TrimSize:6inx9in - download.e-bookshelf.de
Transcript of TrimSize:6inx9in - download.e-bookshelf.de
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page i�
� �
�
A NEW SWING-CONTRACTDESIGN FOR WHOLESALEPOWER MARKETS
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page ii�
� �
�
IEEE Press445 Hoes Lane
Piscataway, NJ 08854
IEEE Press Editorial BoardEkram Hossain, Editor in Chief
Jón Atli Benediktsson David Alan Grier Elya B. JoffeXiaoou Li Peter Lian Andreas MolischSaeid Nahavandi Jeffrey Reed Diomidis SpinellisSarah Spurgeon Ahmet Murat Tekalp
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page iii�
� �
�
A NEW SWING-CONTRACTDESIGN FOR WHOLESALEPOWER MARKETS
Leigh Tesfatsion
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page iv�
� �
�
Copyright © 2021 by The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey.
Published simultaneously in Canada.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form orby any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except aspermitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the priorwritten permission of the Publisher, or authorization through payment of the appropriate per-copy fee tothe Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax(978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission shouldbe addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission.
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts inpreparing this book, they make no representations or warranties with respect to the accuracy orcompleteness of the contents of this book and specifically disclaim any implied warranties ofmerchantability or fitness for a particular purpose. No warranty may be created or extended by salesrepresentatives or written sales materials. The advice and strategies contained herein may not be suitablefor your situation. You should consult with a professional where appropriate. Neither the publisher norauthor shall be liable for any loss of profit or any other commercial damages, including but not limited tospecial, incidental, consequential, or other damages.
For general information on our other products and services or for technical support, please contact ourCustomer Care Department within the United States at (800) 762-2974, outside the United States at(317) 572-3993 or fax (317) 572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print maynot be available in electronic formats. For more information about Wiley products, visit our web site atwww.wiley.com.
Library of Congress Cataloging-in-Publication Data
Names: Tesfatsion, Leigh, author.Title: A new swing-contract design for wholesale power markets / Leigh
Tesfatsion.Description: Hoboken, New Jersey : Wiley-IEEE Press, [2021]. | Series: IEEE
press series on power engineering | Includes index.Identifiers: LCCN 2020027003 (print) | LCCN 2020027004 (ebook) | ISBN
9781119670124 (hardback) | ISBN 9781119670131 (adobe pdf) | ISBN9781119670148 (epub)
Subjects: LCSH: Electric utilities. | Electric utilities–Costs. | Options(Finance)
Classification: LCC HD9685.A2 T466 2021 (print) | LCC HD9685.A2 (ebook) |DDC 333.793/23–dc23
LC record available at https://lccn.loc.gov/2020027003LC ebook record available at https://lccn.loc.gov/2020027004
Cover design by WileyCover images: (Top image) © Sam Robinson/Gettyimages, (Center image) © dowell/Getty Images
Set in 10/12pt TimesLTStd by SPi Global, Chennai, India
10 9 8 7 6 5 4 3 2 1
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page v�
� �
�
This book is dedicated to Herman C. Quirmbach for his steadfastsupport of this challenging book project and for his insistence,by word and example over many years, that reasoning be both
precise and clear.
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:48 P.M. Page vi�
� �
�
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 12/12/2020 5:35 P.M. Page vii�
� �
�
CONTENTS
PREFACE xiii
AUTHOR BIOGRAPHY xiv
ACKