manage that shortfall through out-of-merit order dispatch. In the Nordic market, TSOs grant Nord Pool a monopoly to all available transmission capacity between price zones for day-ahead trade. Thus, all FTRs are allocated to Nord Pool, who collects the congestion rent. The transmission capacity is firm, meaning that the transmission capacity made available for Nord Pool can be taken fully into account in the day-ahead settlement of spot prices. It is up to the TSOs to manage any deviations between the transmission capacity made available for trade and the actual physical transmission capacity, usually through out-of-merit order dispatch. Nord Pool allocates congestion rents to Nordic TSOs, who own Nord Pool, according to a formula. Nordic TSO’s are bound by the European Union regulation on conditions for network access for cross-border exchanges in electricity according to which congestion rents can be used for three purposes: to guarantee actual availability of allocated capacity through out-of-merit order dispatch; to network investment; and as a part of the regulated income base, which must be taken into account in methods to regulate network tariffs.47 Nord Pool also offers a financial product, known as a “contract for differences”, that allows market players to hedge price differences between areas. In Australia, NEMMCO controls the transmission capacity made available for trade by transmission owners. NEMMCO conducts quarterly auctions for settlement residues, resulting from price settlements based on the transmission capacity that is physically available at the moment of operation. In that sense, the settlement residue is not based on a firm transmission right. The auction of settlement residue provides an instrument for hedging risks of price differences but some of the value of the instrument is undermined by the lack of firmness. Settlement residues and auction revenues collected by NEMMCO are used to lower transmission use of service charges. In the trading arrangements used in Australia, PJM and the Nordic market, transmission congestion is priced and managed simultaneously with the settlement of bids and offers from market players. The pricing of congested

47. European Union, 2003b.

83

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

transmission capacity is implicit in the settlement of market prices. An alternative approach is to make the auction of transmission capacity explicit. In reality, this is less efficient compared to the implicit and simultaneous pricing, which co-ordinates all aspects of the transaction. However, if a transmission line is not included in a uniform trading arrangement, the explicit approach can be a necessary, second-best solution. The German and Danish TSOs, E.ON Netz and Eltra, established an auction of transmission capacity over the Danish-German border, launched in 1999 when the western Danish system joined the Nord Pool trading arrangements. In the following years, other countries also established similar auctions along several other European borders, including the Netherlands-Germany and England-France borders. The European Commission’s benchmarking reports indicate that both implicit and explicit auctions are characterised as market based and, therefore, compliant with the European Union market directive and regulation. ETSO, the association of European TSOs, and EuroPEX, the association of European power exchanges, drafted a proposal for co-ordinating price settlement in neighbouring power exchanges. This would allow for coordinated action and thereby making congestion management implicit.48 Simultaneous price settlement, in which transmission congestion between neighbouring electricity trading arrangements is taken into account, has been discussed in Europe for several years. Such co-ordinated action between neighbouring power exchanges goes by many names but in the most recent reports from ETSO it is called “market coupling”. The proposal from ETSO and EuroPEX also includes a methodology to take into account – at least to a certain extent – loop flows in the highly meshed European transmission network, which is why they call the latest proposal “flow-based market coupling”. The approach focuses on cross-border trade between jurisdictions but does not address the need for congestion management within countries and control areas.

Box 2 . Nodal pricing at work in PJM PJM control areas contain highly meshed transmission networks. To meet the challenges inherent in meshed networks, PJM developed a trading arrangement based on locational marginal pricing (LMP), in which all relevant nodes are taken into account in the market settlement process. PJM serves 45 million people in 12 mainly northeastern states. Since its

48. ETSO & EuroPEX, 2004.

84

3 PRICE SIGNALS ARE THE GLUE

extensions in 2004, the entire PJM market area comprises 143 GW of installed capacity with a peak load of 115 GW. In that same year, PJM market participants generated, consumed and traded some 474 TWh of electricity.49 Currently, the LMP model includes 7 542 nodes (or buses), all of which are taken into consideration in price calculations and dispatch schedules. More than 6 000 nodes are with load; the rest are with generation. Buses (nodes) in the LMP model range from 765 kV to 4 kV level; almost half are higher than 100 kV, almost a third fall between 99 kV and 25 kV and the remainder are at lower voltage levels. LMP price calculations and dispatch schedules require input regarding thermal limits on all involved transmission lines - specifically, measures for the resistance on the lines, load bids and generation offers. PJM also requires information about start-up costs on all generators, which allow the LMP to account for the so-called “unit commitments”. Basic function of PJM price settlement is as follows: all bids and offers for day-ahead trade must be submitted to PJM before noon on the day prior to operation. PJM then uses the LMP model to calculate prices and dispatch for every hour for the following day. The dispatch aims to minimise overall system costs while properly taking into account bids, offers, thermal limits and grid losses. Congested transmission lines lead to price differences between nodes. The confusion of managing congestion within such highly meshed systems, compared to congestion management in zonal pricing systems, is that all nodes in the network – not only the two nodes connected by a congested line – may, in principle, be priced at varying levels. That is, a congested line between two nodes may lead to price differences amongst several other nodes, reflecting the consequences of loop flows. In real time, the LMP model runs every five minutes. The real-time market is accustomed to balancing any deviations from generation and consumption schedules. PJM charges imbalances with real-time LMP prices, i.e. hourly prices that are integrated over all the five-minute intervals of each given hour. Both the day-ahead and real-time LMP models include hundreds of additional rules that try to accommodate the various needs and characteristics of generation and system management. This includes rules that blur the final price signal somewhat, such as a price cap of USD 1000 /MWh.

49. PJM, 2005a.

85

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

There is no obligation to include all contracts in PJM’s LMP market settlement. Self-dispatch is allowed to fulfil bi-lateral contracts or to meet own retail commitments. In 2004, PJM had a market share of 35% of total load in its own control areas. Self-dispatch and bi-lateral contracts are included as must-run offers in the LMP calculation. In that way they are not relying on PJM to be dispatched but they are still subject to locational prices that reflects the costs of transportation. At the end of 2004, PJM comprised 34 control zones (divided into PJM midAtlantic Region and PJM Western Region), each covering a geographic area that was customarily served by a single utility. LMPs are based on a very complex model that accounts for the characteristics of more than 7 000 nodes, although the main congestion points in the market are concentrated on far fewer transformers, lines and interconnections. This facilitates the definition of trading hubs that comprise a number of important and normally un-congested nodes, and which ultimately serve as good reference prices. PJM calculates weighted average prices for three groups or hubs of specific nodes. The western hub is the most important reference point, with 111 nodes generating the largest share of liquidity in financial trading. There is also an eastern hub (237 nodes), as well as a hub comprising interfaces with neighbouring regions. Weighted average prices of several control zones are also used as references. PJM measures congestion costs from being forced to dispatch out of merit order due to congested transmission lines. In 2004, the market’s total congestion costs were USD 808 million, which corresponds to 9% of total PJM billing. The level of congestion can be measured in terms of number of hours during which a node is congested. In the same year, total congestion-event hours for the more than 7 000 nodes totalled some 11 205 hours; the most congested node experienced 1 784 congestionevent hours. Several markets, including the Regional Transmission Organisations (RTOs) for New York (New York ISO), New England (New England ISO) and the mid-west (Mid-West ISO), use similar nodal LMP price and dispatch calculations for their own trading arrangements. These four LMP-based RTOs serve approximately 75 million consumers across 26 U.S. states.

86

3 PRICE SIGNALS ARE THE GLUE

Markets for Ancillary Services ............................................. Three elements define electricity as a product for wholesale trade: standardisation in volume (MWh), time (hour or half-hour) and location (node or zone). Standardisation results from trade-offs between capturing the fundamental factors that determine the value of electricity and the transaction costs of managing the trade. However, some important aspects of electricity cannot be captured by this standardisation. For example, an aggregation in time does not give the system operator all the instruments required to operate the system at a satisfactory quality. The system operator will need reserves for automatic regulation and frequency control in the case of outages, for safe operation in the very short run. Supply of reactive power is necessary to maintain voltage levels. The network also needs black start capabilities in case of blackout. Reserves may also be needed to balance the system on a minute-by-minute basis in case of more dramatic imbalances due to outages of large generation plants or transmission lines. Such operational reserves could be considered necessary for various reasons, including because it usually takes time to start up a plant and in integrated markets, the entire generation capacity can, in principle, already be reserved for domestic consumers and exports. The necessity of operational reserves in competitive and liquid markets is debated and is also argued to be an instrument that blurs signals that would trigger appropriate market response. All these needs carry costs. The system operator must be prepared to pay for these services in order to maintain the incentives that prompt market players to supply them. Capacity reserved for ancillary services is not available for the ordinary market; thus, a capacity owner may incur a loss in supplying ancillary services rather than considering alternative sales in the market. At the same time, excessive payments for these services, particularly to dominating incumbent players, can distort the entire market. The approach taken in liberalised electricity markets is to define the needs for ancillary services as separate, specific products and to create mechanisms so that they can be bought by the system operator in a separate competitive market. The Australian, British, PJM and Nordic markets are all in the process of developing specific markets for ancillary services. They are continuously adjusting products to include as many sellers as possible, without losing the necessary characteristics of the service. Australia realised substantial cost reductions through the establishment of a market for frequency control. The Norwegian TSO, Statnett, established an advanced trading platform for weekly purchase of operating reserves. These reserves are a right that enables the system operator to draw 87

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

electricity under certain conditions; they resemble what is called an “option contract” in other markets. Statnett call their trading arrangement the “regulating capacity option market”. Statnett boasts one marked achievement from careful standardisation through this option market: it has managed to attract a significant number of bids from the demand side. In Britain National Grid has also managed to attract substantial volumes of demand resources in its various market segments for ancillary services. Denmark opened market-based purchase of ancillary services in 2003 attracting a few new market players. Since the launch of the market-based mechanism, products have been constantly adjusted to attract smaller, distributed generators to the market. The Swedish and Finnish TSOs have taken a somewhat different approach. They own gas turbines, intended to serve as operating reserves. In addition, Finnish Fingrid decided to build an additional 100 MW as a means of contributing to required upgrades of operational reserves in response to the construction of a new nuclear power plant which re-enforces their approach of ownership rather than contracting in a market.

Cross-border Trade Creates Benefits................................... Open trade is one of the classical merits of liberalised and competitive markets. Open trade across country or regional borders allows both countries (or, indeed, several countries) to realise mutual economic benefits by finding and exploiting comparative advantages in the division of capital and labour. Electricity generation and transport include many factors that relate to resource endowments, geographical characteristics and national skills; it is also very capital-intensive businesses. Thus, there are many reasons to look for and exploit comparative advantages and many ways to realise large potential gains by optimising the use of assets across as large an area as possible. Clearly, for many countries, cross-border trade is an important source of realising benefits from electricity market liberalisation, particularly for smaller countries with geographically close neighbours. In many situations, cross-border trade is also the easiest and quickest way to enhance competition. If it is not possible to create competition within a market, it may be achieved by enhancing the market itself. Again, the benefit is even larger for smaller countries in which economies of scale may limit the number of companies that can operate profitably. The prospects of extending markets are, of course, limited by the availability of transmission capacity. Ideally, price signals on both sides of a congested transmission line should be used as a signal to invest in both generation assets and transmission assets. Transmission and generation assets are substitutes to 88

3 PRICE SIGNALS ARE THE GLUE

a certain extent, reinforcing the notion that control over inter-connectors is one of the most powerful roles in an electricity market. The “gate keeper” in any given market acquires access to congestion rent and potentially controls the level of competition allowed. At any given level of transmission capacity, congestion rent will be reduced if transmission capacity increases. However, a capacity owner may not have the right incentives to invest – unless there is effective competition in the market. As a consequence, the independence of this “gate keeper” is essential to a competitive liberalised electricity market. The business models for PJM and Australia separate system operation and transmission ownership. Independent system and market operators ensure that all congestion is priced and that transmission needs are transparent. There are two paths to transmission extensions. One follows competitive merchant lines, financed through congestion rents: electricity is bought by the transmission owner at the cheap end and sold at the expensive end. This model will be profitable only if a certain price difference can be maintained after the connection has been established. Optimal inter-connector capacity depends on finding a mechanism that just maintains a level of price separation capable of financing the line. The second pathway relies on enhancement through reliability requirements, using regulated tariffs to finance the extensions. Competitive merchant lines have not been successful thus far so most investment still rely on financing through regulated tariffs (see Chapter 5 for further discussion). The transmission business model, used mainly in Europe, in which transmission ownership and system operation are kept together may offer better scope for co-ordinated planning of transmission lines to fulfil both reliability and trading requirements. However, with a monopoly on transmission ownership, private commercial incentives may not lead to efficient enhancement of transmission grids. Incentives will be distorted, as maximisation of congestion rents does not imply an optimal level of transmission capacity. Rather, transmission capacity is optimal when congestion rents just pay for the costs of the transmission line. Fear of distorted incentives is one of the main drivers behind the European Union’s efforts to promote investments in new transmission lines relieving serious congestion points.50 TSOs in the Nordic market are moving towards a more holistic approach to transmission planning and investment. Nordel conducted a series of studies of the main congestion points and flow paths in the Nordic market, which identified five prioritised transmission projects. The Nordic TSOs agreed to proceed with a joint investment project worth EUR 1 billion.51 The investment 50. European Commission, 2004a. 51. Nordel, 2005b.

89

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

package was developed with a reference to the net economic benefit of the entire Nordic market, rather than local gains and losses. The decision is backed by Nordic energy ministers but will be financed by grid users through tariffs (see Chapter 5 for further discussion). In Europe the focus is not only on new investment. Only half of the 34 countryto-country interconnections between ETSOs 24 member-countries are allocated according to market-based principles.52 Only the six Nordic crossborder interconnections are allocated according to the fully co-ordinated market coupling trading principle. Investment in new interconnections seems less important if existing cross-border capacity is not used efficiently to exploit the benefits of trade through fully dynamic trading arrangements. Cross-border trade does more than create benefits through day-to-day trade. On a broader scale, trade and co-operation also create a more efficient, robust and secure system – particularly by joining forces to meet operational challenges through sharing reserves and other ancillary services. This joint effort to manage situations is particularly valuable to smaller systems, and particularly in circumstances of low probability but with serious consequences. Market integration in Australia and PJM also included integration of management of reserves and markets for ancillary services. In the Australian market, NEMMCO reduced the aggregate minimum reserve levels from 2233 MW in 2003 to 994 MW in 2004 – more than 50% reduction. This reduction is mainly attributed to the ability of inter-connectors to share reserves between regions and to differing demand patterns between regions.53 Trade across jurisdictions has also reduced the aggregate need for reserves in PJM. In 2004, several steps were taken to extend PJM market area; summer peak demand increased 30% as a result of the extension. However, demand for spinning reserves (for which there is a market in PJM) increased only 20%, reflecting the value of increased co-ordination. Electricity systems in Europe maintain a long tradition of co-operation and coordination regarding the use of reserves and other ancillary services, primarily through agreements within UCTE and Nordel. Markets for reserves and ancillary services continue to develop in all the Nordic countries and in several other European countries. There is one example of trade with reserves across borders: In 2003, Eltra, the western Danish TSO, bought operational reserves in Norway in agreement with the Norwegian TSO, Statnett. However, the increasing level of co-ordination still does not enable real reduction of reserves. 52. ETSO, 2004. 53. NEMMCO, 2004.

90

3 PRICE SIGNALS ARE THE GLUE

Box 3 . Nordic market: the Autumn the rain never came In the autumn of 2002, the rain that usually falls in Nordic countries never came. Up until that point, it had been a wet spring and summer – particularly in Sweden. The picture began to change in early autumn as Sweden experienced unusually low levels of precipitation. A few months later, a similar picture emerged in Norway. In both countries, reservoir levels, measured as deviations from the normal level for each specific week, fell significantly (Figure 8). By the end of 2002, inflows to the Nordic hydro reservoirs in Norway, Sweden and Finland were 35 TWh below normal, which corresponds to 9% of total Nordic consumption in that year. (In Norway the situation was particularly bad: inflow was almost 18 TWh below normal corresponding to almost 15% of national consumption.) The droughts themselves were unusually severe in these three countries, but it was even more unusual that such severe droughts would actually coincide. A study referring to historic data assessed the probability of such severe droughts coinciding to 0.5%, or once every 200 years.54 In addition to the autumn drought, early winter was unusually cold and one of Sweden’s largest nuclear power plants was out for refurbishment. The Nordic region, and Norway in particular, faced a severe energy crisis that had to be managed at every instant – even in the context of an uncertain future (i.e. without the clear overview that hindsight now offers). As indicated earlier, both Norway and Sweden were coming from a very wet spring and summer. As autumn progressed, it was valid to assume that the most likely future precipitation would be at normal levels. They did not normalise, and precipitation during the winter remained at very low levels. The first months of 2003 turned out to be relatively mild and the snow began melting relatively early, easing the immediate pressure to the system. However, Nordic hydro reservoirs only recovered to more normal levels in early 2005. As a result of this unusual chain of events, electricity prices increased to unprecedented levels – culminating with a weekly average price of NOK 752 /MWh (EUR 104 /MWh) in the second week of 2003 (Figure 8). Nord Pool spot prices subsequently remained at relatively high levels through 2003 and 2004, and the market responded to the prices on several fronts. The most significant response to the energy shortage was substantial 54. ECON, 2003a.

91

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

production increases from thermal power plants in Sweden, Finland and Denmark. Several thousand MW of old, mainly oil-fuelled power plants were de-mothballed. Output from Nordic thermal power plants (oil, gas and coal) increased by 9 TWh in second half of 2002, compared to the same period in the previous year. In the first half of 2003, the increase doubled to 18 TWh compared with the same period of the previous year. Imports to the Nordic region from neighbouring countries were the second most important source to relieve the tight situation. In the second half of 2002, the net import to the Nordic region was 5 TWh; in the first half of 2003, it was 10 TWh. From June 2002 to July 2003, almost 10 TWh was imported from Russia alone. A final significant response from the market came in the form of reduced consumption. According to a recent study, temperature-adjusted consumption decreased in Norway by roughly 3 TWh from October 2002 to February 2003, which corresponds to some 5%.55 In Sweden, temperature-adjusted consumption for the same period decreased by roughly 2 TWh. In Norway, the main reductions were from large industrial consumers, but households and electrical boilers also contributed. Electrical boilers have traditionally a large share of the heating market in Norway, particularly within process heating for industry. However, many of these electrical boilers can switch to alternative fuels, making this market one of the most straightforward fuel-switching opportunities in the country, even though only a part of the potential seems to have responded to price increases. The reaction from large industrial consumers in Norway was also influenced by a general economic recession. In Sweden, the main reaction also came from large industrial consumers. Marked differences in household responses in the two countries – Norwegians reacted by reducing consumption, Swedes did not alter behaviours – can be quite easily explained. In Sweden, one- to two-year contracts are the norm for residential electricity consumers; thus, price increases in the wholesale market had no immediate impact on household bills. In contrast, a large majority of Norwegian households hold short-term mark-to-market contracts that are adjusted according to spot prices; wholesale price changes can be reflected in individual bills on very short notice – as little as one to two weeks. (Consumer response during the Nordic drought is discussed further in Chapter 5). 55. ECON, 2003b.

92

3 PRICE SIGNALS ARE THE GLUE

As the tightness of the Nordic energy balance intensified during the autumn of 2002 – and prices increased – the situation attracted more and more public attention. The situation awakened a broad Nordic debate about resource adequacy, diversification and market functioning. In Norway, the debate was broader and more intense than in the other Nordic countries, leading to significant pressure on the government to intervene. The high percentage of households with short-term contracts created a clear and fast linkage between current spot prices and prices charged to household consumers, thereby giving the debate a social dimension. One main point of criticism was that Norway continued to export significant amounts of electricity, even as the drought intensified during the last months of 2002. Effectively, Norwegian hydro producers emptied their scarce hydro resources to supply Swedish consumers. The development of the net export from Norway to Sweden, week by week, is shown in Figure 8, together with the weekly average spot price and the deviations from the normal reservoir levels in both countries.

Figure 8

Reservoir levels in Norway and Sweden, trade between Norway and Sweden, and Nord Pool spot price, 2002-2004 Price: NOK/MWh

Trade: GWh 30%

2002

2003

900

2004

20%

600

10%

300 0

0% –10%

–300

–20%

–600

–30%

–900 1

8

15

22

29

36

43

50

5

12

19

26

33

40

47

2

9

16

23

30

37

44

51

Week Sweden – Deviation from Median Norway – Deviation from Median

Nord Pool System Price Net Trade Norway to Sweden

Source: Nordpool, adapted from ECON, 2003a.

93

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Trade between Norway and Sweden is affected by many factors, particularly during winter when loads are high. Nevertheless, the relative tightness of the hydro reservoirs seem to be an important driver for the flow between the two countries during this period. During October and November 2002, reservoir levels in Norway were low but the situation in Sweden was even tighter: Norway exported to Sweden. Despite public pressure to intervene to protect short-term national interests, governments allowed the market to allocate scarce resources according to market players’ assessment of the value of water at any given point in time. The Nordic drought in 2002/03 put the Nordic electricity market under tremendous pressure. In that sense, it was an important test for the longterm stability of the market. The governing structures – from governments to regulators to TSOs – followed the development carefully but refrained from direct intervention. The market responded to the crisis by allocating scarce resources according to economic criteria and by attracting new resources from existing Nordic generation plants and through imports from neighbouring countries, as well as through demand reductions. The full range of potential opportunities seems to have been exploited to a great extent. The Norwegian government presented a report on its view of the situation and proposed future measures to the Norwegian Parliament in December 2003.56 The report declares that the Norwegian government will continue to base its energy strategy on a well-functioning electricity market with trade between countries and concludes that the market managed the situation well. At the same time, it also recognises that attention must be given to the prospects for investments in new generation capacity such as renewables and gas.

Market Models...................................................................... This book focuses on issues that are important to establishing competitive and robust electricity markets. PJM, Australian, British and Nordic markets were chosen as focal points because of their relatively long and successful experiences. These markets are similar in the sense that they have all seriously addressed key parameters necessary to creating competition: unbundling; regulated third-party access; opening of retail markets; and establishment of comprehensive trading arrangements. Yet the market models or designs 56. Ministry of Petroleum and Energy, 2003.

94

3 PRICE SIGNALS ARE THE GLUE

chosen by each country vary significantly. While they all address basic aspects that determine the value of electricity – time, volume, location and quality – they apply different weights on the different aspects. Much of the transmission network in the United States is highly meshed, including the network in PJM control areas. It is natural, and probably necessary, to apply a nodal pricing principle in such a market. In contrast, the Nordic and Australian markets have load and generation centres that are interconnected in more radial networks and have fewer loop flows. In such systems, the loss of efficiency that comes with applying a zonal pricing principle seems more acceptable; considering the increased complexity of a nodal system, zonal pricing appears to be a natural choice. That said, the inefficiencies that arise when pricing principles fail to transparently reflect the real congestion continue to prompt debate in both markets. The England & Wales electricity system is relatively well interconnected internally but has relatively weak interconnections with neighbouring systems. In this context, the efficiency losses inherent in a single price zone appear acceptable. However, integrating Scotland into the British Electricity Transmission and Trading Arrangements (BETTA) created more debate, primarily because the interface between Scotland and the rest of the market is a clear congestion point and a decision was taken that it would not be priced transparently. Debates concerning more fine-tuned, locational pricing in the Australian, Nordic and British markets appear to be heavily influenced by political concerns for regional re-distribution. Even if such social aspects can be managed through other policy instruments, they appear to get a higher weight than considerations for cost-reflective price signals and well-functioning markets. Timing of trade also differs considerably from market to market. In the Australian and British markets, trade is concentrated close to real time; it is up to market players to make contracts that reach more than two to three hours into the future. Considering that both these are isolated (or in the British case, relatively isolated), there is no particular requirement for careful and timeconsuming co-ordination with neighbouring systems. In Australia, the need for co-ordination was addressed very comprehensively by forming a joint system operator with responsibility for operating the system and market across all involved states and territories. In the absence of requirements for external coordination, it appears natural not to complicate things by introducing several official trading cycles. These choices, however, raise debate as to whether transaction costs in forward markets will become so high as to discourage participation of some market players. In Australia, attracting active demand-side participation in the market remains a serious challenge. In the British market, there seems to be a trend of vertical re-integration within the generation and retail segments of the market, indicating that transaction costs in forward 95

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

markets are too high. In addition, operators of small plants, particularly intermittent wind generators, have complained about the market structure. In PJM and Nordic markets, market integration and trade across jurisdictions has been a top priority from the start. Both the Nordic and the northeastern U.S. markets use day-ahead price settlement as the main reference point and both have long traditions for cross-border trade, so the need for co-ordination is a natural focal point. Real-time markets provide corrections to the dayahead price settlement and longer term forward contract markets use the dayahead price as the reference. Such practice allows for more careful coordination and planning well in advance of operation. Also, day-ahead market settlement has the advantage of opening the market to players who do not have the resources to maintain 24-hour market monitoring, seven days per week. This contributes to liquidity, competition and flexibility but may, however, remove some of the continuous alertness from market players and dull their responsiveness to sudden changes. It is also likely that such an organisation of the market requires relatively more operational reserves Day-ahead commitments related to transmission capacity made available for trade create a framework for firm contracts for market players but potentially at the cost of loss in overall efficiency. In fact, day-ahead market prices may not correctly signal the real physical conditions of the system at the time of operation. It becomes critical also to operate an effective real-time market, including the option to trade across interconnections. PJM includes a careful real-time market settlement that also integrates and optimises flow across PJM control areas in real-time trading. Nordic TSOs operate an integrated real-time market for balancing services that ensures optimal flow across Nordic interconnections in real-time trading Other physical characteristics of the different systems also play important roles in the development and success of efforts to organise market and trading arrangements. The Nordic market originates from the liberalisation of the Norwegian market, where almost the entire generation capacity is hydro based – mostly from reservoirs, which constitute a very flexible balancing resource. Sweden also has a substantial share of hydro power. In 2004, almost half the electricity generation in the Nordic countries was from hydro power, giving the whole Nordic market access to this flexible resource that effectively enables relatively cheap short-term balancing of deviations from day-ahead schedules. The Nordic market has developed substantially with the inclusion of other Nordic countries and, thereby, with the inclusion of larger and larger shares of thermal-based power plants. Thermal power plants are generally costly to start up, so the short-term marginal cost of operating for one hour is 96

3 PRICE SIGNALS ARE THE GLUE

substantially higher than operating for e.g. 12 hours. As a consequence, Nord Pool introduced “block bids”, which allow a generator to bid at a certain price under the condition that he can operate for a certain number of hours. The sophistication of these block bids has increased continuously. In PJM, such socalled “unit commitment” considerations are included directly into the market dispatch process. There is a natural focus on cross-border trade in continental Europe. Most markets operate on a day-ahead planning and trading cycle, which facilitates co-ordination between TSOs and market operators. The ETSO-EuroPEX proposal on flow-based market coupling also builds on that by focusing on flows across borders between countries and jurisdictions. Considering the highly meshed nature of the European transmission system, which is similar to the transmission system in United States, it appears questionable whether a trading arrangement focusing on countries and jurisdictions will provide sufficient locational signals in the long term. To date, only limited efforts have been made to integrate real-time and balancing markets across Europe. All markets are under continuous development. Trading arrangements in the British market underwent a major reform in 2001. The common drivers for change include considerations about market power, demand-side participation, intermittent resources, small distributed generation, risk management and transparency. In summary, it is now evident that the design of an appropriate market model is a matter of creating trading arrangements that fit the specific circumstances of each electricity system while also addressing the key issues of unbundling, third-party access, cost-reflective pricing principles and transaction costs. It is also clear that there are important trade-offs between efficiency and complexity. There is no winning, “one-size-fits-all” market model. However, there are some common features in the pressures for development, particularly with respect to technology trends and the need for more demand participation. In that sense, there may be a certain level of convergence in market design.

97

4 RISK MANAGEMENT AND CONSUMER PROTECTION

RISK MANAGEMENT AND CONSUMER PROTECTION Uncertainty breeds risks. This is certainly true of the considerable uncertainty connected with many fundamental factors that determine electricity generation, transport and consumption. But it is also true that uncertainty – and thereby risk – can be reduced to some degree through skilful operation, analysis and information management. At the same time, a number of factors connected to energy generation will always leave some uncertainty: weather, technology development, fuel prices, competition, etc. Uncertainty can also be found in the regulatory framework. Here, it is in the hands of regulators and policy makers and it is important that these two stakeholders consider the relationship between regulatory uncertainties, business risks and costs. At the same time, a commitment to remove such uncertainty entirely is almost certainly not credible. The bottom line is that in the electricity business – as with any other business – risk will always be involved in every part of the value chain. For market players, it is important to minimise risks through competence and manage the remaining real risks to align their appetite for risks compared to the rewards. Efficiency requires that all elements are properly taken into account in all decisions, including the fact that information will always be imperfect – there will be uncertainty. The framework of incentives should be structured such that risks are allocated to those who make decisions and who hold responsibility for taking uncertainty into account. This is not the case in a vertically integrated and regulated sector in which all costs can be passed on to consumers. In such a system the consumer typically bears all the risks but investment and operational decisions are made within the vertically integrated utility, although the decisions may be subject to regulatory approval. Liberalised markets allocate risks to the real decision makers; utilities face the risks directly and must take them into account when making decisions. Market liberalisation does not change the risks or uncertainties in the electricity business. However, certain business risks may appear new to utilities and financial institutions lending money to these utilities. In fact, the risks are simply more transparent and appropriately shifted from consumers to the relevant market players. The challenge of managing risks in liberalised electricity markets is multifaceted. Electricity is inherently volatile with large variations both from hourto-hour, day-to-day and season-to-season. At the same time, most assets in electricity have economic lifetimes of 25 to 60 years. In the past, traditional cost considerations determined most aspects of investment decisions 99

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

including choice of technology. Fixed and variable costs were assessed for the economic lifetime of the asset and discounted back to present value, applying a discount rate that reflected the time-value of money – i.e. the levelised lifetime cost approach. This approach created a situation in which there were no or very small risks for decision makers, and low discount rates were applied to reflect that. This picture changes dramatically when risks must be fully considered in liberalised electricity markets; there is a greater appreciation of technologies that can be developed and implemented in small, incremental steps with low, up-front capital costs – especially where demand growth is low, irregular or otherwise uncertain. For example combined cycle gas turbines (CCGT) have many advantageous attributes for liberalised markets and are often a preferred technology. In electricity markets, risks are managed through contracts – contracts that connect generators with consumers. Electricity is physically delivered only in the actual moment of operation: a generator and a consumer/retail company make the final physical commitment when a schedule is forwarded to the system operator or the system operator makes the final dispatch. In the minutes, hours, days and even years preceding this moment of physical exchange, market players can change their commitments many times by trading contracts. Thus, it is easier to conduct business through financial contracts – i.e. agreements to exchange money as opposed to agreements to exchange the actual physical product. A generator and a consumer will establish a contract on electricity supply during a specific period in the future and will agree on a specific price. If the market price turns out to be lower than the agreed price, the generator must meet its commitment to compensate the consumer. The consumer may be compensated by just receiving the price difference, which allows him to buy the electricity in the market at the agreed price, or by actually receiving the electricity “physically”. As long as generators and consumers can rely on an underlying “physical” market, a purely financial contract carries the same value to them as a contract with physical commitments. These financial contracts are easier for market players to manage, without losing any value as a risk management instrument. Physical contracts, on the other hand, create significant challenges in market design. Trying to maintain physical attributes in contracts puts overall economic efficiency critically at stake. In many markets, physical contracts prompt generators to commit transmission capacity in order to fulfil the requirements of their contracts. It has been difficult to integrate these commitments efficiently into a set of market rules and trading arrangements that allow for sound, cost-reflective pricing. Often these contracts were established prior to liberalisation and, as a consequence, they may be a barrier to competition. 100

4 RISK MANAGEMENT AND CONSUMER PROTECTION

Electricity is inherently volatile. Therefore, efficient outcomes require continuous pricing of all factors, including the value of transmission capacity. If a contract is physical and commits a generator and a retail company to a specific time frame, it must be backed by guaranteed transmission capacity for the same period. For example, two market players may agree to transmit electricity from A to B at a fixed price for a year. If electricity is suddenly offered at a lower price in area B than in A – even for a short period of time – the efficient outcome would be to use transmission capacity in the opposite direction (i.e. to transmit power from B to A) during that period. If the contract physically reserves the transmission capacity, such a shift of flow would not be possible: transmission and generation assets would be used less efficiently as a consequence. If market power is also a factor in areas A and/or B, the dominating player may even have an incentive to block transmission capacity to maintain this dominating position. In these kinds of circumstances, physical contracts become a real threat to the functioning of the market. In the Nordic market, initial market rules allowed for contracts with “physical” attributes to a certain extent; these rules were made obsolete with the development of credible spot and balancing markets. Price volatility is a perceived risk factor for both generators and consumers. A generator runs the risk that the price will be too low to cover all costs. A consumer runs the risk that, at times, the electricity price will be too high compared with the benefits electricity consumption creates. A market for contracts offers generators and consumers an opportunity to meet and create a mutual risk hedge. Both parties will probably even be prepared to pay a premium to avoid risk. In this sense, contracts offer protection both for generators and consumers.

Power Utilities Use Contracts to Manage Risks................... In a vertically integrated and regulated electricity industry, risks are managed through decisions within the utility and all costs are passed on to consumers. Unbundling breaks up the value chain thereby bringing risks into the open – making them transparent. Competition effectively shifts business risks from consumers to those who actually make decisions at each point of the value chain. In this unbundled and competitive electricity industry, contracts become one of the instruments that market players will use to manage the business risks, just as they are used by many other sectors with homogenous products. A range of such instruments are now available, their development having been driven by the needs of market players, new IT tools and recent economic research. Many of the risk management tools and contracts previously applied 101

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

in other markets for commodities, currencies, stocks and bonds, are now being introduced in liberalised electricity markets. Competition and cost-reflective pricing of electricity are pre-requisites for successful electricity market liberalisation. Building on that, it is equally important that the unbundled industry have access to instruments to manage the risks that have now become transparent. Development of an effective market for financial contracts is an integrated part of a well-functioning electricity market. Financial contracts are necessary for market players to operate, but they are also important from the perspective of regulating competition. Economic research demonstrates that contracts committing market players over various future time-spans serve another important purpose: they effectively minimise the scope or incentive for market power abuse. In the final planning for the actual operation, when all information is on the table, a dominating market player may be able to analyse all options and possibilities and to manipulate prices in day-ahead or real-time trading. By holding back generation capacity or simply by bidding higher than marginal costs, the player can increase the market price. The player may lose market share temporarily but even this factor can be considered and integrated into revenue calculations. If the same dominating player also made commitments prior to day-ahead, these commitments must also be taken into account. At the same time, if the player made a previous commitment to deliver electricity at a certain price, there is no chance to benefit from momentarily manipulating that price to a higher level. Longer duration contracts remove the incentive to abuse a dominating position – or at least make the manipulation more complex and less profitable. From that perspective, an effective market for financial contracts is a sign of faith amongst participants in the market place. However, such an effective market will develop only if there are many players and if they trust that the underlying market price is not being manipulated by large incumbents. Thus, the transition period while a liquid financial market is developing becomes critical. The competitiveness of the underlying “physical” market depends on the presence of long-term contracts that limit the scope for market power abuse and the development of a liquid financial market depends on a credible underlying “physical” market. At the same time, maintaining previous long-term contracts negotiated prior to liberalisation may be an instrument for incumbents to maintain their market power. The California experience suggests that it may not be a good idea to cancel all contracts at the implementation of a reformed wholesale market. During the transition period, it may be important to accept old “physical” bi-lateral contracts (as was done in Norway) as long as they are not allowed to distort the basic pricing principle. 102

4 RISK MANAGEMENT AND CONSUMER PROTECTION

An effective market for financial contracts is characterised by many factors. Some of the most critical factors are: the presence of institutions for trading; access to products that meet the needs for risk management; and sufficient liquidity (a market is liquid when users are able to buy and sell products at necessary volumes without problems and without affecting the price significantly). A market player who decides to buy a contract always wants to know that there are many other players willing to buy the contract back at the market price. One measure of liquidity is that the distance between the “best offer” and the “best bid” for each product is not too large (i.e. there is a “small price spread”). In Nord Pool’s market for financial contracts there are so-called “market makers” in many of the most important products. A market maker commits to always give bids and offers of a certain volume at prices within a certain spread. Nord Pool compensates the market maker through various preferential conditions, including reduced trading fees. The amount of market turnover compared with the underlying product that is physically being delivered is another liquidity indicator (Table 2 shows liquidity in different markets and different market segments, divided according to the institutions involved). Real-time trade is offered by the system or market operator. Day-ahead trade may be offered by an official system or market operator but may also be offered by a private commercial exchange. The segment of the market for longer duration contracts is often offered in a so-called “over-the-counter” (OTC) market, in which brokers organise interaction between market players enabling them to exchange bi-lateral contracts. Real exchanges for contracts have also developed in several electricity markets. With the development of on-line trading platforms and other sophisticated electronic management tools, the distinction between OTC and exchange trade has become less obvious. Depending on the legal framework of each country, the most important differences may be the legal requirements for exchanges to fulfil minimum levels of transparency, market monitoring and binding trading rules. The largest difference is the management of counterparty risk. Data about liquidity in the real-time segment (Table 2) indicates differences in market design rather than providing any meaningful information about differences in liquidity. In Australia, the NEM is a centralised dispatching mechanism and NEMMCO has a 100% market share by definition. In PJM, market players are allowed to trade on their own, including meeting their balancing requirements: the 35% share indicates the level that balances their operations through this market clearing mechanism. In the Nordic market, the 3% real-time market liquidity indicates the actual turnover of regulating power needed to balance the Nordic electricity system. This is similar with the 103

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Table 2

Liquidity in various electricity markets: Turnover in different market segments as share of total consumption, 2004 England and Wales

Australia NEM

PJM

Nordic

Germany – European Energy Exchange

Real time

5%

100%

35%

3%

NA

Day-ahead

NA

26%

43%

11%

Further ahead: Exchange

NA

13%1

24%3

151%5

29%

Further ahead: Over-the-counter

NA

125%2

58%4

309%6

34%7

1. d-cypha trade, 2. Australian Financial Market Association, 3. NYMEX, 4. ICE, 5. Nord Pool, 6. Nord Pool Clearing, 7. EEX Clearing Source: d-cypha trade, AFMA, FERC, Nord Pool, EEX

turnover in the England and Wales balancing market. In Britain, Sweden and Finland, market players are allowed to balance generation and consumption internally which decreases liquidity. In Norway and Denmark, all regulation must go through the market to minimise the regulating requirement on a system-wide level. Day-ahead markets are the main reference in PJM and in the Nordic market. In the latter case, they accounted for 25% of contracts over a long period of time (PJM reports similar figures). The day-ahead share recently increased in the Nordic market to the current 43%, mainly as a result of a change in Nord Pool’s trading fees. Considering the key position of Germany in the European electricity system, development of the German electricity exchange (the European Energy Exchange or EEX), is very important for the development of electricity trade in Europe. EEX has been in operation since 2000, with a relatively slow but constant increase in traded volumes in the day-ahead market segment; spot trading in this segment now corresponds to 11% of the total German consumption. With EEX and Germany at the centre of European electricity trade, liquidity data from Germany may reflect this role more than actually being an indicator of the effectiveness of the German market. European electricity traders perceive the EEX day-ahead market as a “balancing market” for trades in continental Europe. 104

4 RISK MANAGEMENT AND CONSUMER PROTECTION

The real challenge lies in developing liquid markets for contracts with longer durations. Some of the first experiences in the British market, with so-called “contracts for differences”, developed into a relatively liquid market under the Pool trading arrangements. In the Nordic market, there was a strong increase in the liquidity of Nord Pool and OTC from the mid-1990s to 2002, at which point turnover in the financial contract market corresponded to more than eight times the total Nordic consumption. Turnover, measured in MWh, more or less collapsed during the Nordic drought but, when measured in traded values, the decrease was not equally dramatic. Liquidity began picking up again in mid-2003; in the first half of 2005, total turnover almost corresponded to the 2002 levels. Norwegian market players undertook some 50% of the trade in Nord Pool financial contracts and Nord Pool clearing; another 15% to 20% was undertaken by non-Nordic market players. Liquidity is also increasing in PJM and the FERC reports very rapid growth in the OTC market since mid-2004.57 In the British market, only the trade in the real-time balancing segment is a formal market. It is up to market players and commercial market institutions to create liquid markets for contracts with a longer duration. The markets with longer durations that have been established are not fully transparent and seem to be fairly illiquid. The exchanges UKPX and the International Petroleum Exchange organise electricity trading in the British market. Liquidity in these markets and in OTC trading is assessed by commercial consultancies. Market players in the British market seem to prefer a much less dynamic and costly risk management strategy of integrating generation and retail businesses through mergers and acquisitions. In such a situation, liquidity in terms of the possibility to trade real physical assets becomes even more crucial. This part of the British market is liquid, and indeed the competitiveness of the final outcome in terms of prices is another important indicator of liquidity. Another pre-requisite for creating liquidity is that contract specifications can be aligned in standardised products. A long-term contract could be a very thick document that requires substantial legal input but has the advantage of being designed to meet the exact needs of all parties. Liquidity requires that contracts are standardised in relatively few forms, depending on the needs of market players. A retail company has short-term volume risks (due to the uncertainty of customer consumption) and longer term volume risks (due to the risk of losing customers). A retail company faces the risk of prices that deviate from the sales contracts to final customers. On the shorter term, generators have price/volume risks related to the difference between the price of fuel and the other costs associated to operating the plant; on a longer term, they face the risk that the market price will not sufficiently compensate for investment cost and profit. For a hydro plant, there is a fuel-related risk tied to reservoir development. 57. FERC, 2005.

105

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

To fulfil these needs, a variety of contracts are on offer in all markets: daily, weekly, monthly, quarterly and annual contracts. In many markets, products are also divided into “peak” and “base” products, a base product being one for all hours of the contracted period and a peak product applying only during working hours on work days. The Nordic market is, so far, an important exception to this pattern; no standardised peak-load contracts are traded at this point in time. Due to the central role of hydro power in the Nordic market, there has not been great variation in the hourly prices – although this seems to be changing now. These basic contracts for financial delivery during a specific period in the future are known as “forwards” or “futures”. Forward contracts are generally traded OTC, with settlement at the time of delivery. Futures contracts are generally traded on an exchange, with continuous settlement of differences between the agreed price and the reference market price. In addition to these basic contracts, “swaps” may be offered – contracts that link electricity trade with other risk elements such as fuel prices. Also, the notion of “value of optionality” is becoming more and more recognised in many risky business environments, including electricity markets. There is growing appreciation for the notion that there is a value in being able to postpone a decisive choice of action; thus, the option of postponing an action has a value. It is valuable for a trader who owns a futures contract to be able to postpone the decision to sell the contract until he knows how the price will develop. It is equally valuable for an investor to build a new generation plant in small steps, being able to wait with the next step until it is called for and financially a sound investment. For a consumer, it is valuable to have some time to consider a new contract offer. Markets for option contracts that standardise these risk elements – with respect to standard features, terms and conditions – are developing, but are still relatively illiquid. Risks in the day-to-day management of electricity generation and consumption have led to the creation of liquidity for contracts extending two to three years into the future in some markets. Many commitments reach even further ahead in time. For example, an investor in a new production plant might prefer to hedge the risks of the investment in a contract that spans a substantial part of the plant’s economic lifetime. In general, there are few buyers for such contracts, which indicates that the investors are not willing to accept the discounted price in such a contract – the risk premium that a buyer would require. The need for liquid financial markets in the context of new investments is not a matter of allowing investors to cover all risks; investment in power generation is inherently risky, as is the case for all other investments. What is important, however, is that 106

4 RISK MANAGEMENT AND CONSUMER PROTECTION

an investor knows that there is a liquid market for sale of the power during actual operation and that there is also a liquid market that makes it possible to manage the risks of buying fuels and operating the plant. Markets for financial electricity contracts took a serious blow in late 2001, in the wake of the collapse of the global electricity utility and trader, ENRON. ENRON was a very large player in many markets; in that sense it contributed substantially to the liquidity in many electricity markets. ENRON operated its own on-line trading platform on which it posted its own bids and offers across several markets and market segments. The disappearance of such a large market player decreased liquidity in its own right. However, the most important effect was the reinforced focus on another important risk parameter: namely, the risk that a trading partner cannot fulfil its commitments. Apart from volume and price risks, this is the third most important risk to consider. Counterparty risk, as it is known, attracted more attention after the ENRON collapse. The standardised service offered to manage this risk is called “clearing”, in which an institution offers to step in between two market players who have entered into a contract. The clearing institution manages the counter party risk of each market player by assessing credit-worthiness and collecting collateral. Its presence allows individual players to forego costly management of the creditworthiness of all trading partners and to concentrate on fulfilling its own requirements from the clearing institution. The clearing institution holds a key role in market transactions: a default on a contract becomes the problem of the clearing institution rather than of the trading partner. One of the differences between an exchange and an OTC market is that the exchange implicitly offers clearing, although many electricity markets also have some market institutions that offer clearing services for OTC traded contracts. Nord Pool offered clearing services for OTC contracts from 1997; ICE, the large and quickly growing OTC trading platform in the PJM market, also offers clearing services. Liquidity in the Nordic market did not suffer significantly from the ENRON collapse, even though ENRON was also an important player in this market. Contract clearing may have been a key factor in minimising the impact. As liberalisation made risks more transparent – and an integrated part of the decision-making process – it also prompted the development of very sophisticated risk management instruments in some markets. One of the most important remaining challenges is that of identifying the drivers for liquid markets for financial contracts and enabling this development in all liberalised markets. What are the main drivers behind the relatively liquid financial market in the Nordic countries? Many years of development lie behind the success of the 107

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

liberalised market in Nordic countries. Norway liberalised its electricity market in 1991 but financial trading only really began taking off in 1998-2000. The ownership structure of Nord Pool may also have an influence on the development. Nord Pool Holding owns Nord Pools financial trading activities. In turn, Nord Pool Holding is owned by the Norwegian and Swedish TSOs, Statnett and Svenska Kraftnät. Even if OTC trading still has the largest share of the Nordic financial market, Nord Pool undoubtedly played a very important role in developing the relatively liquid Nordic market. However, commercially, Nord Pool has not been profitable. The relative success of the Nordic financial market could be partly due to the fact that it is backed by a certain level of support from the two state-owned TSOs. The character of risks associated with the large concentration of hydro power may also have an influence. Liquid financial markets contribute an important share of the potential benefits of liberalisation. The sophisticated risk management instruments offered in liquid financial markets improve risk management within utilities while also creating a platform for developing new customer products. To the electricity consumer, electricity is by-and-large a standardised product: the only things that matter are quality, reliability, price and the level of service in connection with billing. Retail prices vary with the wholesale price of electricity and with the revenue the retailer requires to cover costs and make a profit. Billing is a relatively straight-forward activity with substantial economies of scale, so it is difficult to diversify electricity as a product in that dimension. The most obvious way to differentiate electricity as a product – and thereby add value to the consumer – is to offer risk management. In a competitive retail market, consumer decisions on how to manage price risks are probably much more important for the actual annual price paid than the choice of retailer. The decision to sign a one- or two-year contract at a specific moment can lead to substantially different annual prices than a decision to accept the spot price. In this sense, a modern electricity retail company is more like an insurance company than anything else and the product for sale is risk management. Product development focuses on finding ways to offer customers the contracts that best fit their appetite for risk; this, in turn, requires a liquid market for financial contracts.

Contracts Offer Protection for Consumers.......................... Electricity prices reflect the volatile nature of electricity, and will rise to very high levels from time to time. During periods in which the electricity system is constrained by available generation capacity to meet demand, the price will 108

4 RISK MANAGEMENT AND CONSUMER PROTECTION

go up in a short spike. If the electricity system is constrained by energy supply (e.g. due to drought in a hydro-based system or high fuel prices), the price will increase to a higher level over a longer period of time. For large consumers with a high share of electricity costs in their overall costs, such price volatility is a real business risk. They may have committed to selling their products at prices that are not profitable or competitive if electricity prices soar beyond a certain level. On the other hand, prices can also fall to low levels – triggered by low fuel prices, abnormal temperatures or high inflow to hydro reservoirs. Taking advantage of periods with low prices would be the reward for taking the risk of not signing a contract with fixed prices over a longer time period (e.g. one or two years). Small household consumers generally spend only a low share of total income on electricity. Compared with other risks they accept in dealing with loans for houses and cars, petrol prices, foreign currency, etc., the financial risk from electricity prices is often relatively low. On the other hand, the variations of real-time and day-ahead electricity prices may create uncertainty for household consumers who do not understand and closely follow the development of electricity markets. A threat to the competitiveness of large industrial electricity consumers, coupled with uncertainty on the part of household consumers, can very quickly turn into strong political pressure to intervene in the market. There are, however, no price-risks that cannot be managed fully through contracts. The most normal contract is to agree on a fixed price for a fixed term. More sophisticated contracts exist, based on the spot price but with a cap (retailers are able to offer these contracts by using option contracts in the wholesale market). Many of the largest electricity consumers spread risk by contracting a certain share of their anticipated consumption in one form of contract and other shares in other forms. The key point is that contracts can effectively shield consumers from short-run price fluctuations For example, a large industrial consumer may have established contracts to sell aluminium over the following two years. In a competitive retail market, the consumer can sign electricity contracts that fix electricity price at levels that render these deals profitable. If the electricity price turns out to be very high, the aluminium smelter may consider re-selling the electricity in the market and fulfilling its commitments to supply aluminium through the world market for aluminium. This happened to a great extent in Norway during the winter drought in 2002/2003. To reduce risk associated with the volatility of electricity, a household consumer might consider whether it is worthwhile to pay an insurance 109

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

premium to avoid fluctuations in monthly electricity bills. If not, the consumer should sign for a contract based on the spot price plus costs for billing. If the consumer prefers to avoid risk of fluctuating monthly bills, the best option may be to consider a one- or two-year contract at fixed price. It is evident that contracts can offer consumers protection from the inherent price volatility of electricity: but what about the question of political intervention if prices change dramatically? If the underlying wholesale price is competitive and the retail market itself is competitive, a political intervention to protect consumers would primarily be a risk-management decision taken on their behalf, and not necessarily an efficient decision. On top of that, a political intervention carries a cost in the form of lost credibility. The market foundation is undermined, potentially to a large extent and for a long period – well after the political intervention, and its underlying causes, have ended. Political intervention also withdraws the added value that comes from freedom of choice. In wholesale markets, it is not a sustainable solution for consumer protection. Other critical decisions are necessary in relation to consumer protection. As many consumers will not be very active in retail markets, at least not in an initial phase, it becomes important to define how these customers are treated, i.e. what type of contracts are consumers left with as default? An option used in some markets, including New South Wales, Denmark and Spain, has been to maintain regulated tariffs for consumers who do not switch from the incumbent supplier. This helps protect the consumer, but it makes it very difficult for retailers in the market to compete with these regulated tariffs. If the benchmark for regulation is set at a low level, it will be impossible for a competitive retail company to propose an alternative offer that is sufficiently below the regulated tariff. In the short term, this may seem to be in the consumer’s interest. But, if the result is that competition is undermined – and thereby also begins undermining innovation and development – it may serve the consumer poorly in the longer run. A tariff that undercuts a competitive market price will probably also be economically unsustainable; eventually it would have to be based on crosssubsidisation from other activities such as an associated network business. The issue of default contracts also arose during the Nordic drought in the winter of 2002/03. In Norway, most residential consumers traditionally purchase electricity under a contract form in which prices can change from week to week, depending on the wholesale price (this flexible contract is the standard for historical reasons). During the winter of 2002/03, these flexible contracts created a situation in which retail electricity prices increased at unprecedented rates, within weeks of the increased wholesale prices. Consumers responded by reducing consumption to a certain extent, but it also 110

4 RISK MANAGEMENT AND CONSUMER PROTECTION

raised intense public debate about the functioning of the market and how to address the social consequences. Sweden faced a similar drought and saw more or less the same increases in wholesale prices. But in Sweden, the norm for residential customers is a one-year contract at a fixed price. As a consequence, retail electricity prices increased only following a substantial time-lag and only at a fraction of the short-duration increases experienced in Norway. The ensuing public debate in Sweden had a fundamentally different character and did not include social aspects. Unsurprisingly, residential customers in Sweden barely responded to the price by reducing consumption. In the wake of the drought (during the 3rd and 4th quarters of 2004), the share of Norwegian residential consumers with variable price contracts decreased from almost 95% to 85%. However, it has since increased again, indicating that Norwegian households still prefer variable price contracts.58 In most liberalised markets, the approach is to let all retail prices be determined through competition within the retail market itself. This raises concerns for consumers who do not want to take the time to evaluate differing electricity contracts. The transaction time and cost connected with actual switching also become critical factors. In some markets a centralised dissemination of price information eases the process by allowing consumers to compare retail prices; other markets use other approaches, but always with the same goal of facilitating comparison – and often through on-line services. The United Kingdom Utilities Act 2000 established an independent body, Energywatch, to protect and promote consumer interests by providing impartial information and advice, and by taking up complaints from consumers. The Norwegian Competition Authority publishes detailed price information on its Web site, as does the Swedish Consumer Agency. In Denmark, the electricity association collects and publishes retail price information. In Finland, incumbent retailers in a specific local area are required by law to list competitor prices. Various public utility commissions in the United States, in those states where there is freedom of choice, operate Web sites with price information (e.g. the Public Utility Commission of Pennsylvania). Consumers will only be protected through contracts if the wholesale market behind such contracts is competitive – and if competition between retail companies leads to product offers that meet consumer needs. In that sense, consumers are best protected through a well-functioning, wholesale market with cost-reflective pricing, liquid financial markets and competitive retail markets. Several markets have found the recipe for relatively successful wholesale markets. Experiences with competitive retail markets are more uncertain. 58. NVE, 2005.

111

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Retail Competition................................................................ A competitive retail market is the last link to ensure that consumers can receive some of the benefit derived along the value chain in electricity market liberalisation. In fact, effective retail competition creates a natural protection for consumers. Retail companies under pressure from competition may develop innovative contracts and products that create even more added value from liberalisation. As much as it is a crucial part of the liberalisation package, retail competition has proven to be one of the more difficult challenges and it carries the cost of rolling out the necessary infrastructure to make retail competition feasible. Delivering benefits from competition to small residential consumers is the most difficult as the costs relative to the benefits may be significant. Installing interval meters that can be read remotely on a daily basis is the option that adds the most value, but it is relatively costly to execute and operate. The alternative for small consumers is a system that uses calculated load profiles based on monthly, quarterly or annual meter readings. But this is also costly to manage. Moreover, it creates distorted incentives and it fails to enable a pass-through of hourly prices. In most jurisdictions in the British, Nordic, PJM and Australian markets, all consumers have the right to change retail supplier (Figure 9).

Figure 9

Year in which all consumers are allowed to switch retailer

South Australia

Victoria New South Wales

District of Columbia

Maryland Pensnylvania

United Kingdom

Finland

Sweden

Norway

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Denmark

Texas

New York

Delaware

New Jersey

Examining the shares of consumers who have exercised this right shows interesting trends, even though the switching rates are not directly comparable (Table 3). Data is collected from various sources and the oldest dates back to 2001 in the Finnish case.

112

4 RISK MANAGEMENT AND CONSUMER PROTECTION

Table 3

Switching rates amongst contestable customers: Number of customers no longer supplied by incumbent retailer as share of total number of residential and non-residential customers Norway Sweden Finland England New Pennsylvania & Wales Jersey Customer class

Residential Non-residential

By 4th quarter 2004

By 2004

24%

32%

31%

N.A.

Maryland New York Customer class Residential Non-residential

During By March By 2001 2003 2004 1%

38%

3%

N.A.

D.C.

Texas

By 2004

6% 18%

Australia – NEM

15% Denmark

By 2004 By 2004 By 2004 By 2004 By May During 1st 2005 quarter 2003 4% 29%

5% 33%

11% 38%

10% 62%

1% 16%

8%

Source: NEMMCO, OFGEM, Elfor, NVE, STEM, Finergy, Pfeifenberger, Wharton & Schumacher (2004)

Even though these data are not directly comparable, the tables illustrate several general tendencies. Non-residential consumption generally represents between 50% and 75% of total electricity consumption in a given country. In general, non-residential consumers with higher consumption have substantially higher switching rates; in fact, their switching rates can be high enough to undermine the market share of a particular retailer in this segment. According to the fourth benchmarking report carried out by the European Commission in 2005, more than 50% of large industrial consumers switched supplier since they acquired the right to do so in the following countries: United Kingdom, Norway, Sweden, Finland and Denmark. In contrast, residential switching rates are relatively low, except in countries with many years of full retail contestability, such as the United Kingdom, Norway and Sweden. It must be noted that these significant residential switching rates can also be attributed to other factors. In the United Kingdom, electricity and gas retail have been pooled, which makes switching more attractive for many consumers. In Norway and Sweden, residential electricity consumption is high due to the high shares of electric heating.

113

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

A consumer’s decision to switch supplier will typically depend on how much he can gain and how much it costs to switch. The benefits will depend on the level of competition between retail companies. Retail companies under pressure from competition will cut profits to a minimum, optimise customer management and develop new innovative products, probably focusing on risk management. The decision to switch typically depends on the direct switching costs and on the efforts necessary to assess offers and perform the actual switching process. In a policy context, the most straightforward path to generating the necessary competition in the retail sector has been by minimising the costs of switching and lowering the cost of new entry through the creation of a framework for an efficient balancing market and a liquid financial market. A first serious barrier to switching is the actual switching management. A local network company with its own retail arm will not have incentives to make consumer switching easy. In most markets, setting up standardised procedures that correspond with regulated third-party access on the wholesale level has been an integral part of the regulatory framework. In Australia, NEMMCO manages a fully centralised system for consumer switching. Timely billing is a recurring problem in many markets: there are typically long delays in transmitting correct billing data from local network operators to commercial retail companies. This is a constant threat to truly independent retail companies. Free flow of information is another important barrier. Local network companies usually possess all historical metering data; unless competing retailers have easy access to the same data, an incumbent retail supplier may stand to benefit from this information. Another important issue, particularly for residential and small consumers, is the necessary metering equipment. Electricity consumption priced according to the normal hourly or half-hourly market settlement requires metering at the same time intervals, and the metering data must be passed on for processing within the normal settlement cycle, which is usually a daily cycle. At larger consumers, this function is carried out by remotely read interval meters installed on site (in the Nordic market, more than 50% of the load is metered by remotely read interval meters).59 The price of installing, operating and maintaining this equipment is easily justified for higher levels of consumption. For low commercial and residential consumption, the cost-benefit comparison is more questionable. The alternative attempted is to construct a consumption profile for small consumers, which can be used to compute electricity consumption by a specific residential consumer in a specific hour according to a standard formula. The 59. Nordel, 2004b.

114

4 RISK MANAGEMENT AND CONSUMER PROTECTION

computation will always be an approximation, however. Increasing the level of sophistication would improve the approximation, but it would also increase the costs of computing and managing the system. This approach, known as load profiling, prompted high shares of residential retail switching in England and Wales, Norway and Sweden. However, this practice will inevitably introduce distorting incentives – at least to a certain extent. Some particular consumer profiles will benefit at the cost of other types of profiles. For example, consumers with a load profile which has higher peaks than the assumed load profile – and with higher consumption during high price hours – will benefit. An example of marketing by retail companies that illustrates some of the consequences of such distortions comes from Australia, where retailers offered cash-back incentives to encourage customers to buy air-conditioners. A retail company may benefit from peaky demand in the market segment with load profiles. This, in turn, increases peak demand in the total NEM. Recently, more and more jurisdictions are giving serious consideration to fullscale roll-out of remotely read interval meters. In Italy, the utility Enel decided to install remote metering equipment in approximately 30 million households, at an estimated cost of EUR 2.1 billion (corresponding to a unit price of EUR 70).60 By the end of 2004, more than 20 million meters had been installed.61 Several local network companies in the Nordic region initiated similar full-scale, roll-out projects. The local Danish network company, Sydvest Energi, aims to install remote metering for all of its more than 150.000 consumers by the end of 2007.62 The main motivation for these projects is that they lower the administrative costs of managing the metering process while also contributing to improved function of the electricity market. Experiments and research on intelligent meters and communication via the electricity grid are ongoing in many countries. In the European Union, the Electricity Market Directive commits all member countries to provide all consumers with free choice of retail supplier by 1 January 2007. But many countries still debate the overall costs and benefits of full consumer contestability. In 2001, the Government of Queensland released a report assessing the costs and benefits of full retail competition and subsequently decided not to introduce the concept. It should be noted that the only benefits regarded in the report were those that resulted directly from the assessed savings due to competition; more dynamic effects, such as those from product innovation, were not taken into account. Full retail contestability was also abandoned in several U.S. states. 60. IEA, 2003b. 61. ENEL, 2005. 62. Sydvest Energy, 2005.

115

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Recognising that consumers without the freedom to choose are essentially captive to electricity suppliers, several alternative approaches have been suggested to address this issue. One of the most serious threats of leaving residential customers in a permanently captive state is that they will become subject to cross-subsidisation to large consumers. One approach is to allow competitive bidding for the right to supply captive customers, but it carries the risk that the incumbent utility will have an important information advantage. The debate on full retail competition remains inconclusive, but more or less all consumers in the relatively successful markets (the United Kingdom, Australia, the Nordic market, and PJM) have the freedom to choose, and many are using this freedom. In the end, it is a matter of considering two inter-related elements: a) how best to protect the consumer from inefficiency and crosssubsidisation; and b) how best prepare for innovation opportunities that future technology development may offer.

116

5 INVESTMENT IN GENERATION AND TRANSMISSION

INVESTMENT IN GENERATION AND TRANSMISSION Liberalised electricity markets are improving the efficiency of electricity system operations in many different parts of the value chain, from generation to consumption. Competition and cost-reflective pricing keeps pressure on all decisions and provides a clear signal for actions. However, operation costs represent only a share of total electricity costs. A substantial share of the electricity bill paid by the electricity consumer comes from financing generation and transmission assets. Improving decision-making processes for investment related to generation capacity may be one of the most significant benefits of liberalisation. But it is also one of the most serious challenges in liberalised electricity markets. Many investment projects have long lead times and carry financial implications for several decades after implementation. In a transitional phase while markets are established, there will be uncertainty that may undermine the investment climate. A liberalisation process may end in a “Catch-22” situation in which successful liberalisation depends on investments and investments depend on a successful transition to liberalised markets. Investments in power generation seem to be the big test for the development of robust and sustainable markets. Many markets, including the Nordic, benefited from an initial overcapacity, which initially reduces pressure on the market, allowing it to pass through a transitional phase while market institutions are developed. Policy makers have time to build credibility for the market without being pressured by a tight supply-demand balance. At the same time, overcapacity also stresses the need for improvement. Markets with overcapacity are also often characterised by relatively slow demand growth, which implies that it is more difficult to identify the right timing for new investment and when tightness kicks in after some years, it will still lead to political pressure. In contrast, markets that start out with a tight situation, such as South Australia and Ontario, are brought into a particularly challenging situation shortly after market launch. In essence, it is put to the test before it can develop properly. But the high growth in demand that caused the tight balance, as in South Australia, creates a situation that allows for swift and clear price signals for investment. All in all, it is inevitable that new investments will be needed at some point, and that point will be a test to liberalised markets, regardless of the point of departure. 117

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Market-driven Investment in Generation ........................... A competitive bidding process determines prices in liberalised markets. Consumers and generators find the marginal resource that balances demand and supply, and a price is established. In periods with low demand, prices will be set by the generation plants with the lowest marginal costs. In periods with higher and peak loads, the generation plants with the higher marginal costs set the price. During the few hours with very high demand, the most expensive resources set the price – usually plants with low capital costs and high marginal costs, such as open cycle gas turbines (OCGT). Price can also be set by a reduction or shift of demand if consumption has a lower value to some consumers than the alternative market price. Marginal prices are set where marginal costs equal marginal benefits. A generator under competition will be willing to produce at a price that pays the cost of each additional MWh, but this marginal cost will not cover depreciation or provide returns on the invested capital. Generation plants recover invested capital during periods in which the price is set by the more expensive plants. Thus, plants with high capital costs must operate most hours of the year to be profitable, even if marginal costs are low. These base-load plants will recover the invested capital when prices are set by plants with higher marginal cost. Mid-merit plants will only recover the invested capital during hours in which the most expensive resources are setting the price. The most expensive resources are only activated during the very few situations with very tight supply/demand balance. Traditionally, resources for peak load rely on generation plants such as OCGTs. Thus, the profitability of investment in OCGTs depends on the owner’s ability to bid OCGTs into the market at prices above marginal costs. This is not usually a problem because the owner of this “last” resort will also have substantial market power in the specific hours within which it is needed. The owner will be able to collect a scarcity rent. On the other hand, this market power may pose a threat to the economic efficiency of the entire market and raise political considerations. An important point for the functioning of the market is that a generator should never be the last resort. The value of electricity consumption is not indefinite. Electricity is used for millions of different purposes by millions of different users. Each user will assign a different value to each purpose. It ranges from the value of being able to heat a private swimming pool on a Tuesday at noon to being able to safely perform heart surgery. At a certain price, there will be some consumer who is willing to make the effort to reduce consumption or shift it to another 118

5 INVESTMENT IN GENERATION AND TRANSMISSION

point in time. If market rules are adjusted to make it as easy as possible for the consumer to participate, more consumers are likely to be able to respond and lower price will result. Initially in liberalised markets, large industrial consumers are the most relevant, depending on their marginal benefit of consumption and on the costs of managing load reduction in terms of equipment, organisation and time. Market rules that allow for smooth participation of the demand side will effectively “cap” the market power of the generator with the last pivotal supply. If cost-reflective pricing in competitive markets can effectively incorporate a portion of the demand side in system balancing, liberalisation has, in effect, created a new resource. This illustrates well the new investment paradigm that is developing in liberalised markets. Technologies are seen in new roles; the previous division into base-, mid-, and peak-load technologies is less clear. Risks have become more transparent and are shifted from consumers to those market players who make decisions on investment and operation. As a result, capitalintensive technologies with long construction times are viewed with much more scepticism – even if marginal costs are low. On the other hand, cross-border trade is increasing and now provides a larger market for capital-intensive technologies. Access to larger, growing markets may sufficiently reduce the risks of large, up-front investments. Still, various barriers and complexities stall the development of markets that are robust enough to balance supply and demand in extremely tight situations. First of all, there is a strong, traditional focus on the supply side in the electricity sector – partly because it is costly and complicated to involve consumers in decision processes. This focus on supply also affected the establishment of market rules: they are rarely conceived with careful considerations on enabling demand-side participation. The policy perspective reveals an inherent transitional problem. To be able to manage their own active participation in the market, consumers will need to invest in equipment, education and time. Most consumers will also have a marginal benefit of consumption that is substantially above the market price during normal conditions. Average market prices are often in the range of EUR 20 /MWh to EUR 40 /MWh: but consumers with the lowest marginal benefits may only be willing to shift demand at prices of 10 or 100 times the average price. This implies that consumers cannot be expected to respond to prices before they observe some price spikes in the market. Ergo, initially, some generators may be put in a position in which they are the last resort and have full market power. From a political point of view, this may appear unacceptable – but intervention may, on the other hand, undermine the framework upon which active demand participation must be built. This is illustrated by the impact of price caps, set at too low of level, on investments in 119

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

peak generation plants. In that sense, political intervention motivated by fear of a market failure becomes self-fulfilling. The severity of the problem may be even greater if there is risk of real shortage during this transitional phase. If the system is faced with several sudden rare events, it may be necessary to shed load by force to keep the rest of the system operational. It is, however, not socially acceptable to regard forced outages as a natural way to balance supply and demand in electricity markets. As a consequence, many markets introduced price caps. This limits a dominating player’s room to manoeuvre, but it also effectively restricts overall ability to find the most efficient balances between demand and supply. If the price cap is low, it also sends a clear signal to the market that there is a will to intervene. This may have an even more discouraging effect than the price cap itself, reducing willingness to invest in generation, especially in peaking plants, and in demand participation. But restricting market power is not the only reason for introducing price caps. If forced load shedding is the final resource, then this should at least be given a value. The problem originates from the fact that consumers do not disclose the values they see from consuming electricity in the first place. So the value or costs of forced load shedding is, by definition, unknown. This so-called “value of lost load” (VOLL) needs to be assessed. Some countries actively use VOLL assessments to calculate price caps and to determine the benefits of other initiatives to improve system security. Table 4 shows the different price caps used in PJM, Australian, British and Nordic markets.

Table 4

Price caps in PJM, Australian, British and Nordic markets /MWh PJM

Australia

USD 1 000

AUD 10 000

EUR 830

EUR 6 250

Britain Denmark Finland None

None

None

Norway

Sweden

NOK 50 000 SEK 20 000 EUR 6 300 EUR 2 100

PJM has a price cap that is mainly designed to act as a shield against market power abuse. Most liberalised markets in United States have price caps of the same magnitude. In Australia, Sweden and Norway, the price caps are motivated by considerations about VOLL. In Australia, the price cap was increased from AUD 5 000 to AUD 10 000 in 2001, when it turned out that the lower cap was used too often. In the Nordic market the spot exchange,

120

5 INVESTMENT IN GENERATION AND TRANSMISSION

Nord Pool, also has a limit in its system price of NOK 10 000 /MWh (app. EUR 1250 /MWh); however Nord Pool regards this not as a price cap but as a technical limit, which will be increased immediately if need be. In Sweden, there seemed to be a real risk that supply could not meet demand without forced outages, particularly after a few very tight situations in the winters of 2000 and 2001. One of the root causes of the problem was that some older, oil-fired plants had been mothballed. The Swedish government commissioned the TSO, Svenska Kraftnät, to conduct a study on investment in new capacity in the open market and propose solutions. The solution suggested – and accepted by the government – was to give Svenska Kraftnät the authority to purchase additional reserve capacity over a transitional period from 2003 to 2008. The rules for activating the reserves are carefully designed to minimise the distortive impact on the market. For example, if the capacity is activated, it is at prices increasing from SEK 8 000 /MWh to SEK 15 000 /MWh over the five year period. The purchase of reserves takes place in a competitive tendering process and an effort is made to attract as many resources from the demand side as possible to participate in the bid.63 It should be noted that Svenska Kraftnät’s study did not include any discussion of their ownership of gas turbines for operational reserves and the role this might play in creating transparency and predictability for investors. In the Australian, British and Nordic markets, governments clearly stated their intent that market price signals will provide market players with the right incentives for investment. Except for the transitional arrangement in Sweden, no additional financial incentives will be used. These markets are “one-priceonly” or “energy-only” markets. It should be noted that it is false to think that capacity will not receive remuneration in energy-only markets. A one-price-only market is intended to provide remuneration for both variable and fixed costs through the same price, which corresponds to the normal pricing principles for most other products. In the Netherlands, the government initiated a study on security of supply, including whether one-price-only markets create incentives for adequate generation capacity. An independent governmental agency that carries out economic policy analysis, the CPB (the Netherlands Bureau for Economic Policy Analysis) conducted a cost-benefit analysis that concluded that the costs of a specific capacity measure could not be justified by the benefits.64 Instead, the Dutch government focused on creating a regulatory and market framework that provides transparency and clear price signals. It also 63. Svenska Kraftnät, 2002. 64. CPB, 2004.

121

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

authorised the TSO, TenneT, to establish a market for last resort capacity reserves along the same lines as the transitional arrangement in Sweden.65 Production capacity is now being built in the Australian, Nordic and British markets. The South Australian market was put to a test quickly after its launch, when prices spiked to reflect the scarcity of supply. The rising prices raised political concerns, but the only intervention was to double the price cap. Investments were made in peak load plants, which was also the appropriate technology to meet the problem of a rapid increase in peak demand. The Nordic market has experienced a long period of overcapacity and no new investment, followed by a period of tighter supply/demand balance, during which investors hesitated to make investment decisions. The only substantial new investments were in subsidised wind turbines, mainly in Denmark. Today, several new plants are under construction: a new 1 600 MW nuclear unit is under way in Finland; Sweden is building a new 400 MW combined heat and power (CHP) plant; and Norway decided to build two new combined cycle gas turbines. In many countries, much political concern remains for the barriers to timely and efficient investment, including whether a one-price-only market will give sufficient remuneration for the invested capital. These concerns are mirrored in Australia, the United Kingdom and the Nordic countries, where the situations are monitored carefully. However, there are – as yet – still no failures to point to. Investments in liberalised markets are made “just in time” i.e. when there is a real need. So far, there is no reason to believe that cost-reflective pricing, real competition and good market institutions will fail to send signals for timely and efficient investments. Instead of creating extra incentives to improve investment decisions through intervention, policy makers can adopt other and more obvious roles to remove barriers to investment. In the Nordic market, environmental policy caused substantial uncertainty for investors. In Sweden and Norway, respectively, long debates have ensued about nuclear phase-out and the use of natural gas for power generation. In Denmark, subsidies for renewable technologies led to a massive increase in renewable generation capacity, including 2 500 MW of intermittent wind capacity. The Nordic market has, so far, managed the situation despite such levels of uncertainty. Apart from the political reasons for hindering use of certain technologies and supporting others, there is also the even more important problem of acquiring permission to build. The Not-In-My-Back-Yard (NIMBY) and the Build-Absolutely-Nothing-Anywhere-Near-Anybody (BANANA) syndromes are an ever-present part of the investment decision 65. Boot, 2005.

122

5 INVESTMENT IN GENERATION AND TRANSMISSION

process in the electricity sector. Smooth and transparent approval procedures for new generation plants and transmission lines are central to creating a framework that allows for adequate, timely and efficient investment.

Box 4 . Market-driven investment in Australia The Australian National Electricity Market (NEM) started operating in December 1998. Average monthly prices from 1999 to 2004 are shown in Figure 10. Peak demand in Australia is during the summer – due to high use of air conditioning - and is increasing at a rapid rate. During the first years after the market opening, the balance between supply and demand was relatively tight. This is reflected well in the prices, with relatively high average prices particularly during the summer months and particularly in South Australia in the summer 2000/01. The NEM is based on a one-price-only principle in which the electricity pricing determined by normal trading arrangements provides remuneration both for variable and fixed costs. Timing, volume, Figure 10

Average monthly prices in national electricity market, Australia AUD/MWh 140 120 100 80 60 40 20 0 Jan-99 July-99 Jan-00 July-00 Jan-01 July--01 Jan-02 July-02 Jan-03 July-03 Jan-04 July-04 Jan-05

NSW QLD

SA SNOWY

VIC

Source: NEMMCO

123

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

technology and location are left to investors to decide. Figure 11 shows the development of installed generation capacity at principal power stations in Australia – past, present and predicted.

Figure 11

Installed capacity principal power stations - 30 June, MW MW 14000 New South Wales 12000 10000 Queensland 8000

Victoria

6000 Snowy Mountains

4000

South Australia

2000

06 20

05 20

20 04

20 03

02 20

20 01

00 20

19 99

19 98

19 97

0

Source: ESAA and NEMMCO, predicted number from NEMMCO Statement of Opportunity, 2004.

Generation capacity in South Australia increased by 49% from 1998 to 2003 adding 1 133 MW to the total installed principal generation capacity. Almost half of that capacity was in the form of OCGTs for peaking purposes. The rapid increase in peak demand in South Australia, e.g. due to increased use of air-conditioning, created a tight balance between supply and demand during peak load periods, particularly in the summers from 1998 to 2001. This tightness was reflected in NEM prices, which peaked up to the price cap at AUD 5 000 /MWh in several hours during the first years of the new market (Figure 12). The price cap was then increased to AUD 10 000 /MWh. South Australia was particularly in need of more generation capacity suitable for peaking purposes. OCGTs offer three advantages that make them well suited for this purpose: they have very low investment costs per kW installed capacity; they can be extended in small incremental steps;

124

5 INVESTMENT IN GENERATION AND TRANSMISSION

Figure 12

Price duration curves for the highest percent in South Australia 10 000

AUD/MWh

9 000 8 000 7 000 6 000 5 000 4 000 3 000 2 000 1 000 0 100,0% 99,9% 99,8% 99,7% 99,6% 99,5% 99,4% 99,3% 99,2% 99,1% 99,0% 1999 2000

2001 2002

2003 2004

Source: NEMMCO

and they are very easy and fast to build compared to other conventional technologies. The relatively high cost of OCGT operation is less important as they will only run on rare occasions. Key financial features of gas turbines in Australia are in the order of AUD 500 /kw installed capacity and short-run marginal costs of AUD 40 MWh.66 Thus, a 100 MW gas turbine would cost AUD 50 million. In 2004, the price rose above AUD 40 /MW during 1 790 hours in South Australia. A 100 MW gas turbine would be expected to operate during those hours and would receive some AUD 29 million, after fuel costs have been paid. In 2003, the same plant would only be expected to operate during 317 hours, with resulting revenue of some AUD 7 million after fuel. This shows that it is, indeed, risky to invest in a generation plant only for peaking purposes – particularly when expected income varies so much from year to year. On the other hand, if the investor analysis of the need for additional peaking capacity is correct there is a pricing and trading arrangement in place that rewards the high risk with very profitable rates of returns. It would take only two years similar to 2004 to finance the entire investment. 66. IEA, 2003a.

125

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Price spikes in the South Australian market raised political concerns and inquiries, prompting the establishment of a South Australian Government Taskforce, which published its report in June 2001. The report addressed issues of market power and inappropriate market rules and considered price caps and other direct market interventions. Regarding possibilities to intervene, the report reads as follows: Further, the Task Force concluded that actions to change existing contracts could seriously damage [South Australia] in terms of investor confidence, future reliability of electricity supply including new investment urgently needed in South Australia, and retail market competition.67 Thus, investor response to market prices effectively overcame the tightness in the first years of the market and the South Australian Government actively stated that it would not intervene directly – it had confidence in general market principles. The joint federal and state decision to increase the price cap to AUD 10 000 /MWh was probably also key to incentivising the necessary market response; it was a much-needed signal of confidence in the market and of the governments’ broad commitment. In Victoria, generation capacity increased by 8% or 615 MW from 1998 to 2003, primarily reflecting investor response to market prices. Victoria’s situation was similar to that of South Australia with a tight balance during peak periods in the first years of the NEM. Prior to a shortfall of capacity, due to strikes at major generating facilities in January 2000, there was a cap mechanism in Victoria, the “Industrial Relations Force Majeure”. This may have deferred investments in peak capacity up to this point in time.68 But in 2001 and 2002, some 552 MW of peaking gas turbines were added to the system in Victoria. NEMMCO issues a so-called Statement of Opportunities (SOO) every year on 31 July, with a brief update on 31 January. The SOO is a critical instrument to disseminate information and analysis on the status of key fundamental factors in the market to all market players and potential investors. It applies a very systematic and comprehensive methodology to collect, analyse and report the data in the following areas: demand information and demand forecasts including peak demand; generation capabilities that account for real conditions (including consequences of increased temperatures of cooling water during summer); and status of the transmission grid. It calculates 67. South Australia Government Electricity Taskforce, 2001, page 4. 68. IEA, 2003a.

126

5 INVESTMENT IN GENERATION AND TRANSMISSION

reserve requirements, based on all the input on demand, supply and transmission and on acceptable reliability requirements. Finally the demand, peak demand, generation capabilities and reserve requirements are used to calculate the future point in time at which maintaining the status quo would lead to a reserve deficit. The SOO thereby specifies future asset needs in terms of time, volume and location. It is then up to market participants to respond to this analysis and choose the appropriate technology. The SOO also reports on plans and ongoing work on new generation and transmission assets. When a new asset is regarded as committed by NEMMCO, it is included in the computation of future reserve deficits. Thus, investors can carefully follow expected development in the total asset base. The SOO 2004 predicted a reserve deficit of 320 MW in the joint Victorian and South Australian markets, which are closely integrated, by the summer of 2006/07.69 This area has seen little investment in new generation capacity since 2002, at which point a project was initiated to connect the NEM with Tasmania and its hydro resources. In July 2003, NEMMCO and the Tasmanian Government signed a Memorandum of Understanding to integrate Tasmania into the NEM through a new 600 MW HVDC transmission line, known as Basslink. For the South Australian and Victoria markets, this will have a similar impact as would a new generation plant; the MOU has probably deferred new investments. According to SOO 2004, the Basslink project is committed to be comissioned by summer 2005/06 and is thus already included in the forecast reserve deficit of 320 MW for 2006/07. In the SOO update from January 2005, NEMMCO reports a new OCGT project of 312 MW, which they regard as committed to be brought into commercial operation in Victoria in December 2005.70 It will thereby compensate for the forecast reserve deficit. There seems to be a will among market players to respond to market needs and the SOO dissemination process plays an important role to this end. Generation capacity in Queensland has also increased substantially, mainly in the form of coal-fired generation capacity, which is financially most suitable for base load operation. The decisions behind these investments predate the opening of the NEM. According to the NEMMCO Statement of Opportunities 2004, Queensland had significant capacity reserves for the 2004/05 summer peak – almost 1 000 MW more than the required level for reliability. A new 750 MW coal-fired plant is 69. NEMMCO, 2004. 70. NEMMCO, 2005.

127

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

regarded as committed to go into commercial operation in 2007. The SOO 2004 projects a small capacity shortage for the 2009/10 summer peak and increasing thereafter. At present, most of the major generation companies in Queensland are state owned, although there has been private sector investment past 1998. In spite of prices having been high at times (in fact, the highest on average in the NEM during the summer of 2004), there has been very little investment in New South Wales. A full 97% of all principal generation capacity in New South Wales is state owned. The state government blocked a private company proposal to build a small coalfired power station – on greenhouse policy grounds. This was the first time such policy action was taken in Australia. Figure 13

Price duration curves - National Electricity Market 2004 AUD/MWh 100 90 80 70 60 50 40 30 20 10 0 100% 90%

80% NSW QLD

70%

60%

50%

40% SA SNOWY

30%

20%

10%

1%

VIC

Source: NEMMCO

Figure 13 shows that even if prices are peaking at very high levels at times, prices remain below AUD 100 /MWh in 95% of the year’s 17 520 half-hourly prices and below AUD 50 /MWh a full 90% of the

128

5 INVESTMENT IN GENERATION AND TRANSMISSION

time. In 2004, the annual average price across the entire NEM was AUD 38 /MWh. In any case, all market players – from generators to retailers – supplying household consumers are free to manage price variations through contracts to hedge risks. The final price in the NEM reflects the bid of the last dispatched marginal generator or consumer. Prices actually paid for large portions of the supply have been agreed through financial contracts between generators and retailers directly. However, the NEM price is the reference because all production and consumption will be settled at the NEM price, even though price fluctuations can be managed to the preference of the market players through contracts.

Box 5 . Market-driven investment in nuclear power in Finland After almost a decade of careful planning, research and public debate, the Finnish generator Teollisuuden Voima Oy (TVO) started construction of its third nuclear unit at the Olkiluotot nuclear power plant in 2004 (the final decision to proceed was made at the end of 2003). The plant will have a capacity of 1 600 MW, generating 13.3 TWh per year with a 95% capacity factor. It is planned to go into commercial operation in mid-2009. Finland is a part of the integrated Nordic market, which offers no specific remuneration for capacity in its one-price-only pricing principle. Thus, investments in the Nordic market will be financed through the contracts traded in the market place consisting of the day-ahead spot market, the exchange for financial contracts operated by Nord Pool and the bi-lateral over-the-counter (OTC) market. Nord Pool’s day-ahead spot price is broadly regarded as the reference price in the Nordic market and is used as the reference in all other contracts with shorter or longer duration. The new Finnish nuclear unit will receive its entire remuneration from contracts in the market place and the Nord Pool spot price will be the reference. The company and ownership structure of TVO make the financing of Finland’s new nuclear reactor an interesting case concerning risk management and investment in nuclear power in competitive electricity markets. TVO is a part of the Pohjolan Voima Group, under the parent company Pohjolan Voima Oy (PVO), which owns some 60% of TVO shares. The rest are owned by the largest Finnish electric utility Fortum Oy (some 25%) and other smaller, municipally owned electric utilities. PVO was established by large industrial electricity consumers, mainly from the Finnish pulp and paper industry. These large industrial electricity

129

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

consumers still represent the largest share of the owners but today PVO is also partly owned by local municipalities – either directly or through municipally owned utilities. Thus, large Finnish industrial electricity consumers have a very important stake in the new reactor. Even though public ownership – directly or indirectly – is still below majority, there is substantial public involvement in the nuclear project in that a large minority of PVO owners are municipalities, many of which also own important shares of TVO. Fortum, the second largest owner of TVO, is controlled by the Finnish state as a majority share owner. The business of TVO is structured so that all electricity production passes from its nuclear power plants to its owners, according to owner shares at cost price.71 Ownership of TVO thus becomes a “physical hedge” towards fluctuations in electricity prices. The role and significance of such a physical hedge must by analysed in the context of other contracting opportunities offered in the Nordic market. In the Nordic market, all participants including TVO are free to make contracts of shorter and longer duration. That means that the “physical” delivery from TVO can be sold on to other market participants and that – if a price can be agreed – the risk of owning the physical assets in TVO can also be passed on to other market participants through financial contracts. The Nordic market is relatively liquid for financial contracts up to three years in advance. Within that framework, it is possible to undertake a quite sophisticated risk management strategy. Some contracts extend even further. In March 2005, one of the largest Swedish pulp and paper companies, Holmen AB, announced the signing of a 10-year contract with the Swedish utility Vattenfall. The contract commits Vattenfall to supply roughly half of the electricity demand from Holmen, some 1.5 TWh/year for the next 10 years. Holmen has its own generation capacity corresponding to roughly 1/3 of the consumption.72 The investment decision by the large industrial consumers owning large stakes in the new nuclear unit is probably an important part of their risk management strategies. As the Nordic market offers opportunities for advanced risk management, the linkage between Olkiluoto III and the electricity demand of the industrial owners should not be over stated, however. If Olkiluoto III ends up generating electricity at a long-run cost 71. TVO, 2005. 72. Holmen, 2005.

130

5 INVESTMENT IN GENERATION AND TRANSMISSION

that is above the average spot price, the owners will incur a loss compared to the alternative. If the costs are below spot prices the investment will be profitable. Even if the investment is an important part of a risk management strategy, this does not change the fact that the investment decision will only be profitable if the analysis of future demand and supply in the Nordic market is accurate and correspondingly that the analysis of the plant costs is accurate. Will the added capacity be needed and will the costs be covered by market prices? Nordel, the association of Nordic independent transmission system operators, publishes an annual report including basic supply and demand statistics and forecasts. In the annual report for 2003, Nordel reports that the forecast supply/demand balance for Finland in the winter 2007/08 is a deficit of production capacity of 795 MW.73 Nordel also projects power and energy balances three years ahead. In the projections for 2006, Nordel anticipates importing 8 TWh with normal precipitation conditions, 10 TWh with low precipitation and 18 TWh with an extremely dry year. At the same time, Nordel expects electricity consumption to increase at an average yearly rate of 1.2% until 2006.74 All in all, there seems to be good reason to believe that there will be an internal Nordic demand for the extra 13 TWh that Olkiluoto III will bring to the market. The question, then, is if Olkiluoto III will be able to compete with alternative investments in the Nordic market and imports from Germany, Poland and Russia, for which interconnections are already in place, under construction (as is a new interconnection between Norway and the Netherlands) or in the planning stages (Finland and Estonia agreed to a new interconnection with PVO being one of the joint owners). In addition, a private company filed for permission by the Finnish Ministry of Trade and Industry to build a transmission cable between Finland and Russia. If approved, the cable will be able to feed 1 000 MW into the Finnish grid. According to information provided for a joint IEA/NEA75 study on Projected Costs of Generating Electricity – Update 2005, the costs of a nuclear power plant in Finland is approximately EUR 40 /MWh.76 These are the levelised lifetime costs of the plant, assuming a capacity factor of 73. 74. 75. 76.

Nordel, 2004c. Nordel, 2003b. NEA is the Nuclear Energy Agency of the OECD. IEA/NEA, 2005.

131

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

85%, 40 years lifetime and a 10% discount rate. TVO has operated the two reactors, Olkiluoto I & II, at average capacity factors of almost 96% from 2000 to 2004. According to the IEA/NEA study, the costs for the invested capital in nuclear power plants account for a large share of the overall costs. Assuming a 95% capacity factor for Olkiluoto III would almost decrease the costs by 10%. The competitiveness of the plant will be tested against the electricity prices in Nord Pool. Figure 14 shows prices for forward contracts traded at Nord Pool for delivery three years ahead in time at the moment they are traded. Three years is the longest contract traded at Nord Pool and, thus, gives the best indication of the expected prices under normal conditions with normal levels in the hydro reservoirs. In 2002, the contract furthest ahead in time was the contract for 2005, in 2003 it was for 2006 etc. Figure 14

Nord Pool forward prices, yearly contract, three years ahead EURO/MWh 45 40 2008 contract

35 2007 contract

30 25

2005 contract

2006 contract

20 15 10 5 0

2002

2003

2004

2005

The three-year-forward prices have increased steadily since 2002, at a higher rate than can be explained by inflation. It seems to reflect the tighter supply/demand balance as forecasted by Nordel but also a whole range of other factors have an influence on the price. These include the increased costs of coal and gas and also the expected increased CO2 emission costs.

132

5 INVESTMENT IN GENERATION AND TRANSMISSION

Demand Participation as an Alternative ............................. As long as some consumers participate actively in the balancing of supply and demand, a liberalised market can provide timely and adequate incentives for investments. On the other hand, lack of participation from consumers can undermine efficiency and reliability of electricity markets, ultimately raising prices for all users. Lack of demand participation remains one of the most serious weaknesses in the development of robust electricity markets. There are good reasons to be optimistic, however. Different types of consumption have different values for the consumer. Consumers are undoubtedly, in principle, willing to shift demand as a response to price; i.e. demand is price-elastic. The main problem or barrier is that managing the response is too complicated and costly: transaction costs are too high. If investment in equipment, organisation and time to manage the response cannot be offset by the revenues from responding, consumers can not be expected to be active. Figure 15 shows the importance and effects of active demand participation. If some consumers can offset the transaction costs for some shares of their demand, they will become more responsive to prices; demand will become priceelastic, putting downward pressure on electricity prices. Demand participation will create or release resources that, in the end, can save investments in new generation plants. It can also curb market power of the generator with the last pivotal generation resources and add to overall system reliability. These factors are a major source of improved efficiency in competitive electricity markets. In a time of transition, it is likely that – at least initially – only a small share of the demand side will be able and willing to respond to price. This will push the price a little lower than the inelastic clearing price. As the market matures, the amount of demand that finds it worthwhile to enable price response will increase. This development will stabilise at a level at which the alternative new generation plant is less expensive than the marginal benefit from consumption, taking revenues and transaction costs into account. The role of a price cap becomes critical in such a process. If the price cap is below the elastic clearing price in a mature market, it is unlikely that any demand participation will be realised. Even if the price cap is between the elastic and inelastic clearing prices, a price cap may undermine the development path towards more active demand participation. Apart from the direct price effect to improve efficiency and market performance, demand participation also has important “volume effects”. Firstly, it adds flexibility that will improve system reliability: the more demand resources that can be activated with the price instrument, the less likely it will

133

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Figure 15

Demand participation improves market and system performance Inelastic demand

Price

Elastic demand

Inelastic clearing price

Lower prices – Improved efficiency – Reduced market power Elastic clearing price

Less demand – Improved Reliability – Less Investment

Supply Quantity

be that the system operator will be forced to non-voluntary shedding of load. It is also likely that clear price signals that reach the consumer will lead to real conservation of electricity and in any case more efficient use of energy. Investment in equipment, organisation and time that enable demand participation will, in most cases, improve the general level of energy management. Knowledge and control is a first prerequisite to efficient energy use, which will be in the financial interest of any consumer. Transaction cost is the critical parameter for activating demand participation. Transaction costs come from upfront investment and day-to-day management. In general, transaction costs per MWh will decrease with the level of the consumption, i.e. transaction costs per MWh will be higher for small consumers and lower for large consumers. Initially, it is expected that demand participation will mainly be of interest to the largest industrial consumers. That being said, transaction costs are also highly dependent on the character of the consumption. Some industrial processes are highly automated, so no significant additional investment and management costs will be required. On the other hand, the marginal benefit of consumption may be relatively high. Some residential consumption, such as electric heating or air-conditioning, will 134

5 INVESTMENT IN GENERATION AND TRANSMISSION

require relatively high up-front investment costs to allow for full automation of demand participation. Once in place, however, the actual marginal benefit of consumption in a specific hour may be relatively low. Reaction times and duration times also have a large impact on marginal benefits and transaction costs. Some industrial consumers may shift load at a relatively low loss of benefit, if it is with a day’s notice, but the loss of benefit may be significantly higher if it is only with an hour or 15 minutes notice. On the other hand, a residential customer with electric heating and enabled to automatic load shifting is able to react on a very short notice without losing greater comfort than if it is with a day’s notice. Different types of demand response will be best suited for different segments of the market. Some are best suited for day-ahead commitments, some are able to respond in real-time and others are willing to commit to a contract of longer duration to stand as operational reserve. One of the operational challenges of active demand participation is that it is not possible to meter load shifting in the same way as electricity generation. Other methods must be considered to measure these “nega-WATTs”. One approach is simply to regard resources from demand response to market prices as an additional statistical parameter to take into account when making demand forecasts. A retail company can offer a spot price product to its customers and then monitor the effect this has on the demand by its customers in total. Over time, the retail company will start to learn how to account for price when making demand forecasts, in the same way as it is today taking temperature and other similar factors into account. The retail company is then able to take this knowledge into account when submitting bids to the market. Policy makers, regulators and system operators have important roles in making the access for demand participation as easy as possible, thereby minimising the transaction costs for consumers. Low price caps and political intervention in the market will deter demand participation. All market rules and trading arrangements should be considered with the demand side in mind, including design of products and services to meet TSOs need for reserves and other ancillary services. All TSOs in the Nordic, PJM, Australian and British markets allow consumers to bid into some of their competitive purchases of reserves and other ancillary services. An interesting example of innovation in the wholesale market is a new bidding rule introduced by the Nordic power exchange (Nord Pool) to meet requests from consumers. Nord Pool now allows bids in the day-ahead spot market that specify a price and a volume, but not a time. This implies that the 135

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

bidder commits to reduce load within any one hour in which the spot price is higher than the bid price. Thereby the consumer avoids having to give bids for several hours, with the risk of having to reduce load in several hours. Retail companies are the link between the consumer and the wholesale market and, therefore, have a key role in making demand participation a reality. They will also be dependent on appropriate market rules and trading arrangements, but they will have to make the necessary product innovations that can link consumers to the wholesale market. Apart from the innovation triggered by competition, several countries have also considered how to contribute to this innovation process through R&D programmes and pilot projects. In PJM and many of its control areas, there are programmes to promote demand participation. In the Nordic region, TSOs are preparing action plans to support the development of demand participation.77 The 2002 review of the Australian National Energy Market stresses the importance of demand response and suggests some action points.78 A project on demand response resources under the IEA Implementing Agreement on Demand Side Management is exploring options and best practices of enabling demand participation. There may also be a role for local network companies. Congestion points in distribution grids are not priced using normal locational pricing – not even in the nodal pricing system in PJM. A local network owner may consider giving consumers special incentives to shift load during peak conditions to avoid additional investment in the distribution grid. In some circumstances this can have a value that is significantly higher than the value reflected in the normal day-ahead and real-time pricing. Table 5 shows demand participation that has been contractually committed by TSOs, observed in the market and assessed to be additionally available at a minimum. The numbers cover demand participation in many different forms, with many different duration times and many different reaction times. It is difficult to observe demand participation with certainty. The point with liberalised markets is that decisions are made on a decentralised level. The demand side is free to participate in the market through bids. In Australia and in England & Wales, very low levels of demand participation have been observed with certainty: yet a participation corresponding to 1% of peak demand is an important contribution. It does, however, seem unlikely that all consumption in Australia has a marginal benefit corresponding to the most extreme price spikes during peak-load hours. There seems to be barriers to enable demand participation. The extreme peaks in Australia are driven by air77. Nordel, 2004a. 78. Council of Australian Governments Energy Market Review, 2002.

136

5 INVESTMENT IN GENERATION AND TRANSMISSION

Table 5

Demand participation: Committed by system operators with minimum additional assessed and observed demand participation PJM

Nordic

England and Wales

Australia

Alberta (Canada)

Committed

3 598 MW

2 075 MW

4 329 MW*

NA

NA

Percentage of peak load

3%

3%

7%*

800 MW**

334 MW**

800 MW

1%

1%

7%

Additional – 7 964 MW** 10 000 MW*** observed and assessed Percentage of peak load

7%

15%

* Britain ** Observed, *** Observed and assessed Source: PJM, OFGEM, NORDEL, NEMMCO.

conditioning. It seems unlikely that there are no large industrial consumers who would find it profitable to reduce some consumption during the most critical hours and save AUD 5 000 /MWh to AUD 10 000 /MWh. This lack of participation may be a sign of inadequate competition in the retail market. Another aspect that may contribute in Australia is the market design. There is a strong focus on real-time market operation, which provides a very flexible environment for supply but may make it more difficult for the demand side to participate. To many large industrial consumers that may be price responsive, the cost of responding on very short notice may be very high. In markets with day-ahead spot fixations, it is possible for consumers to lock in the “trade” one day in advance and they are, thereby, left some time to plan the reduction. Data from the Nordic market are drawn from a Nordel study on demand participation. The 10 000 MW is an assessment of the minimum capacity on the demand side that should be able to respond to prices without discouraging loss of benefit or too high transaction costs. The Nordic drought in 2002/03 is one of the clearest examples of demand participation. Residential consumers in Norway and industrial consumers in Norway and Sweden reduced consumption significantly over a period of several months. Another important example from the Nordic market is a tight situation during a cold spell in Sweden in February 2001, during which prices spiked to unprecedented levels and peak demand was reduced with some 2% to 3% 137

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

during the critical hours compared with expected levels.79 This made a significant difference for the total system balance. The Swedish TSO also appealed to electricity consumers to reduce load through national media, which may also have had an effect on consumption. In the Nordic and British markets, and in PJM, significant capacity from the demand side is committed to operational reserves. Committed capacity in PJM also includes other programmes intended to promote demand participation. According to a recent review, the observed demand participation in PJM comes from similar programmes with PJM’s member utilities. Responses by PJM customers included additional demand response resources adding up to a total of approximately 15% of PJM’s peak demand.80 In the British market, National Grid has been very successful at attracting the demand side in their purchase of reserves and ancillary services. The more than 4 000 MW of contracted and committed demand is for frequency and fast reserves. This market has also undertaken pilot projects with aggregation of demand for other types of reserves. In Alberta, Canada, a liberalised and competitive market has been operational since 1998. Since then, some 3 500 MW of new generation capacity has been built in a one-price-only market. In a recent study on the demand responsiveness to electricity prices, the Alberta Department of Energy conducted a survey among the 15 to 20 largest industrial consumers. The results indicate that some 630 MW were already responsive to real-time prices, without participating in any specific programme. Among the surveyed consumers, some 800 MW was able to respond to price, representing almost 7% of peak load in 2004.81 Transaction costs and marginal benefit from electricity consumption set the framework for potential demand participation. This will vary with market models but also according to the characteristics of electricity consumption in different countries. How much demand participation is required to make markets robust and competitive? This depends largely on the characteristics of the electricity system and the level of competition. Researchers at George Mason University (Virginia, USA) conducted economic experiments on demand participation by letting students trade electricity in a realistic experimental environment with assumed generation and demand. In some trials, generators had market power; in others they had no market power. Some trials allowed consumers to respond to price; others did not. In the trials 79. Nordel, 2004b. 80. PJM, 2005b. 81. Alberta Department of Energy, 2005.

138

5 INVESTMENT IN GENERATION AND TRANSMISSION

with demand participation, it was assumed that up to 16% of peak load could respond. Demand participation lowered prices significantly, particularly peak prices and particularly in the trials in which generators had market power.82 In a small system in which a peak plant is the pivotal supplier, just 50 MW to 150 MW can make a big difference to the outcome. In South Australia and Victoria peak demand is projected to increase by 2.9% in 2004/05. The difference between summer peak under normal weather conditions and extreme weather conditions corresponds to 9% of peak demand. Demand participation corresponding to just 5% of peak demand would add significant flexibility to the South Australian and Victorian markets.

Planned Investment and Capacity Measures ...................... Two of the most serious concerns in liberalised electricity markets are the threat of abuse of market power and the lack of demand participation. Effective competition and demand participation are two critical factors in establishing a robust market framework for timely and efficient investment. In several markets, these concerns have motivated specific capacity measures that provide extra incentive for investment in generation capacity. The point of departure for these measures is that the necessary level of generation capacity required to supply demand is decided centrally, for example by the system operator. This is fundamentally different from the one-price-only markets in which the necessary level of generation capacity is a matter for the decentralised decision process within the market itself. Two approaches have been taken to implement capacity measures. One is to give a direct up-lift on the price through a capacity payment. This was used in the England & Wales pool until it was replaced with NETA in 2001. The capacity payment was computed for each hour and the point of departure for the computation was the value of lost load (VOLL), set to GBP 2000 /MWh in 2001 and thereafter set to increase with the inflation rate. For each hour, the probability of shortage was computed, based on the capacity available to generate, and the assessed peak load. The capacity payment was calculated as the VOLL multiplied by the probability of losing load. This implied that the capacity payment was low in hours with plenty of capacity surplus and high when the balance was tight. This made the capacity payment highly dynamic; it automatically adjusted to the level of investment. While theoretically ideal within a supply focused framework, it turned out to be vulnerable to gaming by dominating players. Much of the abuse of market power in the England 82. Rassenti, Smith & Wilson, 2001.

139

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

& Wales Pool took place in this segment of the market; withholding capacity resulted in high capacity payments. It was this gaming that triggered the change into NETA. Capacity payments are also used in Spain but the payment is fixed for all hours regardless of the actual supply and demand balance of the hour. It is adjusted on an annual basis. The other approach to capacity measures is volume based instead of price based. Adequate generation capacity to meet projected peak demand is assessed and turned into an obligation, i.e. retail companies supplying consumers are obliged to contract for an amount of generation capacity that meets a certain percentage of contracted load plus a reserve. The advantage of a volume-based system is that it can be based on a cap-and-trade system, where capacity can be contracted and traded using a competitive market mechanism. PJM, the New England ISO and the New York ISO have established such capacity measures based on market mechanisms and known as installed capacity markets (ICAP). The capacity markets in these ISOs are not identical but the main principles are similar. In PJM, load serving entities are given capacity obligations and generation capacity is given credits corresponding to the installed capacity. Load serving entities must document that they have contracted a sufficient amount of credits to fulfil their obligation on a daily basis. Credits can be traded in the market, PJM organises auctions for the capacity credits and self-fulfilment is also accepted. Unfulfilled capacity obligations are penalised. In case of unfulfilled obligations, PJM purchases the missing credits at a high price, until now at a price almost 10 times the normal market price. Load serving entities that have not fulfilled the obligation are charged with this high price. The ICAP markets are accompanied with price caps of USD 1 000 /MWh in the normal energy market to limit abuse of market power. Concerns about market power and lack of demand participation were not the only motivations for introducing ICAP obligations and markets in PJM. When utilities were unbundled and competition was introduced, incumbent generators were concerned about stranded costs. They had invested in new generation capacity to meet peak demand under the previous regulated environment. Opening up the market also implied allowing competition from neighbouring jurisdictions. The concern was that competitors, particularly those from jurisdictions that had not yet opened their markets, would be able to bid into the market at marginal costs while incumbents also had to consider recovery of invested capital in generation capacity. The solution was to guarantee recovery of invested capital through a capacity measure. The model of capacity obligations had been used in PJM before competition was introduced, so it was a natural choice. 140

5 INVESTMENT IN GENERATION AND TRANSMISSION

In the northeastern United States, capacity markets face a variety of problems and challenges, and are adjusting continuously. One of the more serious problems is the implications of cross-jurisdictional trade and competition. Neighbouring jurisdictions with fundamentally different approaches to remunerating generation capacity will inevitably create some market distortions. The biggest problem has been that ICAP markets have, so far, been prone to abuse of market power. As with the England & Wales capacity payment, it seems to be too easy to game the balance of demand and supply when both are well known. Bidding volumes above the required amounts will lead to very low or even zero prices, so there is a great incentive to withhold capacity. PJM assesses the capacity market in their annual report on the state of the market. In the 2004 report, they conclude that “market power is endemic to the structure of PJM capacity markets” even though market results for 2004 were regarded to be competitive.83 The demand side can also participate in supplying capacity credits. In PJM approximately 1% of the 3% committed demand participation as a share of peak load (mentioned in Table 5) comes from capacity credits on the demand side. The amount of demand participation in the ICAP market will, however, always be restricted to those resources that fit to the exact requirements needed to receive a capacity credit. Functioning of capacity markets is at the centre of much research in electricity market design. Several researchers have suggested adjustments to improve competitiveness and efficiency in capacity markets. The New England ISO proposed a major reform of their capacity market that would make capacity obligations and credits more reflective of locational needs and give credits based on actual performance, rather than just availability.84 Other researchers propose solutions that introduce more of the features used in modern financial derivates for risk hedging. A capacity obligation can instead be thought of as an obligation to contract a call option. The penalty can be thought of as the strike price of the call option, where a buyer can demand that the call option goes to delivery. Longer duration times and time-lags are also proposed to allow new investors to sell call options to help finance a plant before it is even built. The longer duration and time-lags would help new-comers curb market power.85 There has been theoretical research that effectively illustrates the key issues in capacity measures.86 What research shows is that, if a price cap to curb market power is in place, a capacity market can restore efficiency compared to a one83. 84. 85. 86.

PJM, 2005b. Cramton & Stoft, 2005. Oren, 2005. Joskow & Tirole, 2004.

141

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

price-only market by compensating for the investment incentives lost by the price cap. One of the most critical assumptions is that consumption is homogenous, meaning that all consumers and all their uses are identical in the sense of having similar profiles and similar marginal benefits from consuming. Consumers can be divided into classes but for every division a new set of regulatory instruments must be added to restore efficiency, thereby increasing complexity. Consumption is, quite clearly, not homogenous, nor can it be divided sensibly into two or three similar classes. The result is that capacity measures will induce a loss of efficiency resulting from the lack of demand participation. Capacity measures are introduced to allow price caps to curb market power, assuming that demand is in-elastic and thereby accepting a loss of efficiency. One-price-only markets aim for efficient and competitive outcomes resulting from active demand participation. What seems to divide the approaches is that, in one-price-only markets, the lack of demand participation is regarded as a transitional phenomenon that will pass with proper price signals and adjustment of market design to minimise transaction costs. In markets with capacity measures, the lack of demand participation is regarded as an inherent market failure that has to be addressed. Contracting for operational reserves and other ancillary services could also be regarded as capacity measures in one sense and most system operators find such contracts necessary. The system operator contracts generation capacity to be ready when called upon. Plants need to be warm and ready to increase or decrease generation at short notice in the event of sudden large imbalances, even if the probability that it will be necessary to draw on them is very low. This is required to meet operational challenges but there are close linkages with the normal energy market. Payment for operational reserves becomes an element of the incentive package for new investment. Sufficient demand participation in the real-time market, which is able to respond on very short notice, should make operational reserves obsolete. The way in which operational reserves are acquired and activated has implications for the price signals in the normal market. Operational reserves must be carefully defined and linked with the management of real contingencies. It should be as easy as possible for the demand side to participate. The reserves should be bought using a competitive process to ensure that when sufficient real-time demand participation makes them obsolete, competition will drive the contract price to zero. Operational reserves should only be activated in case of real contingencies. Loosely defined operational reserves with loosely defined rules for activation can severely distort investment incentives. If the need for operational reserves is assessed from a normal evaluation of the balance between demand and supply,

142

5 INVESTMENT IN GENERATION AND TRANSMISSION

contracting of operational reserves will slide into a normal capacity measure. If operational reserves are activated at low prices during normal tight situations, where there are no contingencies, they will dampen the market price that should have signalled the tight situation. As a consequence, investments will not be made and the TSO is forced to further increase acquired operational reserves. They are no longer operational reserves but, simply, reserves. Active demand participation and market power remain serious concerns in all markets which could be a good reason to introduce capacity measures. On the other hand, capacity measures have proven to be prone to market power themselves and one-price-only markets have so far not failed to deliver. Competitive one-price-only markets have not led to excessive market power and evidence so far indicate that clear and transparent cost-reflective price signals do give incentives for timely and adequate investment. There is some evidence emerging indicating that demand will, indeed, participate if the framework is right and when consumers have something to respond to, i.e. uncapped prices. The merits of capacity measures still needs to be proven and the failure of one-price-only markets have still not materialised. This suggests that capacity measures should be avoided, particularly if the general policy approach is only to regulate the industry when necessary. Capacity measures are still at the heart of the debate in North America, Europe and Asia. This raises additional issues concerning the co-existence of various approaches to incentivising investment. In a EU directive concerning security of electricity supply soon expected to enter into force, it is proposed that EU member states can use both a one-price-only-market and an obligation-based measure (a capacity measure) to maintain balance between electricity supply and demand.87 Regardless of the individual merits of the two approaches, their co-existence raises issues of efficient incentives for investment in an internal market and free-riding across borders. Capacity incentivised through special capacity measures in one jurisdiction will distort investment incentives in neighbouring jurisdictions with one-price-only markets. In relation to the development of an internal market on the Iberian Peninsula, the market implications of the Spanish capacity payment is one of the critical issues. This could call for a stronger role for the European Union in guaranteeing consistency across the intended internal European market.

87. European Parliament, 2005.

143

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Investment in Transmission Networks ................................ Effective performance of transmission networks shapes the operation and development of efficient wholesale and retail electricity markets, particularly the emergence of effective regional markets. Effective performance of transmission networks also underpins efficient development of inter-regional trade, which contributes to effective competition and reduces scope for market power abuse. It also improves capacity utilisation (both generation and networks) and helps to defer expensive investment in generating capacity. In addition, more effective regulation reduces excess transmission capacity. As a result, transmission systems are becoming more stressed and prone to congestion. Other policy priorities have led to increased use of less greenhouse-intense forms of distributed and intermittent generation, making the operation of transmission networks increasingly complex and expensive. At the same time, more stringent environmental approval processes are making timely network investments more difficult to realise. The combination of these factors creates new challenges for maintaining reliable transmission services and maximising transmission network performance. Transmission owners and operators have far less certainty about the demands that will be placed on their networks and less capacity to undertake integrated planning and development of transmission networks as a whole. Transmission system operators’ capacity to manage system balancing and efficient network development is, thus, greatly reduced, particularly within regional markets incorporating multiple owners and operators. At the same time, persistent concerns about regulatory and policy uncertainty hold the potential to magnify the business risks faced by transmission network owners, with the potential to discourage investment and encourage more conservative network performance. The situation can be further complicated when vested interests serve to weaken incentives for efficient network performance and where jurisdictional boundaries unduly affect the emergence of efficient market structures or design. The rate of investment in transmission systems appears to be slowing, particularly in North America and Europe, and especially in jurisdictions with liberalised electricity markets and reformed economic regulation. IEA projections of the need for investment in transmission systems in OECD countries are USD 498 billion for the period 2003-30, corresponding to roughly 25% of total investment needs in the electricity sector. There is much focus on the role of interconnections to enhance competition and the efficient use of resources. In 2002, the European Union even adopted a target for minimum transmission capacity on interconnectors between member states, to support the development of integrated and reliable regional 144

5 INVESTMENT IN GENERATION AND TRANSMISSION

electricity markets. The priority projects pointed out by the European Commission in its TEN-E project can be seen in connection with those targets.88 The European Commission projects that the total costs of the interconnections across the prioritised bottlenecks will be some EUR 5 billion until 2013. IEA projections for transmission investments in the European Union between 2001 and 2010 are EUR 49 billion, illustrating that investment in transmission interconnectors to relieve important congestions is only a share of total required investment. Following a surge in investments in new transmission lines in the 1960s and 1970s, the increase has slowed down. Figure 16 shows the annual increase in lengths of transmission lines in selected European countries. Figure 16

Annual average increase in length of 220-400 kV transmission lines in 16 European countries 7% 6% 5% 4% 3% 2% 1% 0% 1975-1979

1979-1989

1989-1999

2000-2003

Source: UCTE and Nordel.

A similar development has occurred in the United States.89 Transmission investments in the United States fell to unprecedented low levels in 1997 and 1998 but have now recovered to levels seen in the early 1980s in terms of USD real. Investment in transmission networks has also become an important policy issue. It has been suggested that chronic underinvestment in transmission capacity, especially interconnector capacity, was an important contributing 88. European Commission, 2004a. 89. IEA, 2002.

145

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

factor leading to the cascading blackouts experienced in 2003 in Europe and North America. Thus, reinforcement of the main transmission network ought to be a top priority to ensure reliability and efficient market development. Investment in new transmission lines could add strength and resilience to transmission systems, enabling them to face contingency events. But such investments may have the character of gold-plating. Changes in the way transmission assets are maintained and systems and markets are operated may be a far more cost-effective way to enhance system security. This issue is discussed thoroughly in a recent IEA publication.90 Concerns about under-investment in transmission network infrastructure raise the issue of whether electricity market reform, particularly structural reform of the electricity sector and reformed regulatory arrangements, create undue barriers to efficiently timed and sized development of transmission networks. Recent investment trends may be partially explained by pre-existing excess capacity in transmission networks and more effective economic regulation, which introduced new financial and regulatory disciplines on transmission investment projects and decision-making processes. The question now is whether regulators have struck the right balance between passing efficiency dividends to consumers and ensuring adequate returns for efficient transmission projects. The challenge of investment in transmission in liberalised markets is to exploit the price signals for efficiency that effective markets produce. At the same time, effective regulation must be maintained that properly accounts for costeffectiveness, reliability concerns and the role of transmission interconnectors in enhancing competitiveness. This all takes place during a period during which electricity systems are changing fundamentally; liberalised electricity markets are in a transition and, due to insufficient unbundling, vested interests are still affecting the decision process. The goal of market design that creates cost-reflective and transparent locational price signals is to set a price on the value of transmission capacity. With a fine tuned approach, such as nodal pricing, the signals are precise; with a zonal approach, they are less precise. Appropriate market design is a key parameter in creating incentives for efficient transmission investment, particularly for interconnectors. For example, inappropriate regional pricing models that unduly mask intra-regional transmission congestion, or do not provide access to accurate and timely information about the performance of transmission networks, can increase investment risks and create a potential barrier for alternative investments to alleviate congestion. 90. IEA, 2005.

146

5 INVESTMENT IN GENERATION AND TRANSMISSION

Some market models and regulatory arrangements may also create perverse investment incentives, reflecting inherent conflicts of interest. For example, it is possible that arrangements allowing network owners to retain congestion rents resulting from congestion on interconnectors may encourage network owners and operators to refrain from efficient investments to alleviate transmission bottlenecks in an attempt to maximise rents. Such arrangements may encourage network owners to transfer the effect of intra-regional congestion to network boundaries, also in an effort to capture additional congestion rents and defer costs of managing internal congestion through redispatch. In some cases, ineffective unbundling can encourage transmission network owners to discriminate in favour of incumbent generators – at the expense of efficient network investments to maximise trade and competition. Such behaviour would not only discourage efficient network investment but may have the potential to distort efficient operation and development of regional electricity markets. One approach to investment in transmission interconnections is to simply let it compete with generation on equal terms. Interconnections are, after all, the alternative to generation investment. The concept of merchant interconnectors competing with generation, demand response and other network owners to deliver services across regional electricity markets would be financed by congestion rents. An investor in a merchant interconnector would size and locate the investment so that price differences between the two ends of the interconnector multiplied by flow can finance the project. Greater reliance on competitive transmission services, based on competition between merchant lines, may support more effective use of price signals to strengthen incentives for efficient transmission network performance. Such an approach may be more responsive to electricity markets, particularly in the context of interregional congestion management. It may also help to strengthen marketdriven signals for efficiently timed, sized and located new investment, supporting inter-regional trade. Together, these factors may help increase the economic benefits derived from electricity market liberalisation and allow a more coherent and transparent approach to managing the commercial risks associated with transmission network operation and development in competitive electricity markets. There are a number of real-world challenges to this concept, however. Transmission investment involves economies of scale and high fixed costs. This may imply that the most efficient investment response to inter-regional congestion could completely eliminate physical constraints and the related price differentials between regions or nodes, thus removing the underlying cash flows required to fund investment. Development of transmission systems 147

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

is also driven by reliability requirements and such requirements can only be ensured as a regulated investment. Transmission investment motivated by reliability and system security requirements will still compete with merchant lines and generation. This adds considerable regulatory uncertainty to merchant investments. Several merchant interconnectors currently operate in Europe, North America and Australia. Those commissioned in the era preceding electricity market liberalisation tend to operate in a manner that raises concerns about the potential for merchant interconnectors to support market power abuse. However, evidence from Australian NEM, suggests that merchant interconnectors responding to market price signals could provide an efficient means of helping to integrate network services with competitive electricity markets. The Australian example also highlights some of the challenges. Two merchant projects were completed in Australia at the outset of the market when the need for interconnections was relatively clear and seemed quite certain to benefit from interlinking diverse generation portfolios. Directlink was completed in 2000, connecting Queensland and New South Wales. The Murray link was completed in 2002, connecting Victoria and South Australia. In addition, a regulated transmission project was proposed to connect New South Wales and South Australia by TransGrid, the transmission company owned by the NSW government. The profitability of this project changed fundamentally with Murray link, which illustrates some of the challenges in creating a regulatory framework that enables the co-existence of merchant lines and regulated lines in the same system.91 Regulation easily fails to create a sufficiently stable and predictable business environment to create efficient incentives for merchant lines. Murray link has since been transformed into a regulated interconnection. To date, transmission networks have typically been treated as natural monopolies that should be physically separated from the other components of the value chain and separately regulated. Prior to the introduction of competition reform, transmission network regulation was typically based on some form of cost-plus methodology, which allowed network owners to pass through to customers all costs approved by the regulator. This form of regulation provided little incentive for efficient operation or investment and, in some cases, allowed network owners to pass on the risk and cost of poor investment decisions and inefficient operations. Electricity market reform is generally accompanied by regulatory reform. Non discriminatory access provisions were introduced to facilitate efficient 91. Littlechild, 2004.

148

5 INVESTMENT IN GENERATION AND TRANSMISSION

competition and trade. At the same time, incentive regulation, typically built around CPI-X price or revenue caps, was introduced to provide a quasicompetitive discipline to encourage more efficient operation and development of transmission networks. There is some concern that regulators are too focused on reducing operating costs, at the expense of incentives for effective operational performance and efficient maintenance and investment. This can lead to operational decisions with substantial efficiency losses and financial costs for consumers and market participants. Potential examples include removing major transmission lines from service for maintenance during peak periods, inadequate maintenance and investment in transmission networks, and transferring intraregional congestions to interconnectors. Creating regulation for efficient transmission networks is inherently challenging, particularly in terms of properly integrating risk management and asset availability in the regulatory incentive package. Transmission maintenance planning should be conducted with the costs and benefits expressed in the electricity market in mind. Risk management is also, to a large extent, connected to the discussion of firmness of financial transmission rights (FTRs), as discussed in Chapter 3. If network users are able to buy firm FTRs, their value increases as a risk management tool for market participants. Utilisation and, hence, value may be further enhanced if transmission service users are able to make a contract one day, one month or even one year ahead. On the other hand, the length of the commitment also increases the risk of not being able to fulfil the contract. This comes at a cost for the body, often the system operator, who underwrites the contract due to the consequent need to re-dispatch generation. FTRs can only be firm if those bodies taking the risk, such as transmission owners or system operators, are able to manage the risk. In the Nordic market, all users share the risk. The Nordic TSOs are allowed to pass on all costs of re-dispatch to grid users for firmness committed one dayahead and can use congestion rents to pay the cost of re-dispatch. Collection of congestion rents and passing on costs of re-dispatch are two approaches that can be used in the incentive package to ensure efficient utilisation. There is much debate about firm FTRs. Some traders advocate for annual and monthly firm FTRs; some network owners point out that they are not willing to bear the risks of such commitment. An owner of a commercial and competitive merchant interconnector would find the commercially sensible balance. In an environment of regulated transmission ownership and vested interests, risk management of transmission assets must be an integrated part of the regulatory framework, particularly if the independence of TSOs is questionable. In the England & Wales market, a regulatory change in 1994 made re-dispatch costs a 149

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

part of the incentive package for National Grid. The availability of transmission assets improved markedly and re-dispatch costs decreased correspondingly. Uncertainty may be magnified in the case of transmission network investments designed to reinforce network reliability. Emerging regional markets have the potential to fracture responsibility for maintaining reliability across many network owners and operators. Although regulations may require individual transmission owners/operators to meet certain minimum reliability and safety standards, they are not usually directly exposed to the financial consequences of catastrophic system failure in electricity markets. Regulated processes to assess proposals for reliability augmentations have proven problematic in practice. Reliability-based investment proposals, like other regulated network augmentations, are typically developed by network owners and subject to regulatory approval. However, the regulator can be at a considerable disadvantage in assessing such proposals, particularly where it is reliant on information submitted by the network service provider that is proposing the investment. Cost-benefit assessment in these circumstances can be problematic and open to dispute, possibly leading to delays and uncertainty. Access to accurate and reliable information about the operational condition of the network is crucial to help improve the effectiveness, timing and credibility of such processes. Some regulators are exploring options to improve assessment processes, including the use of probabilistic analysis to analyse the potential financial implications of network failure on electricity markets. Probabilistic assessment is currently used in PJM and in Australia (VECorp), and has been proposed in New Zealand. New transmission network investment is also affected by other approval processes including construction, siting and environmental approvals. Greater environmental sensitivity, combined with increasing local community opposition, sometimes slows approval processes, making licenses more difficult and time-consuming to obtain. Inefficient and inconsistent approval processes can create uncertainty and regulatory risk, which may undermine timely and appropriately sized investment responses. To date, incentive regulation has been very successful at driving down transmission network operating costs and passing savings through to consumers, usually in the form of lower tariffs. However, concerns are emerging about the wider operational and investment incentives created by regulation. Underinvestment in transmission networks and interconnectors has been raised as a critical issue in the United States and appears to be an emerging issue in Europe. 150

5 INVESTMENT IN GENERATION AND TRANSMISSION

A particular challenge for regulators is to achieve an appropriate balance between incentives to minimise costs and network charges on the one hand, and augmentation of transmission networks on the other. This challenge is likely to be magnified by the inherent information asymmetry associated with regulation of transmission network services. In Australia, a newly implemented regulatory test assesses costs and benefits, both in terms of enhanced trade and system security. Experience with economic regulation is also raising concerns about regulatory uncertainty, which principally relates to the amount of discretion afforded regulators in their interpretation and application of regulatory provisions. One suggested solution is to tighten legislative frameworks to reign in regulatory discretion by clearly prescribing the nature, scope and limits of discretionary powers, thereby minimising uncertainty and regulatory risk that may otherwise have the potential to hinder efficient market operation and development. This may prove a substantial challenge in practice, especially where the boundaries of regulatory and policy responsibility may be overlapping and somewhat ambiguous. Experience to date suggests that, in practice, economic regulation in electricity cannot be simply divided between policy and administration. Issues of detail matter in electricity markets, and regulatory interpretation and decisions on matters of detail can critically influence strategic policy outcomes. Transparent, consistent and balanced regulatory decision-making and outcomes are required to help address perceptions of regulatory risk and uncertainty. Challenges associated with minimising regulatory risk and uncertainty are greatly magnified in regional markets that span several jurisdictions. Typically, policymakers have sought to minimise these risks by adopting consistent regulatory rules across jurisdictions within regional markets. However, concerns about the potential for inconsistent interpretation and application of regulatory provisions amongst different regulators within regional markets remain and continue to create regulatory risk and uncertainty, particularly in relation to the operation and development of interconnectors linking different jurisdictions within regional markets. Policymakers and regulators are taking steps to address these risks. For example, regulators in Europe, North America and Australia have created multi-lateral fora to pursue greater co-ordination and harmonisation of regulatory interpretation and administration. Some regulators sought to improve transparency and certainty in relation to regulatory interpretation and application by issuing non-binding statements of regulatory intent. Some jurisdictions have gone further by consolidating regulatory functions across 151

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

several jurisdictions into a single regional regulatory body that reports to member governments collectively. A recent example is the new National Energy Regulator established by the governments participating in the Australian National Electricity Market.

Co-ordination of Transmission Investment ......................... In a regime of vertically integrated utilities, transmission investment is an integrated part of an overarching investment plan. The need for transmission lines can be carefully incorporated into a plan for investment in generation capacity. System security issues are a priority concern, particularly when considering development of interconnectors between jurisdictions. When responsibilities are unbundled in liberalised markets it raises questions about the efficacy and compatibility of such planning regimes. This is particularly true in a competitive market environment characterised by independent, decentralised decision-making: network owners and operators can no longer accurately predict or control key decisions determining future network use, especially in larger regional markets. It also raises practical issues associated with coordinating planning processes in larger regional markets where responsibility for planning is shared across several network owners/operators. Furthermore, existing processes may unduly favour investment in network solutions over potentially more efficient generation and demand response options, reflecting the vested interest of transmission owners/operators responsible for planning processes. Regulators responsible for managing this risk may be at a particular disadvantage in attempting to moderate this behaviour in practice, particularly when they must rely on information and technical advice provided by incumbent network owners/operators who have a pecuniary interest in the outcome. This could prove to be a particularly difficult issue in the context of assessing reliability-based investment proposals. Fractured ownership responsibility for transmission networks across regional markets may also affect incentives for transmission network investment. Where there are many transmission network owners and operators within a single regional market, it is unlikely that any of them will have a clear incentive to undertake transmission network development from a market-wide perspective. Unclear or shared ownership responsibilities, combined with practical challenges of coordinating investments amongst different network owners, have the potential to weaken incentives and increase risks, particularly for interconnector investments. Regulatory uncertainty may increase these risks, especially if construction approvals and regulated returns 152

5 INVESTMENT IN GENERATION AND TRANSMISSION

are subject to different regulatory processes and decisions on either side of an interconnector – which are likely to be further magnified where interconnectors cross national boundaries. Greater coordination becomes essential in liberalised markets, especially amongst network owners and operators within larger regional markets, to ensure that planning processes reflect a more holistic and regionally integrated perspective. After experiencing some of the challenges these issues raise, several markets took the approach of transforming planning processes into a tool for accurately and transparently informing market participants and regulators about transmission network use and resources, as well as emerging trends. The intent is to improve independent, decentralised decision making within electricity markets and to support more effective regulatory assessment of transmission network investment proposals. The annual Statement of Opportunity published by NEMMCO for the Australian National Electricity Market is one such example. As a response to a NEM review in 2002, NEMMCO extended its Statement of Opportunity with an Annual National Transmission Statement (ANTS). The Statement of Opportunity carefully describes and analyses demand and supply fundamentals. The ANTS analyses national transmission flow paths, forecasts interconnector constraints and identifies options to relieve constraints. It is then up to market participants, including network owners, to respond to the needs by adding either generation or transmission assets. In the Nordic market, the opportunities, challenges and caveats of transmission planning in liberalised markets is emphasised in the co-ordinated transmission plans developed by Nordel. All Nordic TSOs have a long history of transmission planning. Now Nordel has developed a co-ordinated plan in which important congestion points in the Nordic transmission system are identified and assessed by modelling the Nordic market. Economic modelling has been used to assess the value of relieving single congestion points. A characteristic of the Nordic market is that large hydro-electric resources, located primarily in the north, are connected with demand and thermal generation plants in the south and on the European continent. In such a system, relieving one congestion point often just has the effect that the congestion is moved to another point without benefiting substantially from the added transmission capacity. The economic modelling used in the Nordel transmission planning identified flow paths or series of bottlenecks. Benefits of relieving congestion points and paths are assessed for both generators and consumers, and for both the whole Nordic region and country by country. This 153

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

points to some of the political challenges of transmission investment and again stresses the importance of impartiality and independence from incumbent generators to enable efficient transmission investment. The Nordel transmission planning exercise of 2004 identified five major transmission investment projects. Some will create losses for generators in some Nordic countries and for consumers in other Nordic countries, but in aggregate they will add considerable economic benefits to the joint Nordic economy, also outweighing the assessed investment costs of approximately EUR 1 billion. A process towards final project decision, including financing the investments, is currently evolving.

154

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL? The perfect market is a theoretical ideal rather than an achievable goal. No markets are perfect and all markets are in constant evolution. Competition is rarely perfect. Policy and regulation change. Technology and taste change. Innovation pushes development forward in steps that can create sudden big changes. Electricity markets are no different in that sense. No electricity market is perfect, or is ever likely to be so. It is more relevant to assess the performance of liberalised electricity markets by comparing against the alternatives. Does a less-than-perfect liberalised electricity market perform better than a less-than-perfect vertically integrated and regulated system? There are some serious concerns in liberalised electricity markets. Competition and cost-reflective pricing does not come easy; it can only be accomplished with a high level of commitment by policy and decision makers. Even with the necessary commitment and determination, markets will still take a long time to prepare and even longer to be fully implemented and developed. Electricity market liberalisation is a process. With that in mind, are there still some factors in liberalised markets that simply cannot be managed in a satisfying way? Are there some factors in the economic optimisation of markets that will never be taken into account, regardless of how competitive these markets are and how well they are designed? It has been argued that such so-called “market failures” are inherent in electricity markets and can be found in many parts of the value chain. For example, some argue that electricity markets are inherently uncompetitive: economies of scale will inevitably lead to large companies that will have the power to collude. This claim seems to contradict the recent development of more varied and more flexible units of generation technologies, including combined cycle gas turbines (CCGTs). If large economies of scale are apparent in a market, it may rather suggest that regulation of the market failed to remove the barriers that allow small companies to enter and operate. Another claim is that the demand side is, in fact, inherently inelastic – that demand-side participation cannot be expected, and that is a serious market failure. Considering that all uses of electricity will have different values for all users, the thought that all demand is inelastic is certainly counter-intuitive. In the end, it may not be possible to lower transaction costs sufficiently to make sufficient demand participation profitable, but given the high focus on supply in the current market organisation this conclusion seems premature. In any case, innovation and technology are constantly lowering barriers. The danger is that efforts to 155

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

involve the demand side may fail if markets are designed assuming that demand is inherently inelastic. Another concern is the prospects of research and development in liberalised electricity markets, when the priorities of market players change with the emergence of competitive pressures. To date, private commercial companies have not undertaken sufficient R&D in areas that are mainly of public interest, such as renewable energy and energy efficiency. Experience shows that R&D requires policy attention within areas of public priority, as is required in nonenergy sectors when markets fail. That being said, one general lesson from liberalisation processes in other sectors is that competition also creates incentives for active innovation. In two areas, alleged market failures in the electricity sector seem more related to the inherent characteristics of electricity rather than the result of underlying regulatory failures. Electricity is a unique product. Modern society is increasingly dependant on electricity so the demand for reliability of supply is high. At the same time, it is a highly complex product mainly due to the fact that it cannot be stored. How unique is electricity and at what point might markets fail to secure a reliable supply? Also, electricity generation with most conventional technologies has impacts on the environment and the climate, yet markets do not properly account for external environmental costs – a classical market failure. With policies now in place to address external environmental effects, what are the implications for liberalised electricity markets and how might liberalised electricity markets contribute to meeting new challenges such as climate change? When they arise, real market failures due to the inherent characteristics of electricity must be addressed with policy measures that try to compensate for the failure or internalise the costs. Again, the appropriate policy instrument may not be found by referring to some theoretic ideal, but rather by comparing alternatives. Some market failures may leave the electricity market less than perfect and, thus, policy options designed to address the failures should be assessed on their ability to improve the overall outcome.

Reliability of Supply in Liberalised Electricity Markets ...... In liberalised electricity markets, the intent of price signals is to direct all actions toward efficient outcomes. Thus, it is reasonable to ask: Are there any points along the value chain – from electricity generation to transport to consumption – at which price signals can be expected to be insufficient? Are 156

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

there any parts of the value chain in which markets fail to produce efficient outcomes? The value chain can be broken up into three parts, each of which addresses different critical components of a reliable supply of electricity (Figure 16). Another way to put it is that the concept of reliability of supply needs to be “unbundled” to be analysed in an unbundled sector.

Figure 17

The value chain for reliable electricity supply: Energy security, adequacy and system security

Reliability of electricity supply

Energy security

Adequacy

Coal, natural gas, uranium...

Generation capacity, transmission and distribution networks

System security

Operation, control, contingency management...

First of all, there must be a secure fuel supply as input for power generation. Secondly, adequate generation and network assets are needed to transform the fuel into electricity and transport it to the consumer. Finally, because electricity cannot be stored, it is a significant challenge to operate the system securely – to actually keep the lights on. In which parts of this value chain are markets likely to fail?

■

Energy Security of Supply for Electricity Generation Security of upstream energy supply for power generation is a matter of how well fuel markets function. Electricity is mainly produced from coal, natural gas, uranium and hydropower. Other renewable energies are attracting more and more attention. While oil was a major fuel for electricity generation until the oil crises of the 1970s, its share has since decreased to insignificant levels

157

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

in most countries. The global coal market is competitive, resources are abundant and it can easily be stored. Uranium can also easily be stored. Considering the minor role of oil in electricity generation the focus ends with natural gas. Natural gas resources are relatively concentrated in few countries, storage is relatively costly and it has been grid bound even if this constraint is changing with the development of liquefied natural gas (LNG). Natural gas has an increasing importance in electricity generation. Due to their low capital costs and short construction times, combined cycle gas turbines (CCGTs) are now a preferred technology in many situations in many liberalising electricity markets. Natural gas also has an environmental comparative advantage to coal with its lower emission of greenhouse gases per generated MWh of electricity in CCGT use. Markets for natural gas are also liberalised or liberalising in several IEA member countries. In the European Union, natural gas markets are liberalising in a process that parallels electricity market liberalisation, even though gas market liberalisation lags behind somewhat. Natural gas markets are also under strong influence from the development of liquefied natural gas (LNG), which can potentially add significant flexibility to regional and global gas markets. A recent IEA study on security of gas supply in open markets describes the status of gas markets and discusses the role of LNG.92 As is the case with electricity, liberalisation of gas markets is a long and complex process; in several IEA member countries there are still many challenges to overcome. Natural gas reserves are concentrated in relatively few non-IEA member countries; this raises considerations of geo-political dimensions concerning the functioning of the upstream part of natural gas markets. Even though this concentration raises concerns similar to those associated with oil reserves, one major difference is that natural gas is only one of several fuel options for power generation. The possibilities for substitution of fuel for electricity generation in the mid-term are much greater than is the case for substituting oil with other alternatives for transportation. Clearly, there are reasons for concern regarding the security of gas supply: will sufficient investments be made upstream, will markets be successfully liberalised down stream and will LNG reduce dependence of specific suppliers and increase competition? In relation to liberalised electricity markets, the issue is whether investors fail to take security of gas supply considerations into account. CCGTs are a preferred technology, but the profitability of a CCGT investment will depend on a secure supply of natural gas which is competitively priced. At present, lack of market liberalisation for natural gas is 92. IEA, 2004a.

158

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

regarded as a barrier for investment in CCGTs.93 Investors in CCGTs consider the full natural gas value chain, from well-head to generation plant. Some electric utilities choose to enter the natural gas business as a part of an overall strategy that also includes security of gas supply considerations.94 If investors fail to fully account for the economic consequences of gas supply disruptions or market manipulation – considering both probabilities and costs – policy intervention may be appropriate. A potential risk is that the total electricity generation system could become too dependent on natural gas through increased use of CCGTs. Two main policy options have been applied to address concerns for gas supply security: one is to promote alternative technologies to ensure diversification; the other is to focus on the natural gas market itself. Policy makers can improve the functioning of the down-stream gas market through liberalisation and regulation. Dialogue with supplying countries can help address up-stream issues. Application procedures for LNG and other gas infrastructure investment projects can be improved. A policy to promote alternative technologies spills directly over into liberalised electricity markets. Direct subsidies to promote specific alternative technologies or regulatory barriers to build CCGTs are intended to distort investment signals. Distortions implemented in one country will affect internal electricity markets in neighbouring countries. Are such distortions warranted and efficient? The CPB, the Netherlands Bureau for Economic Policy Analysis, is an independent governmental agency that conducts economic policy analysis. It recently conducted a cost-benefit analysis on various aspects of reliability of supply, including the case for policy intervention to decrease the Netherlands’ dependence on natural gas for power generation. Considering assessed probabilities, costs and benefits in relation to gas supply interruptions or market manipulation intervention could not be justified.95 Recognising that secure supply of natural gas is a great concern for investors in CCGTs and that intervention based on judgements by governments are not likely improving decisions alternatively made by commercial investors it seems more relevant for governments to focus on minimising regulatory uncertainty in the gas market.

■

Adequacy of Generation Capacity and Transmission and Distribution Networks Whether there is adequate generation capacity to meet all demand at all times depends on whether investments are made at sufficient volumes, at the 93. Speckler, 2005. 94. de Tomás, 2005. 95. CPB, 2004.

159

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

right location and in a timely fashion. There is no reason to believe that competitive electricity markets cannot provide incentives for timely and efficient investments. The key requirements are: real competition (including regulated third-party access to unbundled networks); cost-reflective locational pricing; and market rules that ensure transparency and low transaction costs. In addition, there is the regulatory challenge of creating a smooth regulatory framework for transparent and clear approval processes of new generation plants. At present, market players are investing in liberalised electricity markets, even without additional capacity measures. Networks tend to be natural monopolies. Markets will, in general, fail to provide signals for adequate and timely investment. Investment in networks depends largely on the regulatory framework and the incentives incorporated therein. Thus, liberalisation does not, in principle, change the situation for networks – except perhaps if they are privatised as part of the liberalisation process. However, liberalisation does increase the focus on network operating costs, and many countries have introduced models for regulation of network tariffs with strong incentives for cost cutting. This high initial focus on costs raised concerns for the quality aspect in some countries, particularly in the sense that quality is directly linked to adequacy. In several countries, models for regulating network tariffs are now being adjusted to also include direct incentives for quality – and, thereby, incentives for investment. Markets do not necessarily fail to provide signals for adequate investment in transmission networks. In fact, locational pricing should create incentives for investment in a way that treats generation and transmission assets as substitutes for one another. So far, this is one of the disappointing failures of liberalised electricity markets. Almost all investment in transmission networks remains regulated; there are only few on-going examples of so-called “merchant lines”.

■

System Security in Transmission System Operation The value of electricity to its consumers is determined by time, location, volume and quality. Liberalised electricity markets can set a price in terms of the first three factors, but quality is so far a uniform product delivered to all consumers, regardless of price signals. Technically and economically, it is still not possible to deliver electricity at varying quality, depending on the willingness to pay. All consumers receive electricity at a certain standard quality; if problems arise that force the system operator to shed load, this will happen more or less randomly without taking notice of willingness to pay. In this context, load shedding is thought of as load that is cut involuntarily after all other options have been exploited including voluntary load reduction 160

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

contracts with consumers. A full blackout happens to everybody. Quality is also a characteristic of electricity supply that everybody can benefit from without reducing the benefit for others. Just because someone pays, does not mean he will get a higher quality product or have a lower probability of being disconnected. The case is the same for consumers who decide not to pay for quality, i.e. there is an incentive to try to free ride. Consumers cannot claim individual property rights to quality of electricity supply, which makes it a classical public good. In general, markets fail to balance supply and demand for public goods at a level that makes everybody happy. Without intervention from policy makers, markets will deliver public goods, such as electricity quality, at a level that is lower than consumers would prefer. To date, establishing independent system operators is the primary means of addressing the failure of liberalised and competitive electricity markets to deliver quality. These system operators ensure quality by balancing supply and demand in real-time operation. They manage the interface between electricity as a normal private good traded in a market and the public good aspect of the product related to the quality. System operators have a large toolbox from which to draw instruments: balancing real-time supply and demand using real-time markets and other ancillary services; a fixed set of reliability criteria that determines the limits of the transmission system; immediate and transparent disclosure of critical information to the market place; skilled personnel and equipment to carefully monitor and control the system, etc. Many of these instruments have close links to electricity as a private good traded in a market. If they are not used according to very strict standards – and in very disciplined way – they will blur the interface between electricity as a private and public good. Opaque and non-transparent definition and management of congestion blur price signals: markets will respond less comprehensively and precisely, and the public good sphere of electricity will increase. In 2003/04, there were some remarkable blackouts and incidents related to system security. The most publicised events included blackouts in North America, Italy, southern Sweden, and eastern Denmark. An important incident also occurred in Australia, but was managed well enough to avoid a major blackout. In each of these cases, official investigations do not blame market liberalisation. They did, however, raise some issues concerning system security in competitive markets that must be better addressed in the future. Liberalised markets significantly increased cross-border trade, in terms of both volume and variability. These exchanges contribute to the overall benefits of liberalised markets. However, they must be carefully managed by system 161

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

operators to avoid becoming a threat to system security. The above incidents stressed the importance of co-ordination and co-operation between system operators in general. Initially, this was directed primarily through voluntary agreements in North America and Europe. After the blackouts, the United States implemented legislative changes and UCTE made their operational handbook binding. The incidents pointed to the importance of monitoring the actual fulfilment of cross-border agreements. The question is whether compliance monitoring can be effective unless these agreements are made legally binding. In reality, reliability criteria extend far beyond the role of system operators. In the North American and Italian blackouts, insufficient tree trimming along power lines played an important role in the actual failure of transmission systems. This raises the issue of including a quality assurance dimension to regulation of transmission systems, as a means of ensuring good and secure availability of transmission lines. As is evident, this relates to sophisticated regulation rather than the impact of liberalisation. A recent IEA book that focuses on these case studies also reviews and discusses the issues of transmission system security in competitive electricity markets.96 Its main conclusion is that liberalisation fundamentally changed the usage of transmission systems and management – and that operation of transmission systems still needs to adapt to these changes. What is most important is that the much-publicised blackouts and the day-today management of system operation in liberalised markets do not point to unsolvable issues. They do make it clear that management in all parts of the value chain – from generation to consumption – must be re-considered in liberalised markets, including all aspects of system operation. Very few of the issues raised with liberalisation relate to specific problems from the unbundled structures in liberalised markets. Almost all of the issues relate to aspects that would improve system operation in terms of security and efficiency, regardless of the organisation of the market. For example, cross-border trade improves the efficiency of the overall system and the usage of transmission assets. It changes the usage of the system itself in a way that forces it closer to its limits, which makes the need for reform of system operation principles more pressing. Wellmanaged co-operation between system operators is not a threat to system security but rather an important potential for improved security. Well coordinated system operation across several systems should make all systems stronger than each would have been on its own. Policy intervention from

96. IEA, 2005c.

162

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

governments may be needed to secure these obligations and to ensure that market players and independent system operators are taking steps to achieve these aims. If players are not co-operating and co-ordinating their efforts, governments must take policy action.

Addressing Environmental Issues and Climate Change ..... The environment and climate are classical public goods – to the extent that markets fail to account for the costs and benefits associated with activities that have an impact upon them. There is no reason to expect that liberalised markets will voluntarily begin to account for external environmental effects. This will only happen through policy intervention. In theory, it is straightforward to internalise an external environmental cost. The damaging activity can be penalised through a tax that corresponds to the cost such activity inflicts on society. However, environmental costs are very rarely computable in a way that is precise and can be commonly agreed upon. This raises a whole range of issues that will interfere with markets, including questions of international competitiveness in open markets. Environmental costs that are borne only by generators and only in some countries will give competitive advantages to economies in countries that do not consider these costs. This relates not only to competition in electricity but also transfers to the competitiveness of products produced using electricity. In that sense, differing approaches to the internalisation of environmental costs distort both liberalised electricity markets and all other global markets, particularly those with high electricity content. The issue of how to assess and internalise environmental costs is an issue for liberalised electricity markets, as it is an issue for other open markets. At present, the issue remains beleaguered by international disagreement. Again, it is more relevant to consider policy options with reference to the alternatives, rather than using some theoretic perfection as the benchmark. Generating electricity creates consequences for the environment and the climate. These should be addressed by all governments through policy initiatives. The key question for liberalised electricity markets is how various policy options will affect the market functioning and its outcomes. Policy initiatives are intended to create the exact effect warranted by the external cost. But it is clear that various policy options have different distortive effects in a liberalised electricity market. This is particularly true in regional markets in which environmental costs are addressed differently in different countries. The fundamental question is this: are there policy options that undermine the 163

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

way liberalised markets intend to create efficient outcomes, in terms of both operation and investment? Among the policy options considered and implemented in liberalised markets two general types stand out as having the highest impact: those that support, promote and subsidise specific technologies and those that impose a cost on damaging activities. The first includes R&D programmes for renewable energy and other initiatives to ease the construction of renewable energy generation plants. Such initiatives are common in many other sectors and can hardly be seen as detrimental for the functioning of liberalised electricity markets, as long as they do not involve large-scale deployment. Different countries have different requirements regarding accepted standards for generation plants. These affect the costs of generation plants and distort international competition. On the other hand, such differences are also evident in most other economic sectors, where they are generally accepted in the short run and become the subject of international harmonisation efforts in the longer term. Environmental risks and their consequences are perceived very differently from country to country, even when some of the consequences have potential impacts far beyond national borders. In some countries, the general public broadly supports nuclear power or at least accepts it; in others, it is strictly prohibited. Binding international security standards are in place, but various aspects in the approval process, in the management of spent fuel and in the management of long-term waste disposal vary from country to country. All in all, this creates differences that influence the costs of nuclear power from one country to another. Many aspects related to nuclear power have a broad policy interest within and across countries. However, as long as conditions for nuclear power are transparent, related policy initiatives are unlikely to undermine the functioning of liberalised electricity markets. Support for R&D and differing legal conditions for different technologies both affect competitiveness in open electricity markets, but only indirectly and on a level comparable with many other economic sectors. Liberalisation significantly alters decision-making processes, transferring them from the centralised and planned structures within vertically integrated systems to the decentralised decisions made by individual market players, based on direction of market prices. Direct financial subsidies to specific technologies, such as renewables and nuclear power, perhaps in specific volumes and perhaps even on specific locations, fundamentally undermine the decentralised decision-making process in liberalised markets. Alternative policy options must be applied to internalise external environmental costs while at the same time maintaining the decentralised decision process. When 164

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

taxes are not viable, the least distortive policy option is the cap-and-trade market mechanism. As with the tax option, cap-and-trade systems still focus on the actual harmful activity but it does not stipulate or dictate the solutions as direct subsidy schemes do. For example, a cap-and-trade system is stipulated for greenhouse gases within the flexible mechanisms of the Kyoto Protocol. The European Union implemented a cap-and-trade system for CO2 emissions. During the first half of 2005, trade of CO2 emission permits increased dramatically. Liquidity and price of the permit market reached a level at which this market is commonly accepted as one of the explanatory factors for price increases in most European electricity markets. National governments determine the conditions for entering the European CO2 emission permit market for each individual generator. These conditions differ significantly and leave the European electricity market severely distorted. However, the marked-based mechanism itself has already demonstrated its ability to create the intended and anticipated incentives. The first experiences with CO2 emissions trading are explored in a recent IEA publication.97 Most IEA member countries support renewable energy. It, too, can be supported in a cap-and-trade system. Several member countries, including the United Kingdom, Australia and Sweden, as well as several U.S. states, introduced systems based on quotas for renewable energy and tradable renewable certificates, known as renewable portfolio standards (RPS). Policy intervention determines volumes but investors decide on technology, time and place. If the justification for supporting renewable technologies accrues to the fact that these technologies does not emit greenhouse gases, full implementation of a cap-and-trade system for greenhouse gases will make an RPS obsolete. Hence, if the justification is based on the assumption that renewable energy is a cost-effective way to reduce emissions, an increasing CO2 emission price will eventually reduce the price of renewable certificates to zero. RPS systems will only be robust, cost-effective policy options with minimal distortion of electricity markets if they are international and consistent with efficient operation of open electricity markets. A purely national RPS will be vulnerable to political changes and will lose the tremendous efficiency potential of international trade. Most renewable resources are highly locational and, thus, significantly more cost-effective if deployed where resources are abundant or diverse, e.g. wind turbines where it is windy, bio-mass plants where there is a lot of cheaply available biomass and photovoltaics where it is sunny. 97. IEA, 2005d.

165

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Many renewable energy technologies operate in smaller unit sizes than conventional generation technologies and tend to be more distributed and connected to the grid on lower voltages. Most newer technologies, such as wind power, are also highly intermittent. In the traditional vertically integrated and regulated electricity industry there is a natural focus on large conventional generation plants and systems are developed and operated to meet both the needs and constraints of these technologies. Moreover, the costs are passed on directly to consumers, so there are no built in incentives to enhance transparency. With such a focus it becomes difficult to fully assess the costs and benefits of distributed and intermittent renewable technologies. But without an overview of costs and benefits, it is impossible for policy makers to fully assess the merits of renewable technologies, including their cost-effectiveness in terms of mitigating greenhouse gas emissions. Liberalised electricity markets change the framework for distributed and intermittent renewable technologies on several accounts. Cost-reflective pricing and competition determine the real costs of managing intermittency. The larger focus on risks, coupled with the value of financial and operational flexibility, in liberalised markets sheds new light on the merits of small, distributed generation resources. Cross-border trade makes it possible to use local renewable resources much more effectively, for example, by exploiting a diverse resource base in wind. Some of the costs and benefits of renewable energies accrue to the impact that distributed and intermittent technologies have on network development and operation. On one hand, high shares of intermittent wind necessitate network extensions; on the other, wind turbines and other distributed resources connected to local networks may save some costs to grid losses and some investment in transmission networks. Creating incentives for network tariffs that reflect these costs and benefits is a challenge for effective economic regulation of networks. Economic impacts of integrating wind power into electricity grids are discussed in greater detail in the annex of a recent IEA/NEA study on projected costs of generating electricity.98 There is intense debate over the merits of renewable energy technologies, particularly wind power which seems to be one of the more cost-effective options in some circumstances. For those countries with the highest shares of wind power as a percentage of consumption (e.g. Denmark, Germany and Spain) there is concern for the security of national and neighbouring electricity systems. In addition, there is increasing concern about the costeffectiveness of wind power when all costs of managing intermittency are 98. IEA/NEA, 2005

166

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

taken into account. In liberalised electricity markets, all of the issues associated with management of intermittent renewable technologies are more – or even fully – transparent. Competitive locational marginal pricing of energy and competitive pricing of reserves and ancillary services also make the costs and benefits of intermittency and geographically diverse generation more transparent and the total costs possibly lower. Box 6 . Managing intermittent wind power in the liberalised Danish electricity market Denmark is divided into two electricity systems that are not interconnected. The eastern system is electrically synchronised with the Nordic system; the western with the continental European UCTE system. By the end of 2004, some 3 122 MW of wind power capacity was installed in Denmark. During that year, wind turbines produced 6.6 TWh, corresponding to 19% of the consumption in a year with average energy content in the wind at 90% of the normal value. Approximately 2 379 MW of the total installed wind power capacity was located in western Denmark, where turbines generated 4.9 TWh, or 23% of consumption for that region. High shares of wind power in the Danish system are at the heart of a serious debate about costs, system security, industrial policy and impact on landscape that has been ongoing throughout the past decade of wind power development. Many issues regarding the impacts on the Nordic electricity market still remain unresolved. Until recently, Denmark subsidised wind power through feed-in tariffs given for every MWh produced from wind power; the tariff was reduced in 2003 to a level that has literally halted further development. Instead, most new wind power development is now driven through politically decided tenders for off-shore wind parks and politically decided refurbishment schemes. Such an approach deviates fundamentally from the principles of decentralised decision making in liberalised markets. For example, Sweden implemented a cap-and-trade system to support wind power and Sweden and Norway decided to develop a common cap-and-trade system. Some harmonisation issues related to policy - and its impact on broad investor confidence – in the Nordic market remain open. However, there is no doubt that the Nordic electricity market has been very important for the development of wind power in Denmark. This is evident when one compares data on hourly wind power production in western Denmark and the hourly Nord Pool dayahead spot price for December 2003 (Figure 18).

167

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Figure 18

Wind power production and Nord Pool day-ahead prices in western Denmark, December 2003 MWh/h

Euro/MWh 100 90

2500

80

2000

70 1500

60 50

1000

40 30

500

20 10

0

0 01

04 07 10 13 Nord Pool Spot – Left axis

16

19

22 25 28 31 Wind power – Right axis

Many factors influence electricity prices in western Denmark; Figure 18 shows clearly that wind power is one of them. The price tends to drop when wind power production is high and rise to prices in neighbouring markets when production is low. Transparent spot and balancing prices have significantly improved understanding of costs and benefits of wind power. Moreover, the liberalised market has directed trade across the Nordic region in a way that is sufficiently dynamic to enable integration of Danish wind power across a larger area – even while the direct costs of Danish renewables policy remain with Danish electricity consumers. Figure 19 shows the impact this price development has on trade, as well as hourly wind power production and hourly aggregate exchange with Norway and Sweden. Exchange between the Nordic countries is entirely determined by the price of electricity for every hour. Figure 19 shows that the price directs wind power production to Norway and Sweden when the level is high. This is a sensible outcome, particularly considering that Norway has large hydro reservoirs, which are excellent for storing electricity. Danish electricity consumers pay the direct costs of Danish renewables policy, but the costs are minimised through international trade.

168

6 WHEN DO LIBERALISED ELECTRICITY MARKETS FAIL?

Figure 19

Wind power production in western Denmark and trade with Nordic neighbours MWh/h 2500

Import to DK west

2000 1500 1000 500 0 –500

01

04

07

10

13

16

19

22

25

28

31

–1000 –1500 Export from DK west Trade – DK West/Nordic

Wind power – DK West

169

ANNEX 1

ANNEX 1: BRITISH ELECTRICITY TRADING AND TRANSMISSION ARRANGEMENTS Privatisation was the main driver for the launch of the liberalisation process in the late 1980s. The process was triggered by the publication (February 1988) of a government white paper, Privatising Electricity. The white paper outlined a number of principles including: competition is the best guarantor of customers’ interests; regulation should be designed to promote competition, oversee prices, and protect the customers’ interests in areas where natural monopolies will remain; and security and safety of supply must be maintained. The white paper stipulated an eight-year transition period. Prior to the publication of the white paper, the telecommunications and gas businesses had been privatised and a reform of the state-owned coal business had been initiated. The white paper was followed up with a legal framework enabling the effective launch of liberalisation; the Electricity Act 1989. The most critical initial parts of this framework were the establishment of an independent regulator and a restructuring of the sector. Under the Electricity Act 1989, the post of Director General of Electricity Supply (DGES) was created to regulate the natural monopoly businesses. The government established the Office of Electricity Regulation (OFFER) as an independent body to be headed by the DGES. Its role was to promote competition and regulate the network businesses. OFFER established a new regime for network regulation based on a cap of prices, with a focus on cost-cutting. It was the Retail Price Index (RPI-X) regime, which limited the development of prices by a cap consisting of the increase of the RPI minus a specified demand for improvement in efficiency. This model was to be copied by many regulators in the following years. Prior to the Electricity Act 1989, the entire electricity industry in England & Wales was state owned. Generation and transmission were managed by the Central Electric Generating Board (CEGB) and distribution was managed by 12 area electricity boards. Under the Electricity Act 1989, the entire sector was reorganised, corporatised and eventually privatised. CEGB was split into four companies. All transmission assets and responsibility for system and market operation was transferred to National Grid Company (NGC). All generation assets were divided between National Power (40 conventional power stations with 30 GW capacity), Power Gen (23 conventional power stations with 20 GW capacity) and Nuclear Electric 171

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

(12 nuclear power stations with 8 GW capacity). Initially it was intended to place the nuclear assets with National Power but these plans were abandoned at a late stage on financial advice saying that the nuclear assets were not saleable at a reasonable price. The 12 area electricity boards were corporatised into 12 regional electricity companies (RECs). Ownership of NGC was transferred to the 12 RECs. The vesting of CEGB and the 12 area electricity boards took effect on 31 March 1990. A series of three-year contracts were signed. National Power and Power Gen signed contracts for purchase of coal from the still state-owned British Coal at a price above world market prices. These generators held contracts with the 12 RECs enabling a pass through of the extra cost to captive customers. On 1 April 1990, retail competition opened to the 5 000 consumers with a load higher than 1 MW and the Pool commenced operation. In Scotland, the North of Scotland Hydro-Electric Board was also restructured under the Electricity Act 1989 into Scottish Hydro-Electric and Scottish Power. Both were privatised as vertically integrated regulated utilities in June 1991. The first round of privatisation happened in December 1990 with the 12 RECs. The government still kept a golden share that hindered mergers and takeovers. These golden shares lapsed in April 1995 and immediately triggered the first trading activities with REC assets. In March 1991, the first 60% of National Power and Power Gen were privatised and the remainder were privatised in March 1995. In December 1995, NGC was sold by the RECs through a flotation of NGC shares on the stock market. Initially, the pumped storage capacity of CEGB had been allocated to NGC but this was sold when NGC shares were floated. The final round of privatisation happened in July 1996. The nuclear power stations in Nuclear Electric were transferred to British Energy (the seven modern reactors, excluding the seven old Magnox reactors) and British Nuclear Fuels Ltd. (the seven Magnox reactors). Also two reactors within Scottish Electric were transferred to British Energy. British Energy was then privatised. The three-year vesting contracts and the five-year golden shares in RECs gave some time to develop the operation and functioning of the market. When the three years ran out, prices started to increase and the DGES raised concern over the development of competition in the market. Several initiatives were taken that were to have significant influence on the development of competition and the structure of the industry. Competition and divestiture were the only sources for increased competition. In anticipation of the importance of new entry by independent power producers (IPPs), it was accepted that RECs could enter into long-term power purchase agreements 172

ANNEX 1

from IPPs and IPPs could sign long-term gas contracts. This ignited the “dash for gas” that was to add significant capacity of new combined cycle gas turbines (CCGTs) and decrease the market power of the incumbent generators. The dash for gas more than halved the size of the remaining deep coal mining industry. When margins between fuel costs and Pool prices did start to increase after 1993, the largest incumbent, National Power, agreed to accept divestiture of 6 000 MW generation capacity within two years under the threat of being referred to the Monopolies and Mergers Commission, the competition authorities. A price cap was also introduced during the two years. When the golden shares of the RECs lapsed, these were bought by other regulated UK utilities and two US utilities. Power Gen and National Power also gave bids for two RECs, driven by the desire to vertically integrate with retail supply businesses. The bids were referred to the Monopolies and Mergers Commission, and in the end blocked by Department of Trade and Industry. The response from the two incumbents was to accept the divestiture of an additional 4 000 MW generation capacity each, in exchange for the admittance to acquire RECs. This happened in November 1998. Several large foreign utilities have entered the British market and some have left again. National Power, with its reduced generation capacity but with its REC arm, was bought by RWE. Power Gen was bought by German E.ON. French EDF bought two large RECs including London Electric. An increasing reliance of gas and increasing problems for the British coal industry losing market shares to gas raised concerns with the new Labour government in 1997. In December 1997, the government imposed a moratorium on building new gas-fired power stations, which halted development. The Department of Trade and Industry estimates that the moratorium delayed the building of 5 200 MW of new gas-fired generation capacity. Retail contestability was extended in steps. The barrier was lowered from 1 MW to 100 kW in April 1994, thereby increasing the number of contestable consumers to 45 000. The remaining 22 million consumers were given access to switch supplier between September 1998 and June 1999. This ended the process that the Electricity Act 1989 had triggered 10 years earlier. But already before that, the first considerations to reform the market which had been implemented were starting to emerge. In October 1997, the Minister for Science, Energy and Technology asked the DGES to review the electricity trading arrangements in the pool. The merits of gas compared to coal in the pricing principles of the Pool were one of the reasons for initiating the Pool review. OFFER published its Pool Review in July 1998, which included 173

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

a far-reaching critique of the existing mandatory pool system and gave recommendations for new electricity trading arrangements (NETA) based on a voluntary approach. The government accepted the recommendations in a white paper with conclusions from the review in October 1998, which triggered the reform of the Pool. NETA replaced the Pool on 27 March 2001. In August 2000, NGC established a separate company to manage the new Balance and Settlement Code, the rule book for NETA that all market participants must sign. The new company, ELEXON, a subsidiary to National Grid Company operates and settles the balancing market in NETA. A new Utilities Act 2000 was passed in July 2000. The main points were: replacement of separate gas and electricity regulators with one regulator, the Office of Gas and Electricity Markets (OFGEM); legal separation of retail supply and distribution; and enabling the implementation of NETA. The Utilities Act 2000 forced the legal unbundling of the two vertically integrated Scottish utilities and discussions about the extension of NETA into Scotland commenced. After a decade with focus on cost-cutting in distribution networks concerns were raised for the quality of service. In 2003, OFGEM introduced a new aspect in the RPI-X regulation of prices, specifically targeted at mimicking incentives for efficient levels of quality. By measuring the quality of service in terms of the number of interruptions of supply, the duration of these interruptions and the information service provided in connection with these interruptions, failures to perform according to acceptable standards could lead to a reduction of prices of up to 1.75%. Scotland was integrated into NETA in April 2005, which is hereafter referred to as the British Electricity Trading and Transmission Arrangements (BETTA). NGC was appointed as market and system operator for BETTA with ELEXON in the role as market operator.

Framework for Governance.................................................. The Utilities Act 2000 establishes the framework for the British liberalised market. It requires independence of transmission ownership, system and market operation from the competitive businesses in the sector. This independence is through ownership unbundling. It requires legal independence of distribution networks from generation and retail supply. The Utilities Act gives British electricity consumers the freedom to choose retail supplier. It gives the authority to regulate the electricity industry to the Gas and Electricity Market Authority (GEMA) with the regulatory body Office of Gas and Electricity Markets (OFGEM). 174

ANNEX 1

OFGEM’s role is to regulate the electricity and gas industries in protecting consumers by promoting competition where this is possible and through regulation where this is necessary. More specifically, OFGEM focuses on making electricity and gas markets work effectively and intelligently regulating monopoly businesses. OFGEM is governed by GEMA, a body appointed by the secretary of state. National Grid Company has been licensed by the GEMA to operate the British electricity system and, through its wholly owned but independently managed subsidiary ELEXON, to operate the balancing market. OFGEM regulates the performance of National Grid Company and approves the grid code that sets the rules for interaction with the British electricity system. The balancing mechanism is the only official market place in BETTA. It is functioning in accordance with the Balancing and Settlement Code (BSC). One of the reasons for abolishing the Pool was that the governing structures made it very complicated and tedious to make any changes to the rules. As a consequence, a very strict and clear procedure for managing changes of the BSC has been established, in which the GEMA has a stronger influence. In accordance with these modification procedures, ELEXON established a BSC panel to ensure an impartial enforcement of the code and to give impartial advice on any proposed modifications. The Panel consists of representatives appointed by GEMA (appoints the chairman), industry, EnergyWatch and National Grid Company. EnergyWatch is the independent watchdog for electricity and gas consumers. ELEXON advises the BSC Panel. Any modifications to the BSC are subject to approval by GEMA, which must justify its rulings transparently. OFGEM has the responsibility to give licenses to distribution network operators, who are then subject to OFGEM RPI-X and quality of service regulations. In connection with quality of service regulation, annual reports are published on the quality of service. OFGEM also has the responsibility to monitor and secure the supply of gas and electricity. In its efforts to fulfil this responsibility, OFGEM publishes biannual reports on the security of supply of gas and electricity.

Basic Market Design Features .............................................. The initial trading arrangement in the Pool was inspired by the principles applied in the calculation of dispatch by the former CEGB and the dispatch in the Pool was calculated with the same software, GOAL. Generators gave bids and specified start-up costs and other technical constraints. Bids were ordered 175

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

in ascending order and the software calculated the dispatch that would meet forecast demand, subject to transmission constraints. The marginal bid set the system marginal price to be paid to all dispatched generators. Grid losses and congestion management (locational aspects) were not reflected in the pricing; they were only taken into consideration in the actual dispatch. Demand was forecast by National Grid Company, so it was a one-sided market assuming no demand response to prices. Bids and prices were calculated for every half hour of the following day. It was a day-ahead trading mechanism, in which bids had to be submitted by 10:00 a.m. one day ahead, and dispatch was calculated during the afternoon before 5:00 p.m. The system marginal price was only one element in the price paid by consumers and the remuneration received by generators. The most significant addition was the capacity payment paid to generators for keeping generation capacity available to the market. It was awarded as a price-per-availablegeneration capacity but the price was calculated dynamically from an assessment of the reserve margin, the probability of losing load and an assumed cost of the lost load for every half hour. Hence, the price increased with decreased reserve margins, also if this was the result of withholding generation capacity. This system for capacity payment was a sincere effort to reflect to true value of capacity for every half hour, but it also proved to be prone to manipulation in a market with dominating generators, as was the case in the England & Wales market in the mid-1990s. In addition, an uplift was charged from consumers to pay for transmission losses, ancillary services, reserves and other necessary services to operate the system. All incumbent generators possessed the GOAL software, which gave them considerable power to maximise profit through their bids, rather than bidding the true competitive cost parameters of their plants. Apart from the problems with market power abuse, there were also a range of other critique points. It was only a one-sided market, which made demand participation difficult. The rules for trading in the Pool were set in the Pooling and Settlement Agreement, a multi-lateral agreement signed by all market participants. It was very difficult to make the necessary changes to this agreement, allowing the market to develop. At the initiation of the Pool review in 1998, a liquid financial market also failed to develop, mainly because the underlying market was uncompetitive. The approach taken in NETA, later extended to BETTA, was fundamentally different. Instead of basing the trade on obligatory trade in the Pool, trading was to be entirely based on voluntary bi-lateral trade divided into four market segments. The intention was that a forward market for standardised financial contracts should develop for the long run, covering periods up to several years 176

ANNEX 1

in advance. A market for bilateral contracts traded in the OTC market should develop for the medium and long run. A short-term OTC market for bi-lateral contracts should develop for the last 24 hours before gate closure. All these markets were entirely voluntary and were intended to develop automatically, driven by the need from market participants. The only market that market participants are forced to link into is the balancing market, operated by National Grid Company and ELEXON. Gate closure is one hour before real-time operation. All registered market participants are obliged to submit schedules (Final Physical Notification) before gate closure and will be held financially responsible for any deviations from the schedule. Schedules are submitted both for consumption and generation, making it a real two-sided market. Actual trade in the balancing mechanism is still voluntary. All market players are free to manage unexpected deviations in generation and demand. The balancing mechanism is an adjustment market. Market participants can submit bids to the balancing markets with prices and volumes that they are willing to increase and/or decrease generation and/or demand with. NGC calls on the cheapest bids to balance the overall system physically. The purchase of balancing services deviates fundamentally with all other similar markets on one critical point. The price paid to those that are called on to deliver balancing services are pay-as-bid or discriminatory pricing, meaning that the prices bid by each individual market player is also what he will receive if called on. This is fundamentally different from the marginal pricing principle used in all other electricity markets. There has been extensive research into the merits of these two pricing principles. The philosophy with pay-as-bid is that only those generators (or consumers) that are able to deliver at lower cost than the most expensive will receive this lower price. Those that buy the service will benefit from this technological comparative advantage. Average prices are thereby thought to become lower. The problem is that those with the comparative advantage are not necessarily willing to pass this advantage on to rate payers. To avoid that, they may want to bid in at prices that are as close to, but slightly below the most expensive resources needed. This puts pressure on them to focus more on the costs of their competitors, rather than their own costs. As a result, average prices may not be lower, and may even be higher as a consequence of speculation. Research has been inconclusive but has illustrated that the alleged merits of pay-as-bid markets are overstated. Market participants do not disclose their true cost structure when bidding into pay-as-bid or discriminatory auctions. Those that cause the overall system imbalance between scheduled and actual operation are financially responsible for the costs of those imbalances. The charges for these individual imbalances have changed since the 177

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

implementation of NETA. In the current pricing principle, those that have imbalances in the same direction as the total system imbalance are charged with the weighted average price of those that were called on by NGC to physically balance the system in each half-hour period. Those that have imbalances in the other direction, and who thereby have helped the overall system imbalance by chance, are charged with a spot reference price. The spot reference price is the day-ahead spot price from the privately owned UK Power Exchange (UKPX). Liquidity in UKPX in terms of turnover as a percentage of total British demand is very low, which could be seen as a threat to the role of this spot price to serve as a reference. Another problem that has been raised with the issue of such a dual pricing principle is that it creates an incentive to pool imbalances. The dual pricing principle generates a profit for NGC. This residual cash flow is paid back to market participants, corresponding to their proportion of total demand and generation. A merger of companies lowers imbalance charges without necessarily improving forecasting or lowering the real costs of imbalances in the system. Another related issue of concern is that it is specifically beneficial to pool generation and load to enable selfmanagement of imbalances within a company. This lowers liquidity in the “official” balancing market and may increase total system costs of imbalances in the end. More expensive resources within a company may be used to manage imbalances instead of benefiting from other cheaper resources owned by other generators. Congestions within the British grid are not priced in the balancing mechanism. There is one price for the entire BETTA. On the other hand, there are locational signals in the transmission network tariffs for connecting to and using the transmission grid. The remaining three markets that were pictured to develop automatically have so far not materialised. An illiquid spot market has developed. Initially, several projects to establish day-ahead spot exchanges were launched, but there is currently only the UKPX with very low traded spot volumes. There is not much transparency in the OTC market, but liquidity is probably relatively low. There are no standardised forward contracts traded on exchanges. Prices in NETA have been deemed close to competitive levels according to models of competitive pricing and physical assets have been traded extensively, so with that measure the British market has been liquid. With a market philosophy based on voluntary trading, transparency has been a matter for the voluntary exchanges. As these have not materialised so far, market transparency has been low. Prices and traded volumes in the balancing mechanism are easily available from the homepage of ELEXON. There are, 178

ANNEX 1

however, no requirements to make changes in the status of generation plants immediately available to the market. National Grid Company submits annual reports to the GEMA with a seven-year statement of forecast development of peak load, installed generation capacity and transmission capacity. The statements are easily available on the NGC Web site and provide an easy access to extensive fundamental data.

Market Structure .................................................................. In 2003, electricity consumption in the United Kingdom was 348 TWh of which 313 TWh was in England & Wales, 35 TWh in Scotland and 8 TWh in Northern Ireland. To serve the 26 million residential and 2.5 million commercial consumers, some 376 TWh was generated in 2003 and 2 TWh was imported. The remaining 30 TWh were transmission and distribution losses. Approximately 22% was generated on nuclear plants, 37% was generated with gas, 35% was generated with coal, 1.5 % was generated with hydro and the remaining was generated with other sources. Demand in United Kingdom peaked at 61 GW in the winter 2004/05 of which a little more than 1 GW was in Northern Ireland, outside the BETTA area. To meet this peak demand, there was an installed capacity of 75 GW of which 16% was nuclear, 37% was coal, 33% was CCGTs, 5% was hydro and the remaining was from other technologies. In addition, there is an interconnector between France and Britain with a capacity of 2 000 MW; Northern Ireland and Scotland are interconnected with 500 MW; Northern Ireland and the Irish Republic is interconnected with 600 MW, and Scotland and England are interconnected with 2 200 MW to keep BETTA united. By mid-May 2004, the three largest generation companies registered by Department of Trade and Industry statistics were British Energy with 11.5 GW of nuclear capacity (16% of total installed capacity), RWE Innogy with 8 GW coal, oil and CCGT capacity (11% of total installed capacity) and E.ON UK with 7.6 GW of installed coal, oil and CCGT capacity (10% of total installed capacity). Other large companies include Scottish Power, Scottish and Southern Energy, EDF Energy and American Electric Power, all with 5% to 6% shares of total installed capacity. An additional 37 companies were operating generation plants in United Kingdom. While market shares in generation are very widespread without any dominating players, the concentration in the retail market is higher with six large companies. By the end of 2003, the largest retail suppliers were British 179

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Gas (24 % of the market), Powergen (21% of the market), npower (15% of the market), EDF Energy (14% of the market), Scottish and Southern Energy (14% of the market) and Scottish Power (11% of the market). The remaining percentage was shared between other small retail suppliers. Comparing to the original distribution companies linked with the 12 area electricity boards and the two Scottish boards, there has been a significant concentration. Several of the retail suppliers are also major generators. British Gas has benefited from marketing electricity and gas jointly.

Further Reading .................................................................... Background and development of the electricity supply industry is described in IEA country reviews. The joint reviews of the United Kingdom since late 1980s constitute a comprehensive description of the development and of governing structures. The Web sites of Department of Trade and Industry (www.dti.gov.uk), OFGEM (www.ofgem.gov.uk), National Grid (www.nationalgrid.com) and ELEXON (www.elexon.co.uk) include many descriptions, information, statistics and rules. Substantial research has been undertaken in analysing the functioning of the Pool, the functioning of NETA, the development of competition and the development of the structure of the sector. Hunt (2002), Evans & Green (2003) and Newbury (2005) are some examples.

180

ANNEX 2

ANNEX 2: THE NORDIC ELECTRICITY MARKET The first step towards a competitive internal Nordic electricity market was taken with the approval by the Norwegian parliament of a new electricity reform act in June 1990. The act came into force on 1 January 1991, and introduced regulated third-party access, freedom of choice of retailer for all electricity consumers and unbundling of the transmission grid and system operation. On 1 January 1992, the state owned utility Statkraft was broken up into two independent state owned companies, Statkraft with all generation assets and retail contracts from the former company, and Statnett with all transmission assets (85% of total Norwegian transmission assets) and responsibility for system operation. In 1993, the Nordic power exchange was established as an independent company under the name Statnett Market AS. It established price quotation on a day-ahead basis and it established the world’s first exchange-based trade with futures contracts in July 1993. An important milestone for the development of the retail market was the reduction of a standard fee for retail switching from NOK 4 000 (app. EUR 500) to NOK 200 (app. EUR 25) in 1995. From 1997, there were no fees for switching and from 1998 every consumer could switch supplier with one week’s notice. A second decisive process towards an internal Nordic market was initiated with the approval of a new electricity reform act by the Swedish parliament in May 1992, which established the framework for initiating competition on wholesale level and set a plan for additional steps to create competition. It had been preceded by steps towards the introduction of competition already in 1991. With effect from 1 January 1992, the State Power Board was transformed into a state-owned limited liability company, Vattenfall AB, and the transmission system, including cross-border interconnectors, were separated out into a state entity, Affärsvärket Svenska Kraftnät. Svenska Kraftnät operated the transmission system and was also instructed to promote competition. The most important reforms were laid out in the Electricity Act passed by Parliament in October 1995, with effect from 1 January 1996. The act required legal separation of generation and network operation. It gave all consumers the freedom to choose supplier. In the wake of the act, it was decided to establish a common Norwegian and Swedish electricity exchange under the name Nord Pool. It was established with day-ahead price quotation from 1 January 1996 and extended with trading in financial futures contracts in 1997. Nord Pool ASA was established as a Norwegian company with 181

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

headquarters in Norway and owned by the two transmission system operators, Statnett and Svenska Kraftnät with 50% shares each. The liberalisation process in Norway and Sweden ran almost in parallel, but actual reform to introduce competition proceeded at a faster pace in Norway. One of the important drivers for regulatory reform in Sweden was the relatively severe economic recession at the time. Consequently, there was a high focus on reducing electricity costs, which were critical considering the high share of heavy, energy-intensive industry in Sweden. The next country to be integrated into the Nordic market was Finland, a process initiated with the approval of the new Electricity Reform Act that came into force on 1 June 1995. It gave regulated third-party access to the grid to everybody except consumers with a power requirement below 500 kW, a threshold that was removed on 1 January 1997. An independent regulator, the Electricity Market Authority, was established with the Act. A Finnish TSO, Fingrid Oyj, was established by the government in 1996 and commenced operation on 1 September 1997. Fingrid was formed after the unbundling of the transmission grids from the two large vertically integrated utilities, IVO (owned by Fortum) and PVO. Fortum was 75% state owned, a share now reduced to 59%. The ownership of Fingrid is distributed between the utilities Fortum (25% shares, 33% votes) and PVO (25% shares, 33% votes), the Finnish State (12% shares, 16% votes) and the remainder with institutional investors. The Finnish electricity act did not establish a framework for the development of trading arrangements. Two electricity exchanges were established through private commercial initiatives in 1995, EL-EX and Voimatori Oy. Liquidity never reached satisfying levels. The two exchanges merged into EL-EX in 1996. When liquidity still did not materialise Fingrid bought all shares in 1998 and from 1 June 1998 EL-EX became Nord Pool’s representation in Finland. Finland, thereby, effectively entered into the internal Nordic market with a day-ahead price quotation at Nord Pool. Denmark entered the Nordic market in two steps, thereby completing the picture.99 Denmark has two physically independent electricity systems, one being synchronised with UCTE and one being synchronised with the Nordel system. The two systems are not physically connected. To a great extent, liberalisation in Denmark was driven by the development in the other Nordic countries, where Danish utilities started to take part in the trade, and by the first EU market directive. In anticipation of the expected development, Danish 99. Iceland is also a Nordic country and a member of Nordel. Iceland has taken some steps to reform their electricity sector, but it is not interconnected with the other Nordic countries. What is referred to as the Nordic market in this book is excluding the Nordic country Iceland.

182

ANNEX 2

utilities started to adapt to the market development ahead of the legislation. The dominant utility in the western part, Elsam, re-organised with the establishment of a separate business entity for transmission ownership and system operation, Elsam System, from 1 January 1997. Elsam System was turned into a separate company, independent of Elsam, with effect from 1 January 1998 under the name Eltra. An amendment of the Danish electricity act in May 1996 transposed the EU market directive into Danish law by giving large consumers (more than 100 GWh per year) access to the grid from 1 January 1998. However, the act did not require unbundling and did not specify the framework for grid access. This only happened with the approval of the Electricity Supply Act on 2 June 1999 which established a framework for competition through regulated third-party access, legal unbundling of networks and the establishment of a regulator independent of government. In the western part of Denmark, market rules very similar to those found in the other Nordic countries had already been developed through 1998 and the spring of 1999. Thus, a western Danish market was launched on 1 July 1999 with a price quotation on Nord Pool, immediately after amendment of the electricity act. Eastern Denmark followed with the formation of an independent TSO, Elkraft System, on 1 January 2000 and Nord Pool price quotation on 1 October 2000. The Danish electricity sector was more or less fully owned by municipalities and co-operatives. From 1 January 2005, the Danish state bought the shares of the two TSOs and a joint, state-owned TSO was formed under the name Energinet.dk, also including system operation in gas. Danish electricity consumers have gained access to the grid in steps with consumers above 10 GWh/yr getting access from 1 April 2000. On 1 January 2001, the threshold was lowered to 1 GWh and it was removed altogether as of 1 January 2003. A final important milestone in the development of an integrated Nordic market is marked by the re-organisation of Nord Pool in 2002. By that time, Nord Pool conducted four key activities: the day-ahead spot exchange, financial trading, clearing and consultancy. Nord Pool was still owned by Statnett and Svenska Kraftnät. In early 2002, the day-ahead spot exchange was transformed into a separate company. From 1 July 2002, the other Nordic TSOs joined Nord Pool Spot as co-owners with 20% shares to each country and the last 20% still remaining in the hands of Nord Pool Holding. The Nordic electricity market developed constantly throughout the period. Market design has been refined and harmonisation for further integration has taken place in steps, driven by agreements at regular meetings between Nordic energy ministers. Nordic energy authorities have a permanent working 183

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

group to facilitate the necessary co-ordination between authorities and TSOs. Nordic TSOs co-operate through the association Nordel, which is the framework for permanent committees and numerous working groups that practically address a whole range of issues related to operation, system planning and market design.

Framework for Governance.................................................. In all Nordic countries, electricity market liberalisation is founded on a relatively detailed legislative framework. Legislation specifies the freedom of choice for consumers, regulated third-party access and legal unbundling of network activities. Finnish and Danish electricity reform acts also established the framework for the formation of regulators independent of government (regulators were already in place in Norway and Sweden). In Denmark, the Energy Agency, a government body, also has some of the roles that regulators traditionally have, including issuing licences for network companies. The roles and responsibilities of TSOs are specified in great detail in the legislations of each Nordic country. The common responsibilities are to ensure, maintain or enhance: operational security of the system; the momentary balance between demand and supply; adequacy of the transmission system in the long term; and efficient functioning of the electricity market. All TSOs co-operate closely with stake holders in the industry. Rules and principles that guide market design and information flow are conceived in various ways, with slight deviations from country to country. TSOs are involved, in one way or another, in most interactions that occur in the market and often play a central role in rule making and market design. Regulators monitor and often contribute to the process; other stakeholders are given a chance to react. Regulators oversee network tariffs and the performance of network companies, including TSOs. The Norwegian regulator, Water Resources and Energy Directorate (NVE), regulates network tariffs according to a cap system that allows revenues to increase with CPI, minus an annual requirement of efficiency improvement. NVE is a subordinated ministerial agency with independence in its day-to-day dealings. The Ministry of Petroleum and Energy is, however, the court of appeal for NVE rulings. The Swedish regulator, the Swedish Energy Agency, oversees the implementation of the electricity act, including ensuring that network tariffs are reasonable. It is governed by its board, which is appointed by the government. In Finland, the Energy Market Authority regulates networks by monitoring tariffs and the implementation of 184

ANNEX 2

Finnish legislation. It is the sole authority to oversee the implementation of the electricity act and it is independent from the Ministry of Trade and Industry. In Denmark, the Danish Energy Regulatory Authority oversees that network tariffs are in accordance with legislation and the directives established by the Danish Energy Agency. It is governed by a board, appointed by the relevant minister for a fixed term. Competition is regulated and monitored by competition authorities in all Nordic countries.

Basic Market Design Features .............................................. A first crucial building block in the design of the internal Nordic market is the postage-stamp grid tariffs. Across the entire Nordic system, there is a tariff for feeding electricity into the grid and a tariff for taking electricity off the grid. Initially, there were examples of specific border tariffs but the last element of that was removed when Sweden abolished its border tariff towards Denmark (March 2002). Tariffs are different from country to country – and even from region to region – but these differences mainly reflect real costs. There is no element in the network tariffs that unduly distorts trade within the Nordic region. The day-ahead spot market organised through Nord Pool is the cornerstone of the internal Nordic electricity market. Bids and offers must be submitted to Nord Pool by 12:00 noon for the following day. Bids and offers are put in merit order: the marginal bids and offers that determine the balance between supply and demand sets the price for the entire market. Unit commitments or other considerations regarding fixed costs are not taken into account in the market clearing but market players have various opportunities to submit blockbids. Block-bids enable generators to make a bid conditional of being dispatched for a block of hours instead of only one. Nordic TSOs give Nord Pool Spot a monopoly to use all available transmission capacity that interconnects the defined areas or zones in the Nordic market. Currently there are three zones in Norway, but they can change if Statnett sees frequent congestions in other places. Denmark has one zone for each of their two separate systems. Sweden and Finland constitute one zone each. Nord Pool calculates market prices for each zone, taking all bids and offers in each zone and interconnection capacity between the zones into account. The resulting zonal prices clear the market. All network companies are responsible for assessing and purchasing electricity resulting from grid losses. Hence, grid losses are reflected in the zonal prices through normal demand bids in the spot market. 185

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Nord Pool also calculates a system price assuming that there are no constraints in the entire Nordic transmission system. This is purely a reference used in the financial market and does not necessarily correspond with the exact prices faced by any market players. So far, market outcomes often show Norwegian zonal prices almost identical with the system price. Prices in the entire Nordic market are regularly identical and equal the system price. The western Danish zone is the most clearly congested area as it constitutes the gate between the hydro-based Nordic electricity system and the more traditional thermal electricity system in continental Europe. Initially, some interconnection capacity was reserved for old, long-term contracts. The last of these reservations was removed in 2000 when Stattkraft, Elsam and Preussen Electra (latter to merge into E.ON) abolished contracts that blocked the Danish-Norwegian interconnection. The parties were financially compensated by Eltra, the western Danish TSO. The transmission capacity made available to Nord Pool, as announced during the morning before day-ahead bids and offers are submitted, is guaranteed by the TSOs. This implies that the transmission right is firm. It is the responsibility of the TSOs to re-dispatch if the announced transmission capacities turns out to be unavailable at the moment of operation. The costs of such a circumstance are a part of transmission network tariffs and are thereby paid by all consumers and generators. Conversely, available transmission capacity is also a source to collect congestion rents, which will be used to lower tariffs if they are not used for financing new interconectors. Nord Pool operates a Web site on which prices, volumes and other fundamental market data are easily available. All market participants that wish to trade on Nord Pool must sign a user contract that outlines a wide range of obligations and responsibilities defining the interaction between market participants and the market. These obligations include a requirement to immediately disclose to Nord Pool any changes in generation and transmission facilities larger than 50 MW. Such changes are immediately posted on Nord Pool’s information service as a market message. Market messages include planned revisions and urgent messages on sudden changes. When Nord Pool has cleared the spot market and announced sales and purchases, all market players are required to submit schedules to the TSOs reflecting forecast demand and generation and all trades for each of the 24 hours of the following day. Trades can be both Nord Pool spot trades and bilateral trades. As Nord Pool has a monopoly on interconnection capacity, bilateral trades can only take place within a single zone. These schedules are binding in the sense that market players are financially responsible for their 186

ANNEX 2

fulfilment. All market players with a physical footprint in terms of generation, consumption or trade after the scheduling deadline are required to register as balance-responsible market players. That is, they must sign a contract with the TSO in the zone in which they want this physical footprint; through this contract they become financially responsible for deviations and are bound to follow the specified rules and formats for communication with that TSO. Balance responsibility is the connection between the electricity market and the physical system responsibility held by the TSO. Various initiatives and efforts have tried to harmonise and even unify balance responsibility contracts within the Nordic TSOs, but some differences still remain. After submitting schedules to TSOs, the framework deviates somewhat from country to country. In Sweden, Finland and eastern Denmark, it is possible to trade intra-daily up to a few hours before real-time operation. Nord Pool operates a trading platform, Elbas, for open intra-day auctions. So far, liquidity has been very low in this segment of the market. Balance-responsible market players in Sweden and Finland are also allowed to change their schedules until the final gate closure, when Elbas closes. There are also differences in real-time balancing. Nordic TSOs operate regulating markets in which they buy and sell electricity to balance the system according to the merit order of bids submitted by market players to TSOs. Prices for real-time regulation are determined by the marginal bid, as is the case in the day-ahead spot market. The Nordic regulating markets have become more and more integrated with enhanced information management systems. Now they function as a common Nordic market in which best bids and offers are called on across the entire Nordic region, subject to available transmission capacity. The purchase of regulating resources is more or less harmonised, but the settlement of imbalances deviates on some critical points. The total imbalance of a system is the aggregate imbalances of all balance-responsible market players. A majority of the individual imbalances will be in the same direction as the aggregate system imbalance, but some imbalances will be in the opposite direction and will, thereby, actually decrease the aggregate system imbalance. These individual imbalances that, by chance, are actually helping the system are treated differently in the different Nordic countries. In all the Nordic countries, the balance-responsible market players that contributed to the imbalance are charged the same price as the marginal price from the purchase of regulating resources. In Norway, the imbalances that help the system by chance are also rewarded with the same price and are, thereby, treated equally with the market players that were actively called on in the purchase of regulating resources. This pricing principle is referred to as the “single-price” system and is cost neutral. Market players that caused the 187

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

imbalance pay to those that alleviated the imbalance. The TSO is merely a coordinator or broker of the transactions. In Sweden, Finland and Denmark, the balance-responsible market players that helped the system by chance are not rewarded. Their imbalances are settled with the day-ahead spot price, which always gives an equal or poorer remuneration than the price settled in the purchase of regulating resources. Otherwise, there would be an incentive to make arbitrage between the two markets. Thus, balance-responsible market players that caused the imbalance pay the settled regulating price to the TSO and the TSO passes this price on to those that were actively called on. In reality, they are settling imbalances of those that helped the system by chance at a less favourable price. This pricing principle is referred to as the “dual-pricing” system and is not cost neutral. In fact, it generates a profit for the system operator. It is intended to provide extra incentive for making good forecasts and maintaining balance, but in reality it creates an incentive to merge and decreases liquidity in the regulating market. If market players can cancel individual imbalances through merger and aggregation, they save the costs that would otherwise end as revenue to the TSO. Imbalances are settled at the cleared regulating prices, usually one or two weeks after the day of operation. Local network companies collect hourly interval meter readings on a daily basis. These are matched with schedules to calculate individual imbalances. All Nordic countries have implemented systems for load profiling for the smallest consumers, primarily to avoid the need to have them install remotely read interval meters. The system with balance responsibility that allocates financial responsibility amongst market participants, the system of firm transmission rights made available to the market and the ex ante price settlement could all be thought of as systems based on forced day-ahead contracts. Binding contracts are made prior to real-time operation and balance-responsible market players are forced into the intra-day and/or real-time adjustment market. In an integrated market, it is a system of forced long-term contracts (longer term than real-time) that enables optimisation and planning across several jurisdictions. The day-ahead spot market is the cornerstone; real-time regulating and balancing markets provide the adjustment market between day-ahead spot and real-time operation. These markets create the interface between decisions made by market players and physical system operation, and could be called the markets with a physical footprint. However, most electricity trade and turnover takes place prior to the day of operation and is purely financial. Nord Pool spot has a market share of 43% of the physical Nordic demand; the remaining 57% 188

ANNEX 2

is traded bi-laterally. This could be thought of as bi-lateral physical trade but, in reality, it mainly reflects that several generators also have retail arms and therefore demand and generation are matched directly within the company. Nord Pool also operates a trading platform for financial trading and a clearing house for bi-lateral, brokered, over-the-counter (OTC) trade. In the short term, Nord Pool offers contracts for one to nine days ahead and for one to six weeks ahead in time. These futures contracts are settled daily. In the long term, they have contracts for one to six months ahead in time, one to eight quarters ahead, and one to three years ahead. These longer forward contracts are settled at the end of the period during which they go to delivery. All these futures and forward contract use the daily average system price as reference. There are also contracts to hedge zonal price differences, either one quarter or one year ahead for which several competing electricity brokers hold the largest market share. Brokered contracts in the OTC market are still based on the same standards as the contracts marketed by Nord Pool. European Option contracts, in which the option can be exercised at a pre-specified date, are also traded at Nord Pool and OTC. The main part of the liquidity in this standardised option market is traded OTC. Nord Pool offers clearing of OTC-traded standard contracts. Volume and price data from OTC trading are published on Nord Pool’s Web site, enhancing the flow of information in connection with its clearing service. This makes the Nordic market highly transparent, including the OTC market.

Market Structure .................................................................. In 2004, total consumption in the Nordic market was 391 TWh with 146 TWh in Sweden, 122 TWh in Norway, 87 TWh in Finland and 36 TWh in Denmark. Some 379 TWh was supplied from Nordic production with 148 TWh coming from Sweden, 111 TWh from Norway, 82 TWh from Finland, 38 TWh from Denmark and the remaining 12 TWh from neighbouring markets. The bulk of the net-imports came from Russia. Net-trade with Germany can deviate substantially from year to year, depending on market fundamentals. In 2004, there was a net-export of some 2 TWh out of the total 11 TWh traded with Germany. In the same year, some 2 TWh were net-imported from Poland. As of 31 December 2004, total installed capacity in the Nordic market was 91 076 MW. Some 47 059 MW was hydro capacity: 28 GW in Norway, 16 GW in Sweden and 3 GW in Finland. Hydro capacity in Norway is mainly in hydro reservoirs; in Sweden and Finland there is more run-of-river with less storing capability. The total maximum Nordic hydro storage capacity is some 120 TWh. Approximately 23 235 MW fossil-fuelled thermal capacity (mainly 189

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

coal) was installed: 8 GW in Denmark, 6.5 GW in Finland and 4 GW in Sweden (most thermal capacity in Denmark was combined heat and power). About 12 142 MW of nuclear power was installed: 9.5 GW in Sweden and 2.5 GW in Finland. The last approximately 8.2 GW of power came from other renewable sources including 4 GW of bio-mass in Finland and Sweden, and 3 GW wind power in Denmark. Another 1 058 MW was added to the Nordic system in 2004 and 574 MW was taken off-line. The installed capacity served a Nordic peak demand of 66 GW. Transmission interconnections between Nordic countries and regions, as well as between the Nordic market and neighbouring countries and regions, are a crucial part of the framework for a competitive Nordic market. Sweden and Norway are interconnected with 3 620 MW over nine different AC lines. Sweden and Finland are interconnected with 2 230 MW over five AC lines. Sweden and Denmark East are interconnected with 1 810 MW over four different sub-sea AC cables. Sweden and Denmark West are interconnected with 670 MW over two sub-sea DC cables. Norway and Denmark West are interconnected with a 1 000 MW sub-sea DC connection. Norway and Finland are interconnected with a single 100 MW AC line. The Nordic market is connected with neighbouring markets through 1 350 MW between Denmark West and Germany, 600 MW between Denmark East and Germany, 600 MW between Sweden and Germany, 600 MW between Sweden and Poland, and 1 560 MW between Finland and Russia. Several of the rated transmission capacities are subject to internal congestions and security requirements. As a consequence, some flows are reduced on a more or less permanent basis. The four largest generating companies in the Nordic market are Vattenfall, Fortum, Statkraft and E.ON Sweden (formerly Sydkraft). In 2001 Vattenfall had a market share of 19% in terms of output compared to the total Nordic electricity generation. Vattenfall is owned by the Swedish state and has expanded with significant stakes into Finland, Denmark and Germany. Fortum had a market share of 16% in 2001. It is based in Finland with the Finnish state having a 60% stake; it recently made significant acquisitions in Sweden. Statkraft had a market share of 12% in 2001. It is owned by the Norwegian state. E.ON Sweden had a market share of 8% in 2001 and is owned by the German utility E.ON. With this, the four largest generation companies had a joint market share corresponding to 55% of the total Nordic generation. There has been a significant turnover of generation assets in Sweden, Finland and Denmark but very limited in Norway. Norwegian hydro resources are subject to restrictive concessions by the Norwegian state and this part of the market is not liquid. No other companies held more than 4% of the market in 2001. In Norway, 160 companies are engaged in electricity generation; the 15 largest had 88% 190

ANNEX 2

of the market share of Norway’s total generation in 2001. In Sweden, the 15 largest generators had a market share of 94% of the total domestic generation in 2001. In Finland, the 15 largest companies had a market share of 95%. The Danish market is undergoing rapid change. The two largest generators, Elsam and Energi E2, are merging but Vattenfall is taking over roughly half of the generation assets. Apart from this, there are hundreds of small combined heat and power plants and some 4 000 wind turbines. Distributed generation and wind power are being integrated into the market in the sense that they do receive a fixed subsidy per kWh produced output. Apart from that they have the responsibility to dispose of their output in the market – a responsibility previously placed on Danish TSOs. In Norway, more than 100 retail companies are registered on the Competition Authorities’ Web site, to serve some 2.5 million Norwegian consumers. Eighteen of these companies are registered as being retailers across the entire country. In Sweden, the Swedish Consumer Agency invites retailers to publish price information on their Web site. Some 80 companies have registered that they supply to the approximately 5 million final customers across the entire country; some 50 registered companies supply only to local customers. The Swedish Consumer Agency notes that almost all retailers that supply the entire country are registered, as are an increasing number of local suppliers. In Finland, there are roughly 80 retail suppliers for 3 million consumers. In Denmark, where there are 3 million consumers, some 80 to 90 retail suppliers are registered on a price information service; most of them are only local suppliers. Even though there appears to be a very large pool of retail companies in the Nordic market, a large majority of them are actually the retail arm of local, often municipally owned, distribution companies that supply only local customers. At Nord Pool, some 400 participants are registered to trade in the day-ahead and/or financial market – either directly or through a handling agent. Roughly 150 of these are based in Norway, some 100 in Sweden and some 25 are based outside the Nordic region. The number of market participants has increased constantly since the launch of the market, up from some 150 participants in 1996.

Further Reading .................................................................... Background and development of the Nordic electricity supply industry is described in IEA country reviews. The joint reviews of the Nordic countries since late 1980s constitute a comprehensive description of the development 191

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

and of governing structures. These can be found on www.iea.org. Several descriptions of market design can be found on the Web sites of Nord Pool and the four Nordic TSOs; www.nordpool.com, www.statnett.no, www.svk.se, www.fingrid.fi, www.energinet.dk and www.nordel.org. These sites also contain comprehensive and detailed data and descriptions of the Nordic market’s fundamental structure. A joint report by the Nordic competition authorities analyses the structure of this market in great detail (2003). Numerous research articles describe the development of the Nordic market and various market design aspects, including Flatabø et al. (2003) and von der Fehr et al. (2005).

192

ANNEX 3

ANNEX 3: AUSTRALIAN NATIONAL ELECTRICITY MARKET Prior to the liberalisation process in Australia, the Australian electricity supply industry was dominated by state-owned utilities active in their own states. There was very little trade between states, except between Victoria and New South Wales (NSW) via the Snowy Mountain hydro system. Regulation of the electricity industry was under state jurisdiction. At the end of the 1980s, NSW, Victoria and Queensland had substantial overcapacity in generation, mainly due to optimistic projections of industrial demand growth made in the late 1970s and early 1980s. In the late 1980s, a range of initiatives were taken to address the various apparent inefficiencies in the Australian electricity industry. The Commonwealth Government Industry Assistance Commission was reviewing industry performance and pointed to a number of inefficient practices in an early stage. Some states were considering private construction and operation of generation plants and it was logical to consider building interstate transmission capacity to enable interstate trade. In March 1990, a new transmission line opened, connecting South Australia to the Victoria/NSW interconnected grid. Just two years later, as much as 24% of South Australia’s electricity demand was met by power from the other states. The liberalisation process in the Australian electricity market was effectively launched with the release of a draft report by the Commonwealth Industry Commission in January 1991. It recommended sweeping changes to the industry, with the main points being: placing transmission and distribution of gas and electricity on a commercial footing within one year; unbundling ownership of generation, transmission and distribution facilities within two years; breaking up electricity generating capacity into competing units; merging all state-owned transmission bodies into one organisation in order to establish a unified, national market; and requiring regulated third-party access to transmission and distribution networks. After privatisation, the final requirement was to bring tariffs in line with costs and remove cross-subsidies. A final report on the strategy was published in May 1991. In July 1991, the heads of Commonwealth, state and territorial governments, through the Council of Australian Governments (COAG), agreed to establish an intergovernmental supervisory body, the National Grid Management Council, to encourage and co-ordinate efficient development of the electricity industry in eastern and southern Australia. The Council’s mandate was to encourage open grid access, competition and free trade in electricity, and to 193

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

co-ordinate planning of generation and interconnected transmission systems. The Council was formed as a co-operative body with an independent chairman; members were appointed by participating jurisdictions. In the “One nation” statement in February 1992, Prime Minister John Major announced that he would seek states’ agreement to establish a National Grid Corporation to operate the transmission network – separately from existing generation and distribution interests. At the same time, the Commonwealth offered to contribute up to AUD 100 million (app. USD 80 million) to upgrade the transmission network, subject to states’ agreement to a timetable for developing the corporation. Additional financial incentives, totally AUD 4.2 billion up to 2005/06, were given by the federal governments. The Commonwealth followed up on the statement with a National Electricity Strategy discussion paper (May 1992) outlining the main elements of a policy to continue structural reform of the sector. In December 1992, the COAG agreed that the transmission sector should be managed separately from the generation sector. At the same time, the National Grid Management Council agreed to a protocol on the principles for development and operation of an interstate electricity grid. Electricity market reform was only one of several sector reforms within a wider effort to improve the competitiveness of the Australian economy. In October 1992, the prime minister commissioned an independent inquiry on national competition policy following an agreement with COAG. This report, released in August 1993, proposed the abolition of a large number of regulations in all economic sectors and led to the first element of federal legislation (1995) that was to have a direct impact on regulation of the electricity industry. A key element in the legislation was the establishment of the Australian Competition and Consumer Commission (ACCC), which was to be involved in the regulation of the electricity sector. In May 1994, COAG agreed to introduce a competitive national electricity market; by August 1994, they had agreed to the key principles and objectives of the market. These included unbundling, regulated third-party access, customer choice of supplier, regulatory reform of networks, merit order dispatch, non-discriminatory access for new entrants and no discriminatory barriers to inter- and intra-state trade. A key player in the National Electricity Market was the National Electricity Market Management Company (NEMMCO), the independent system and market operator. NEMMCO was founded in May 1996 with a membership comprising the governments of Queensland, New South Wales, Australian Capital Territory, Victoria, and South Australia. Further details had been worked out by October 1996, when the National Electricity Code (NEC) was submitted to ACCC for approval. The NEC specifies 194

ANNEX 3

market rules, network pricing, network connection and access, and system security. The National Electricity Code Administrator (NECA) was set up in May 1996 to enforce the code through implementation and monitoring of compliance. At the state level, the necessary restructuring took place in parallel with the development of the key principles of the market, albeit in varying speeds. In Victoria, the electricity sector was privatised beginning in 1994. Generation assets were privatised, plant by plant, into six generating companies; distribution was privatised as five independent companies. Initially, the transmission company, PowerNet Victoria, remained public but was also privatised later. The first Australian power market started to operate in Victoria in 1994, managed by the Victorian Power Exchange. In NSW, transmission was unbundled and a market launched in 1996. The same happened in Queensland in 1998. In South Australia, the state-owned vertically integrated utility was corporatised in 1995 and unbundled by accounting in 1997. In 2000, its assets were put under private management through a long-term lease. The first truly decisive and committing steps were taken in May 1996, when state energy ministers from NSW, Victoria, Queensland, South Australia and the Australian Capital Territory (ACT) signed an inter-governmental legislation agreement to support operation of the national electricity market in all participating jurisdictions. Ministers from NSW, Victoria and the ACT agreed to harmonise their existing electricity markets from 1 October 1996 to allow interstate competition at the earliest possible date, with expected market launch by the beginning of 1998. Queensland, South Australia and Tasmania were to join the market upon completing the necessary structural changes and connecting to the interstate grid. The National Electricity Market commenced operation on 13 December 1998, including South Australia, Victoria, the ACT, NSW and Queensland. After 2.5 years of operation (June 2001), it was decided to review market performance. This timing also coincided with a major reform of the England & Wales market, based on the obligatory pool system – which was somewhat similar to the key design principles of the Australian National Electricity Market. COAG agreed to establish the Ministerial Council of Energy (MCE), comprising energy ministers from the Commonwealth and all states and territories. MCE was to oversee the review process, also called the Parer review. The final report was published on 20 December 2002 under the name Towards a truly national and efficient energy market. Compared to development in England & Wales, where an overarching fundamental reform 195

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

had resulted, the Parer review confirmed the robustness of the NEM’s basic principles but pointed to various problems and gave recommendations on a wide range of necessary changes. One of the most important was the formation of two new bodies; the Australian Energy Market Commission (AEMC), which was to take over responsibility for rule making and market oversight from NECA; and the Australian Energy Regulator (AER), which would take over regulation from 13 state regulators. AEMC and AER were to have been established by July 2004 but the process was delayed to 2005. Consumer contestability on the retail level remains an issue under full state responsibility. NSW and Victoria gave full retail contestability to all consumers from January 2002; contestability was granted in South Australia from January 2003 and from the ACT in July 2003. Queensland gave retail contestability for consumers of more than 100 MWh/yr from July 2004 but decided not to lower this barrier to include other consumers. This decision is based on a cost-benefit analysis, performed in 2001, claiming that the benefits for consumers would not outweigh the costs of enabling them to switch. NEMMCO operates the system that manages all retail switching, including the management of load profiles accepted for smaller consumers. The NEM is now being extended to Tasmania through a new sub-sea DC cable connecting Tasmania with the interconnected national grid. Tasmania became a member of NEMMCO in May 2005, but will only be fully integrated in the market with the commissioning of the new cable, scheduled for early 2006. Retail contestability will develop in steps until the barrier is lowered to 150 MWh/yr in July 2009. At this point full contestability will be reviewed by the state.

Framework for Governance.................................................. The electricity supply industry is under the legal jurisdiction of states and territories, with the exception of issues related to interstate trade, which are in the jurisdiction of the Commonwealth as specified in the Australian constitution. The Constitution also specifies that all interstate trade must be free and fair. All necessary legislation that sets the framework for competitiveness and governance of the NEM in relation to interstate trade must be established in the legislations of states, territories and the Commonwealth. Close co-operation and co-ordination are an indispensable necessity for developing sufficient harmonisation to create a robust framework for governance. State and territory legislation required unbundling of transmission and distribution networks from generation and retail. It has assigned rule-making 196

ANNEX 3

and regulatory responsibilities to specific bodies and required third-party access. Transmission networks are unbundled by ownership. Detailed rulemaking has been left to those joint bodies formed out of co-operation within the COAG. In practice, the National Electricity Law has been passed by parliaments agreeing on one state to take the lead, after which mirror legislations were quickly passed in the other states and territories and in the Commonwealth. The new National Electricity Law appoints a new national body, the Australian Energy Market Commission (AEMC), to undertake rule-making and market development. The AEMC is responsible for the rules that determine market design, system operation and regulation through the new National Electricity Rules, which replaced the previous National Electricity Code. The AEMC reports directly to the Ministerial Council of Energy (MCE) and is governed by three commissioners. Its role is to make decisions on new proposed rules and to review the market, both on its own initiative and on that of the MCE. AEMC commenced operations from 1 July 2005, replacing the NECA. The National Electricity Law specifically requires the AEMC to establish a reliability panel that is to monitor, review and report on the security and reliability of the operation of the national electricity system. In parallel with the establishment of the AEMC, the new National Electricity Law also appoints a new national body, the Australian Energy Regulator (AER), to perform economic regulation of the wholesale markets and transmission networks in the electricity and gas markets. The AER also ensures the compliance of the National Electricity Rules. The intention is also to transfer regulation of distribution and retail to the AER. The AER is an independent legal entity but a constituent part of the ACCC. It is governed by three commissioners, including a chairman, of which one must be a commissioner in the ACCC; the other two commissioners are appointed by states and territories (1) and by the Commonwealth (1). The rulings and decisions of the AER are subject to judicial review by the Federal Court of Australia. In accordance with the National Electricity Rules, market and system operation are undertaken by the National Electricity Management Company Limited (NEMMCO), a company owned by states and territories that comprise the NEM. NEMMCO is governed by a board of directors, with each member appointing one individual. NEMMCO operates on a break-even basis by recovering the costs of operating the NEM from fees paid by market participants. 197

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

Basic Market Design Features .............................................. NEMMCO operates a spot market for wholesale electricity trade. The spot market is compulsory for all generation and consumption connected to the interconnected transmission grid that forms the back bone of the NEM. Generators must submit offers and retailers can submit bids to NEMMCO for generation and consumption. These offers and bids must be submitted before 12.30 p.m., one day ahead of operation, specifying volumes and prices for each of the 48 half-hour periods in the following day. The volume in the bids can be changed up until five minutes before real-time operation, but the price cannot change. A pre-dispatch forecast of prices, volumes and dispatch is calculated for the following trading day. However, final prices and volumes are set only when the actual dispatch to meet demand at least cost has been determined – i.e. after the actual operation. This ex-post market settlement has a wide range of implications for market operation, which are quite different from many other market systems. One implication is that nobody is forced to make binding commitments about generation, demand or transmission capacity prior to real-time. Without these forced commitments, there is no need for retailers to submit bids. As a default, NEMMCO makes load forecasts, which are used as input to make pre-dispatch forecasts. Retail companies can submit bids based on their own forecasts and particularly their own knowledge about demand response to prices. However, this approach makes little difference and has never really been used. In a system of ex post price settlement, it is not necessary to make the market twosided. It does not, on the other hand, offer any implicit possibility to hedge activity prior to operation. All hedging must be performed through voluntary contracts between different market players. This has the advantage of minimising regulatory involvement and clarifying the interface between the responsibilities of NEMMCO and market participants. But it also has drawbacks in that transmission rights are not made firm, which makes it difficult to hedge for price differences and it makes the true independence and impartiality of transmission owners critical. It also creates challenges for demand participation: much demand will be far more responsive to price, if it is possible to plan the response one day ahead of operation. The NEM does not offer firmness for such planning. This should be developed in the normal market for hedging through aggregation and financial contracts. Bids submitted by generators are stacked in ascending order. The generators with the lowest bids are dispatched until demand is met for every five-minute period. For each of these five-minute intervals, a marginal price is determined by the most 198

ANNEX 3

expensive generator dispatched. A spot price is calculated as the average of the six marginal prices in the five-minute intervals over each half hour of the day. This average marginal price is then used to settle the market. Consumers and retail companies are charged this price for all consumption through this half hour; generators are paid this price for all dispatched generation through this half hour. Unit commitments and other factors related to the fixed costs of plants are not incorporated into the market price calculation. Bids of a certain volume and at a certain price are dispatched to the extent that they are necessary to meet load. On the other hand, as market participants are allowed to adjust the volumes of the bids until five minutes before operation, market participants are left some room to manage such constraints. Transmission capacity is made available to NEMMCO by transmission system owners. Prices are calculated on a zonal basis, in which each member state is one zone and the Snowy Mountain region on the border between Victoria and NSW comprises an additional individual zone (due to the special role of its hydro capacity). Grid losses are assessed through computed loss factors at different locations in the grid, relative to specific reference points. Loss factors are calculated annually, based on historical data. Generators must compensate for grid losses and this is taken into account in the calculation of the dispatch schedule. In that sense, generators face a locational signal on two levels: one on the level of the zone, when taking congestion of interconnectors into account; and one on the level of distance to a zonal reference point, when taking grid losses into account. NEMMCO must operate the system to balance supply and demand while at the same time maintaining a minimum reserve margin. The minimum reserve margin is currently set on levels equivalent to the capacity of the largest generators in each zone, but at the same time accounting for the possibility to share reserves across zones. The minimum reserve margin is required to enable the system to withstand the sudden failure of any generator. NEMMCO does not acquire operational reserves on a regular basis; however, NEM rules give NEMMCO the possibility to tender for reserves if they fall below minimum levels. In the beginning of 2004, the Statement of Opportunities predicted a reserve deficit of 356 MW below minimum during the summer months of February and March 2005, in South Australia and Victoria. NEMMCO initiated a tendering process to acquire operational reserves, and consequently 80-90 MW was contracted mainly from demand response. NEMMCO purchased ancillary services in eight parallel markets for frequency control, network control and black start. Competing service providers can bid their services into these eight markets. Financing of ancillary services is based on “causer pays” principles. 199

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

NEM rules also regulate a market for financial transmission rights. Price differences between zones give NEMMCO a congestion rent, also called settlement residues. Price differences create risks for market players and NEMMCO operates a financial market that facilitates hedging these risks. Rights to settlement residues are auctioned on a quarterly basis. Buying a certain share gives the right to the corresponding share of the settlement residues, but the right is not firm. If interconnection capacity is reduced or removed, the rights are lost with no compensation. Settlement residues and auction revenues are used to lower transmission use-of-service charges. Transparency and information management are critical points in the NEM market design. NEMMCO releases an annual Statement of Opportunities, presenting and analysing all fundamental data on demand, transmission and generation. NEMMCO operates a Web site with very extensive dissemination of information. Prices, volumes, changes to the system and other fundamental market data are easily available. A unique feature is that all bids and offers submitted to NEMMCO are publicly available one day after operation. Most trade in the mandatory spot market is backed by financial contracts (between market participants) to manage price and volume risks. Trade with financial contracts, apart from the settlement residue auction, is not regulated in the NEM rules. Financial trade takes place mainly in the bi-lateral OTC market, facilitated by brokers. Electricity derivatives are also traded on the Sydney Futures Exchange. One segment of the financial market is instituted and regulated by the NSW government as a forced hedging contract. On 1 January 2001, the NSW government instituted the Electricity Tariff Equalisation Fund (ETEF), which requires electricity retailers in the state to pay money into a fund when the NSW pool price is below a certain, regulated threshold. If the pool price is above the threshold, retailers receive money instead. In that sense, the ETEF works as a protection that allows retailers to earn a guaranteed margin. But it also undermines competition and incentives for innovation. If the fund develops a negative balance, state-owned generators must contribute funds to keep the fund solvent. Generators are repaid when pool price developments recover the balance through payments from retailers.

Market Structure .................................................................. In 2003, electricity supplied to the 8 million consumers in the NEM was 158 479 GWh, of which 65 TWh (41%) was generated in NSW and the ACT, 200

ANNEX 3

40 TWh (25%) in Victoria, 42 TWh (27%) in Queensland and 12 TWh (7%) in South Australia. The introduction of Tasmania into the NEM will increase load with some 10 TWh. Peak demand in the summer 2003/04 was 12.2 GW in NSW and the ACT, 8.6 GW in Victoria, 7.9 GW in Queensland and 2.6 GW in South Australia. Total NEM peak load was 29.8 GW, including the 1.3 GW in Tasmania, which is 2.9 GW less than the sum of the regional peaks. In 2003, demand in the NEM was met by generation from coal (91%), natural gas (6%) and hydro (3%). In the same year, NSW imported 8 TWh, Victoria was a net exporter of 4 TWh, Queensland exported 3 TWh, South Australia imported 2 TWh and the Snowy Mountain Region exported 4 TWh of the 5 TWh generated in that region. By the end of 2003, total installed principal generation capacity in the NEM was 38 533 MW of which 12 GW (32%) was in NSW, 8 GW (22%) in Victoria, 11 GW (28%) in Queensland, 3 GW (9%) in South Australia and 4 GW (10%) in the Snowy Mountain Region. Extending the NEM to Tasmania will increase installed capacity with 2.5 GW of mainly hydro capacity. Interconnections between various regions are the backbone of the NEM’s integrated national market. NSW and Queensland are interconnected with 950 MW transmission capacity, of which only some 600 MW are made available for imports to Queensland. NSW and the Snowy Mountains are interconnected with 3 038 MW. Victoria and the Snowy Mountains are interconnected with 1 892 MW. Victoria and South Australia with 627 MW, of which only 379 MW are made available for imports to Victoria. Victoria and NSW are interconnected via the Snowy Mountains, but imports to the Snowy Mountains are limited to some 1 100 MW from either side. Generation assets in the NEM are owned both by private and state-owned companies. The three largest generation companies in terms of installed capacity are Macquarie Generation, Delta Electricity and Eraring Energy. All are based in NSW and owned by the NSW government. They dominate the NSW region with shares between 40% and 25% of the total installed capacity. In the NEM, they have between 12% and 8% of the market share and a joint market share of 31%. In total, there are 27 generation companies in the NEM, one-third of these are state-owned. Apart from NSW, Queensland also owns the dominating generation companies in that state. All generation assets in Victoria and South Australia are operated by privately owned companies, several of which are non-Australian. On the retail side, there are 66 retailing companies supplying the 3.2 million consumers in NSW and the ACT, the 2.3 million consumers in Victoria, the 1.7 million consumers in Queensland and the 0.8 million consumers in South Australia. 201

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

By mid-2005, NEMMCO had registered 103 market participants; six of these were engaged only in trading.

Further Reading .................................................................... Background and development of the electricity supply industry in Australia is described in IEA country reviews. The joint reviews of Australia since the late 1980s constitute a comprehensive description of the development and of governing structures. The latest country review from 2005 gives a comprehensive description of the structure of the Australian electricity industry. These can be found on www.iea.org. The Web site of NEMMCO, the system and market operator, also contains a wealth of information, data and descriptions on most aspects of the NEM: www.nemmco.com.au.

202

ANNEX 4

ANNEX 4: PENNSYLVANIA – NEW JERSEY – MARYLAND INTERCONNECTION (PJM) The Pennsylvania – New Jersey – Maryland Interconnection (PJM) has a much longer history than that of the development of electricity market liberalisation itself. PJM has been a pool that enables co-ordination of trade between the three founding utilities since 1927. Prior to 1978, the United States electricity industry was run by vertically integrated utilities, in most cases privately owned. These companies were regulated by the state public utilities commissions (PUCs). The Federal Power Act of 1935 kept most regulatory authority in the hands of the states. The Public Utility Holding Company Act of 1935 (PUHCA) limited the possibility for any utility to do business in other states, thereby also effectively limiting the scope for mergers and acquisitions. On the federal level, the Federal Energy Regulatory Commission (FERC) has authority only over wholesale trade issues. An important development away from this approach was triggered by the Public Utility Regulatory Policies Act of 1978 (PURPA), which gave small generation resources – particularly with environmental merits – access to the grid through contracts corresponding to avoided costs. This led to an influx of independent power producers (IPPs), primarily in those states where the vertically integrated utilities were permitted and encouraged to auction leastcost contracts to IPPs to obtain the power needed. The next important step towards a more competitive market was taken on a federal level with the Energy Policy Act of 1992. This act gave the FERC authority to order open access for wholesale trade between utilities and across state borders. It specifically prohibited the FERC from ordering open access for final consumers. This was still a matter for state legislation. In anticipation of developments the Energy Policy Act of 1992 would trigger, PJM started to transform itself into an independent, neutral organisation in 1993, primarily through the formation of PJM Interconnection Association, which administered the power pool. By 1995, the FERC had developed a draft set of rules that would implement the Energy Policy Act of 1992. In 1996, the final set of rules was issued in Order 888. The key points in the FERC Order 888 include: functional unbundling of utilities’ transmission system operation business from their power marketing business; transmission utilities under FERC jurisdiction must file non-discriminatory, open-access transmission tariffs that offer service to third parties on a comparable basis to the utilities’ own 203

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

uses of their transmission facilities; and guidelines for independent system operators (ISOs) that would be subject to FERC jurisdiction. In 1999, the FERC issued order 2000, which asks for consolidation across jurisdictional borders, meaning that utilities are encouraged to merge ISOs into regional transmission organisations (RTO) that act as co-ordinating system operators across larger areas. In September 2001, the chairman of the FERC made several proposals to encourage the standardisation of market design and push for the formation of RTOs. The proposals were clarified in a standard market design (July 2002) and the FERC issued a white paper (April 2003) with a refined version of the proposals. After strong resistance from some states, the proposals have never materialised as an instrument to force wholesale competition in all states. In the Energy Act of 2005 (July 2005), the FERC was given some of the authority initially suggested in its standard market design. With this act, the FERC has much stronger authority in matters of system security and it has been given some authority in the approval process of new transmission infrastructure and to monitor and enforce competitive behaviour in wholesale markets. In short, the Energy Act of 2005 indicates that the development towards competitive and open electricity markets should be supported, but not enforced on all states. The Energy Act of 2005 also abolishes the limitations set by the PUHCA, thereby opening up for consolidation across states. With the legal framework of Order 888 and the intentions expressed by the FERC in mind, PJM became a fully independent organisation in 1997. Ownership was opened to non-incumbent utilities and an independent Board of Managers was elected. The original PJM Interconnection comprised the control areas of eight utilities, covering large proportions of Pennsylvania, most of New Jersey, large portions of Maryland and Delaware in its entirety. A bid-based spot market for power commenced operation on 1 April 1997, one year before the launch of the Californian market. Later in 1997, PJM was approved by the FERC as the first ISO in the country to be in compliance with Order 888. PJM was thereby responsible for safe and reliable operation of the unified transmission system and for the management of a competitive wholesale electricity market across the control areas of its members. With the FERC’s encouragement of the formation of RTOs in mind, PJM, the New York ISO (NYISO) and the ISO New England (ISO-NE) tried to agree to merge to form a unified RTO. However, an agreement could not be struck and in December 2002, PJM earned full RTP status on its own. The period between the issuance of Order 2000 and the approval of PJM as an RTO also covers a period with substantial turbulence as more and more states embarked on the 204

ANNEX 4

electricity market liberalisation process. California was one of the first states to order full retail access in March 1998: a competitive wholesale market was launched on 1 April 1998 and performed very well initially. It reinforced the merits of liberalisation but the subsequent poor performance from 2000 and the eventual crash in summer 2001 had an even stronger negative influence on the liberalisation process in the US. The first years of PJM operation were used to firmly establish and develop the market to become a robust framework for competition. The initial day-ahead spot market was based on a single market-clearing price for the entire region. High costs for congestion management and poor operational flexibility in the utilisation of the system (due to security restrictions) called for a stronger locational reflection of real costs. One year after the initial launch, the level of sophistication increased by introducing locational marginal pricing (LMP) based on reported costs, in which market-clearing prices were calculated for each node in the system. On 1 January 1999, a daily capacity market was introduced, motivated by the traditional approach to capacity planning in the region and by a concern for stranded costs from incumbents, who would now be faced with competition from generators without former times’ capacity requirements. From March 1999, the daily capacity auctions were extended to monthly and multi-monthly auctions. On 1 April 1999, the cost-based LMP was replaced with LMP based on real competitive bidding. On 1 June 2000, the day-ahead market was extended with a real-time market, also based on LMP and competitive bidding. On December 1 2000, a market for spinning reserves was added. With the implementation of LMP principles in 1999, there appeared a need to offer hedging of the consequential price differences between nodes. In April 1999, PJM introduced an auction of allocated financial transmission rights (FTRs), which gave market participants the needed risk-hedging opportunity. In May 2003, the sophistication of this market increased again by replacing the initial allocation of FTRs with an entirely financial allocation of auction revenue rights (ARRs). By the end of 2001, the market design had developed to create a framework for a robust market. The market design has continued to develop since then, particularly through modifications of the capacity market. Since 2002, there has been a focus on extending PJM system and market operation area. On 1 April 2002, Allegheny Power joined PJM to form PJM West. The extension added more regions of Pennsylvania, large parts of West Virginia, parts of Virginia and small parts of Ohio. In June 2002, American Electric Power (AEP), Commonwealth Edison (ComEd), Illinois Power and National Grid signed a memorandum of understanding with PJM to develop an independent transmission company that would operate within PJM West. Also in June 205

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

2002, Dominion joined PJM to form PJM South, thereby integrating a large share of the electricity system in Virginia and a small share in North Carolina into PJM’s system and market operation. AEP (Ohio), Com Ed (Illinois), the Dayton Power & Light Company (Ohio), Duquesne Light Company (Pennsylvania) and Dominion were successfully integrated into the system and market operation of PJM during 2004 and first half of 2005. The integration of Com Ed alone expanded the PJM market by 20%. Midwest ISO (MISO) and PJM have worked together since 2004, with the goal of developing an integrated wholesale market across the 23 states, the District of Columbia and Manitoba (Canada), which covers their joint control areas. MISO launched a LMP-based market on 1 April 2005. Most of the states in PJM have ordered retail access for all consumers. The first state in the United States to order full retail access was Rhode Island (January 1998), a part of ISO-NE. The first state in PJM to follow suit was New Jersey (August 1999). Pennsylvania, D.C., Delaware, Ohio, Maryland and Illinois gave full retail access at various points between 2000 and 2004. By 2004, 18 states had given retail access to all consumers.

Framework for Governance.................................................. Three major federal acts on energy determine the level of involvement of federal authority in the regulation of the electricity industry in the United States. The Federal Power Act of 1935 limits the federal authority to issues related to wholesale trade. The Energy Policy Act of 1992 ordered open access for wholesale transactions, meaning that there must be open access to trade between utilities within and across states. Finally, with the Energy Policy Act of 2005, the intentions laid out in the Energy Policy Act of 1992 were confirmed and federal authority was strengthened in some critical areas such as system security, transmission system extensions and market monitoring. Some state legislation, including in most of the states within PJM, gives full retail access to final consumers. State regulators, together with a federal regulator, oversee compliance of state and federal legislation. States have public utility commissions (PUCs) and the Federal Energy Regulatory Commission (FERC), which is an independent agency within the Department of Energy, regulates on those areas in which federal legislation gives it authority. PUCs regulate intra-state utility business, such as generation and distribution. The FERC regulates interstate energy transactions, including wholesale power transactions on transmission lines. The FERC consists of five members appointed by the 206

ANNEX 4

President and confirmed by the Senate. No more than three commissioners may belong to the same political party. The President appoints its chairman. In Order 888, the FERC established the set of detailed rules that must be fulfilled to comply with the Energy Policy Act of 1992. In the 1960s and 1970s, large blackouts raised the issue of secure and reliable operation of interconnected transmission systems that reach across state and jurisdictional borders. In 1968, the North American Electric Reliability Council (NERC) was formed by the 10 regional reliability councils that had been formed in preceding years. The NERC is a voluntary organisation with the mission of ensuring that the North American electricity transmission grid is reliable, adequate and secure. The lessons from the widespread northeastern blackout in 2003 pointed to the inadequacy of voluntary reliability rules, particularly concerning lack of enforcement of otherwise appropriate rules. In the Energy Policy Act of 2005, the FERC was given authority to appoint a body to act as regulator of a set of mandatory reliability rules. PJM Interconnection is a limited liability, non-profit company, governed by a board of managers. Members of the board of managers must have no personal affiliation or ongoing professional relationship with – or any financial stake in – any PJM market participant. Users of PJM join as members and are represented with a vote in the members committee. The members committee elects a board of managers and provides advice to this board by proposing and voting on changes in market rules; it also has authority to make specific recommendations. There are other committees and user groups for resolution of issues through discussion and negotiation. Market rules and market design issues are often developed through these governing structures. There is a specific unit within PJM to oversee the functioning of the market; the Market Monitoring Unit (MMU). The MMU is an independent group that assesses the state of competition in each of PJM markets, identifies specific market issues and recommends potential enhancements to improve competitiveness and market efficiency. In particular, the MMU is responsible for monitoring the compliance of members with PJM market rules and for evaluating PJM policies to ensure those rules remain consistent with the operation of a competitive market. The MMU issues an annual report on the state of the market.

Basic Market Design Features .............................................. It costs resources to transport electricity, both in terms of keeping adequate transmission assets available and in terms of energy losses due to resistance in transmission grids. In PJM, this fact has been taken very explicitly into 207

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

account by introducing a nodal pricing system, which means that a price is calculated in every possible choke point in the electricity system. Every bus that connects two transmission and/or distribution lines is a possible choke point and is, thus, incorporated directly into the calculation of prices. All generators defined as a capacity resource in PJM system are obliged to submit an offer into the day-ahead PJM market. The bus that connects a generator to the grid is specified when registering. Offers can include incremental prices, specifying different prices at different generation volumes. They can also specify minimum run times and start-up costs to ensure that unit commitments are incorporated into the market clearing price. Market participants are allowed to self-schedule. On the generator side this is accounted for by indicating that a specific share of a generation unit must run regardless of the price. An offer specifying that a unit must run is basically just a schedule that commits the generator financially. Retailers and consumers must submit bids to the day-ahead spot market. They can do it by bidding prices and volumes, if they intend to respond to the price by decreasing demand, or they can do it without specifying any price. All demand bids and generation offers are tied to a specific bus and transmission capacity between all buses are specified. Load and/or generation at a specific bus will have a specific influence on energy losses from transmission. These losses will be dynamic but sensitivity factors are calculated annually as approximations of the transmission losses they incur relative to a reference point in the grid. With all bids, offers and sensitivity factors, a dispatch schedule is calculated that minimises total system costs of serving total demand. Reliability and transmission system security considerations are taken into account in the total market clearing. A marginal pricing principle is used, whereby the most expensive offer that meets the cheapest bid sets the price. This nodal pricing is the instrument of congestion management internally within PJM market area. Interconnections between the PJM area and other neighbouring areas are also defined as nodes. External bids and offers across these nodes determine the price in the node and the flow across the interconnector. Other details are also included in the market clearing calculation. Certain generators are included in the price calculation, according to a previously agreed “cost of service” due to their position and power in the market. Other generators might be forced to operate due to reliability and stability criteria in the operation of the system. Depending on their location, these generators are not always included in the calculation of locational marginal nodal prices, according to their offers. 208

ANNEX 4

Market clearing prices are calculated at all buses; if there are congestions between nodes, nodal prices may be different. These price differences reflect the cost of electricity transmission. Each generator is paid the market clearing price in its specific node. All loads are charged the market clearing price in their specific nodes. For market participants that have chosen to self-schedule, either by letting own generation meet own load or by entering into other bilateral contracts, the resulting price differences are exactly the charges that reflect the costs connected to a specific location. In 2004, some 26% of the load was cleared in PJM day-ahead market. The remainder was generation offers submitted as must run, meaning it was self-scheduled. Price differences between nodes can imply a significant market risk for market participants. At the same time, these price differences generate revenue streams to PJM known as financial transmission rights (FTRs). Owners of transmission grids and holders of contracts linking generation in one location with load at another can request the allocation of FTRs. Investors in new transmission interconnectors earn the right to FTRs tied to that line. FTRs can be traded, either bi-laterally or through PJM, which conducts annual and monthly auctions. A holder of an FTR for the flow of a specific volume in a specific direction between two nodes will receive the revenues of price differences between those two nodes – or will be charged if the revenue stream is negative. This creates a hedge for a contract that was made at a price referring to one of those nodes. Instead of tying the initial allocation of FTRs to specific congestion points, according to the earned rights, an initial annual allocation of auction revenue rights (ARRs) has been instituted. ARRs give the rights to the proceeds of an annual auction of FTRs. If the holder of the ARR has a specific interest in the underlying FTR, the ARR can be transformed into the corresponding FTR outside of the auction. With ARRs, the initial auction of FTRs has become more liquid and, hence, more flexible. PJM spot market is a two-tier market with a day-ahead market and a real-time market. The pricing principles in the real-time market follow the same principles of nodal-based, locational marginal pricing as the day-ahead market. The real-time market is an adjustment market, in which deviations between scheduled bids and offers submitted the day-ahead are traded. Market participants are free to self-schedule to compensate for any deviations. PJM real-time market is a single-price market. Deviations that contribute to the total system imbalance are faced with the costs of regulation to achieve system balance. Deviations that, by chance, help the total system balance are faced with the same price as those resources that were actively called on to regulate. There is no extra punishment for imbalances in excess of the real costs of managing the imbalance. 209

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

PJM operates a market for installed capacity. All load-serving entities (LSEs: retail companies with contracts to serve consumers) are obliged to contract for an amount of installed capacity that corresponds to the peak load of the customer base plus reserves. The obligation must be met by capacity which is registered by PJM and available to the system. Owners of such capacity receive a capacity credit. LSEs can either contract for capacity credits bi-laterally or through the market for capacity credits organised by PJM. If LSEs are not meeting their obligation, they are charged with a fine, currently at a level of some USD 170 /MW and per day of unfulfilled requirement. PJM conducts daily, interval, monthly and multi-monthly auctions of capacity credits. PJM also acquires ancillary services through market-based mechanisms. There is a market for capacity to meet regulation on a very short-term, automatic regulation. There is also a market for spinning reserve to make sure there is adequate capacity on-line to meet sudden failures in the system. Both markets are cleared every hour. Markets operated by PJM include: the day-ahead spot market, which also constitutes a reference; the real-time spot market; a market for capacity credits; and two markets for ancillary services. A substantial part of electricity trade in the PJM market area takes place outside of the official PJM market places. Much load is met through self-scheduled generation, either within a vertically integrated company (generation and retail) or through bi-lateral contracts. Financial trading is evolving at the New York Mercantile Exchange (NYMEX) and the Intercontinental Exchange (ICE) where liquidity in terms of traded volumes is increasing rapidly, particularly over the past year (2004/05). There are more than 7 000 nodes for calculation of nodal marginal prices in the PJM model. Most of these nodes are never singled out with a specific price because they are not congested. PJM calculates weighted average prices for three hubs that consist of important non-congested nodes. These hubs are used as references for risk management and financial trading. The most important hub is the western hub, which attracts the most liquidity as a reference point. PJM Web site offers extensive information on market prices, the state of the electricity system, principles, governance and other valuable information to understand the overall structure of the market and its day-to-day evolution. Data showing individual bids, without identifying the bidder, are published with a six-month delay. There is no requirement or corresponding site that contains information regarding the status of specific generation units. A generation unit can be taken off-line – with possible significant consequences for involved nodal prices – without immediate pass through of that information to the market place. 210

ANNEX 4

Market Structure .................................................................. Statistical data on demand and supply in the PJM market has changed substantially during the last years with the rapid extension of the market area. In 2004, there where three important phases: Phase 1 with 12 control zones; Phase 2 where PJM also comprised the ComEd control zone, and; Phase 3 where PJM also included American Electric Power and Dayton Power & Light. In 2005, PJM was further extended to also include Duquesne Light Company and Dominion. In 2004, demand peaked at 78 GW in the extended PJM market. Assessed peak demand after the extensions in 2005 is 130 GW; total demand is some 700 TWh. The joint population in the area, after the 2005 extensions, is 51 million spread across 13 states plus the District of Columbia. During Phase 1, PJM was a net importer: during the first four months of 2004, some 7.2 TWh were imported to PJM. In the extension of PJM in Phases 2 and 3, PJM became a net exporter with a net export of some 5.5 TWh during five months in Phase 2 and some 3.9 TWh of net export in the last three months of the year. This made PJM a net exporter of some 2.2 TWh for the calendar year 2004. The net-export and total load was served by 144 GW of installed production capacity as of 31 December 2004: 41.5% was coal-fired; 28.4% was fuelled by natural gas; 19.1% was nuclear; 7% was oil; 3.7% was hydroelectric; and 0.3% was solid waste. In the calendar year 2004, the installed coal units generated 52.1% of total generation; 36.9% was from the nuclear stations; units fuelled by natural gas generated 7%; oil was 1.1%; hydroelectric generated 2.3%; and solid waste generated 0.6%. With the extensions in 2005, the total installed generation capacity increased to 164 GW. With the inclusion of ComEd, a special feature was introduced in PJM market, which is particularly important when analysing the market structure. ComEd is connected to the other PJM area via a pathway through a geographic area that is not a part of PJM market. By the end of 2003, American Electric Power Company was the largest generation company in PJM, owning 17% of the total installed capacity and generating 22% of the total output that year. Exelon was second with 13% of installed capacity and 23% of the total generation. Third was Public Service Enterprise Group (PSEG) with 9% of installed capacity and 6% of generated electricity. Exelon and PSEG have proposed to merge. In the marketscreening process of the consequences of the merger, PJM Market Monitoring Unit has analysed the resulting market shares. After a merger, it is indicated 211

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

that the market share in terms of installed capacity will result in a merged company with a 29% market share, up from 22%, as well as another company with 22% and at least three companies with market shares of 6% each. It is noted that a forced divestiture of 4 500 MW would reduce market concentration to pre-merger levels. The market for retail supply is overseen by state public utility commissions. In states with customer choice, there are lists of licensed suppliers from which consumers can choose. In Pennsylvania, there are 40 licensed suppliers of which roughly half are registered as suppliers to all types of customers; the remaining serve only industrial and commercial customers. A similar picture can be found in other PJM states such as New Jersey and Maryland. PJM has more than 375 members with generation capacity and/or retail supply commitments, or with only trading commitments.

Further Reading .................................................................... The fact that electricity markets in the United States develop geographically on a control area basis, with the extension of markets proceeding control area by control area subject to state regulation, makes it complicated to get a full overview of the development. Hunt (2002) provides a thorough description and discussion of the development in the United States, including PJM. PJM Web site (www.pjm.com) provides substantial amounts of material describing the various aspects of the market model. The Market Monitoring Unit publishes an annual report on the state of the market, as does the FERC, presenting thorough analysis of all markets in the United States. Many research articles have been published on the Californian crisis, the various market design aspects of PJM (including LMP and capacity markets) and on market power analysis.

212

REFERENCES

REFERENCES Alberta Department of Energy, 2005, “Alberta’s Electricity Policy Framework: Competitive – Reliable – Sustainable”, Review of Alberta’s electricity market, www.energy.gov.ab.ca Australian Government, 2004, “Securing Australia’s Energy Future”, www.pmc.gov.au/initiatives/energy.cfm, Commonwealth of Australia, Canberra Boot, Pieter, 2005, “Security of Electricity Supply – the Dutch Approach”, presentation at IEA/NEA workshop Security of Energy Supply for Power Generation - May 2005, www.iea.org Bushnell, James B., & Catherine Wolfram, 2005, “Ownership Change, Incentives and Plant Efficiency: The Divestiture of US Electric Generation Plants”, CSEM Working Paper 140, University of California Energy Institute Centre for Advancement of Energy Markets, 2003, “Estimating the Benefits From Restructuring Electricity Markets – An Application to PJM Region”, www.caem.org Council of Australian Governments Energy Market Review, 2002, “Towards a truly national and efficient energy market”, Commonwealth of Australia CPB, 2004, “Energy Policies and Risks on Energy Markets – A cost-benefit analysis”, CPB Netherlands Bureau for Economic Policy Analysis, The Hague Cramton, Peter & Steven Stoft, 2005, “A Capacity Market that Makes Sense”, conference paper, EPRI conference on Global Electricity Industry Restructuring – in Search of Alternative Pathways, May 2005, http://www.lrca.com/ events/EPRI-AC/ Danish Competition Authority, 2003, “Elsam og Energi E2 afgiver tilsagn så konkurrencen fremmes” (”Elsam and Energi E2 gives consent that supports the improvement of competition”), press release, www.ks.dk ECON Energy, 2003a, “2002: Høsten da tilsiget sviktet”, www.econenergy.com ECON Energy, 2003b, ”Torrår i Norden – Vad hände på kraftmarknaden”, www.econenergy.com ENEL, 2005, ”ENEL Annual Report 2004”, ENEL financial statement, www.enel.it ESAA, 2002,”Electricity Australia 2002”, Electricity Supply Association of Australia Limited, Melbourne 213

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

ESAA, 2004, ”Electricity Australia 2002”, Electricity Supply Association of Australia Limited, Melbourne ETSO, 2004, “An Overview of Current Cross-border Congestion Management Methods in Europe”, ETSO report on congestion management, www.etsonet.org ETSO & EuroPEX, 2004, “Flow based market coupling”, joint ETSO and EuroPEX proposal, www.etso-net.org European Commission, 2000, ”Commission allows merger of VEBA and VIAG subject to stringent conditions”, Press release IP/00/613 on 13 June 2000, www.europe.eu.int, Brussels European Commission, 2004a, “Trans-European Energy Networks: TEN-E Priority Projects”, European Commission DG for Energy and Transport, www.europe.eu.int, Brussels European Commission, 2004b, “EU Productivity and Competitiveness: An industry Perspective”, European Commission DG Enterprise, www.europe.eu.int, Brussels European Commission, 2005, “Report from the Commission on the Implementation of the Gas and Electricity Internal Market: Technical Annexes”, Commission Staff Working Document, European Commission DG for Energy and Transport, Brussels. EFET-European Federation of Energy Traders, 2004, “Reforming the Management of Electricity Transmission Congestion in the EU Internal Market: an EFET Vision”, EFET position paper, www.efet.org European Parliament, 2005b, European Parliament legislative resolution on the proposal for a Directive of the European Parliament and of the Council Concerning Measures to Safeguard Security of Electricity Supply and Infrastructure Investment”, www.europe.eu.int, Brussels European Union, 2003a, “Directive 2003/54/EC of the European Parliament and of the Council of 26 June 2003 Concerning Common Rules for the Internal Market in Electricity and repealing Directive 96/92/EC”, Official Journal of the European Union, L 176/37-55, Brussels European Union, 2003b, “Regulation (EC) No 1228/2003 of the European Parliament and of the Council of 26 June 2003 on conditions for access to the network for cross-border exchanges in electricity”, Official Journal of the European Union, L 176/1, Brussels 214

REFERENCES

Evans, Joanne & Richard Green, 2005, “Why did Brittish Electricity Prices fall After 1998”, Department of Economics Working Paper, WP 05-13, University of Birmingham Federal Energy Regulatory Commission (FERC), 1996, “Order No 888, Promoting Wholesale Competition through Open Access”, www.ferc.gov von der Fehr, Nils-Henrik, Eirik S. Amundsen and Lars Bergman, 2005, “The Nordic Market: Signs of Stress”, The Energy Journal, special issue on European Electricity Liberalisation FERC, 2005, “State of Markets Report 2004”, Annual FERC energy market report, www.ferc.org Flatabø, Nils, Gerard Doorman, Ove S. Grande, Hans Randen and Ivar Wangensteen, 2003, “Experience With the Nord Pool Design and Implementation”, IEEE Transactions on Power Systems, vol. 18, no. 2 FSE-Foreningen for Slutbrugere af Energy, 2003, ”Klage til EU Kommissionen over svensk stop for dansk import af elektrisk kraft fra Sverige” (”Complaint to the EU Commission over Swedish blocking of Danish import of electric power from Sweden”), letter from FSE to the European Commission, www.fse.dk Green, Richard, 2004, “Electricity Transmission Pricing: How Much Does it Cost to Get it Wrong?”, Cambridge Working Papers in Economics, CWPE 0466, The Cambridge-MIT Institute Holmen, 2005, ”Holmen signs electricity supply and energy efficiency agreements with Vattenfall”, press release, www.holmen.com Hunt, 2002a, “Making Competition Work in Electricity”, John Wiley & Sons, New York IEA, 2002a, “Distributed Generation in Liberalised Electricity Markets”, IEA/OECD, Paris IEA, 2002b, “Security of Supply in Electricity Markets”, IEA/OECD, Paris IEA, 2003a, ”Power Generation Investment in Electricity Markets”, IEA/OECD, Paris IEA, 2003b, “The Power to Choose – Demand Response in Liberalised Electricity Markets”, IEA/OECD, Paris IEA, 2004a, “Security of Gas Supply in Open Markets – LNG and Power at a Turning Point”, IEA/OECD, Paris IEA, 2004b, “World Energy Outlook 2004”, IEA/OECD, Paris 215

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

IEA/NEA, 2005, ”Projected costs of generating electricity – 2005 Update”, OECD/IEA Paris IEA, 2005a, “Russian Electricity Reform: Emerging Challenges and Opportunities”, IEA/OECD, Paris IEA, 2005b, “Energy Policies of Spain”, IEA/OECD, Paris IEA, 2005c, “Learning from the Blackouts: Transmission system security in competitive electricity markets”, IEA/OECD Paris IEA, 2005d, “Act Locally, Trade Globally: Emissions trading for Climate Policy”, IEA/OECD, Paris Joskow, Paul L., 2003, ”The Difficult Transition to Competitive Electricity Markets in the US”, Electricity Restructuring: Choices & Challenges, eds. J. Griffin and S. Puller, University of Chicago Press. Joskow, Paul L., 2004, “Integrating Transmission Networks and Competitive Electricity Markets”, presentation at IEA workshop Transmission Network Performance in Competitive Electricity Markets, www.iea.org Joskow, P. & Jean Tirole, 2004, “Reliability and Competitive Electricity Markets”, Massachusetts Institute of Technology Department of Economics Working Paper Series, Working paper 04-17 Littlechild, S., 2004, “Regulated and Merchant Interconnectors in Australia: SNI and Murraylink Revisited”, Cambridge Working Papers in Economics, CWPE 0410, The Cambridge-MIT Institute Ministry of Petroleum and Energy, 2003, “Om forsyningssikkerheten for strøm mv.”, Stortingsmelding nr. 18, www.oed.no NEMMCO, 2004, ”Statement of Opportunities for the National Electricity Market 2004” NEMMCO, 2005, “Statement of Opportunities for the National Electricity Market – Update 2004” Newbery, D. & M. Pollit, 1997, “The Restructuring and Privatisation of Britain’s CEGB – Was it worth it?”, The Journal of Industrial Economics, vol. 45, No 3, pp. 269-303, Blackwell Publishing Ltd Newbery, D., R. Green, K. Neuhoff, P. Twomey, 2004, “A Review of the Monitoring of Market Power”, Report prepared at the request of ETSO, www.etso-net.org 216

REFERENCES

Newbery, David, 2005, “Electricity Liberalisation in Britain”, The Energy Journal, special issue on European Electricity Liberalisation Nordel, 2003a, ”Action plan: Peak Production Capability and Peak Load in the Nordic Electricity Market”, Nordic Council of Ministers and Nordel Nordel, 2003b, “Power and energy balances, today and three years ahead”, www.nordel.org Nordel, 2004a, “Peak Production Capability and Peak Load in the Nordic Electricity Market”, Summary report in the Nordel project on Peak Load Capability and Demand in the Nordic Electricity Market, www.nordel.org Nordel, 2004b, “Demand Response in the Nordic Countries – A background survey”, A working document in the Nordel project on Peak Load Capability and Demand in the Nordic Electricity Market, www.nordel.org Nordel, 2004c, “Annual report 2003”, www.nordel.org Nordel, 2005a, “Enhancing Efficient Funtioning of the Nordic Electricity Market”, www.nordel.org Nordel, 2005b, “Nordel to Finance Jointly Network Investments and to Streamline Market Rules”, Nordel press release 28 February 2005, www.nordel.org Nordic Competition Authorities, 2003, “A Powerful Competition Policy: Towards a More Coherent Competition Policy in the Nordic Market for Electric Power”, www.ks.dk, Copenhagen, Oslo and Stockholm Nord Pool, 2005, “Negative Prices in the Elspot market”, Nord Pool message, www.nordpool.com NVE (Norwegian Water Resource and Energy Directorate), 2005, “Leverandørskifteundersøkelsen 4. kvartal 2004” (Report on retail switching 4th quarter 2004), www.nve.no OECD, 2005, “The Benefits of Liberalising Product Markets and Reducing Barriers to International Trade and Investment: the Case of the United States and the European Union”, OECD Economics Department Working Paper 432, Paris OFGEM, 2004, “2003/2004 Electricity Distribution Quality of Service Report”, OFGEM, London Oren, Shmuel, 2005, “What is a natural capacity mechanism that will meet generation adequacy need s with minimal interference in energy markets?”, conference paper, EPRI conference on Global Electricity Industry Restructuring 217

LESSONS FROM LIBERALISED ELECTRICITY MARKETS

– in Search of Alternative Pathways, May 2005, http://www.lrca.com/ events/EPRI-AC/ Pfeifenberger, Johannes, Joseph Wharton & Adam Schumacher, 2004, “Keeping up with Retail Access? Developments in U.S. Restructuring and Resource Procurement for Regulated Retail Service”, The Electricity Journal, vol. 10, pp. 50-64 PJM Interconnection, 2005a, “PM Annual Report 2004”, www.pjm.com PJM Interconnection, 2005b, “2004 State of the Market”, www.pjm.com PVO, 2005, “Annual Report 2004”, http://www.pohjolanvoima.fi Rassenti, S. J., V. L. Smith & B. J. Wilson, (2001), “Turning off the Lights”, Regulation, v. 24, no. 3, Fall 2001, The Cato Institute South Australia Government Electricity Taskforce, 2001, http://www.treasury. sa.gov.au Speckler, Jörg, 2005, “The Feasibility of Independent Power Production in Germany”, conference paper, PowerGen Europe 2005 – Market Liberalisation: Lessons, realities & opportunities, Penn Well Sydvest Energi, 2005, “Årsrapport 2004” (Annual report 2004), www.sydvestenergi.dk Svenska Kraftnät, 2002, “Effektförsörjningen på den öppna elmarknaden” (”The capacity supply in the open electricity market”), www.svk.se de Tomás, José Antonio, 2005, “Security of Energy Supply for Specific Technologies”, presentation at IEA/NEA workshop Security of Energy Supply for Power Generation - May 2005, www.iea.org TVO, 2005, “Annual Report 2004”, http://www.tvo.fi Vilnes, Olav, 2005, “Chasing the Insiders”, Montel Powernews, No 3, Oslo.

218

The Online Bookshop International Energy Agency All IEA publications can be bought online on the IEA Web site: w w w. i e a . o r g / b o o k s You can also obtain PDFs of all IEA books at 20% discount. Books published before January 2004 - with the exception of the statistics publications can be downloaded in PDF, free of charge, on the IEA website.

IEA BOOKS Tel: +33 (0)1 40 57 66 90 Fax: +33 (0)1 40 57 67 75 E-mail: [email protected]

CUSTOMERS IN NORTH AMERICA

International Energy Agency 9, rue de la Fédération 75739 Paris Cedex 15, France

You can also send

CUSTOMERS IN THE REST OF THE WORLD

your order Turpin Distribution The Bleachery 143 West Street, New Milford Connecticut 06776, USA Toll free: +1 (800) 456 6323 Fax: +1 (860) 350 0039 [email protected] www.turpin-distribution.com

to your nearest OECD sales point or through the OECD online services: www.oecdbookshop.org

Turpin Distribution Services Ltd Stratton Business Park, Pegasus Drive, Biggleswade, Bedfordshire SG18 8QB, UK Tel.: +44 (0) 1767 604960 Fax: +44 (0) 1767 604640 [email protected] www.turpin-distribution.com

IEA PUBLICATIONS 9, rue de la Fédération, 75739 Paris Cedex Pre-press by Linéale Production Printed in France by CHIRAT Code 612005321p1 ISBN 92 64 10959 5 December 2005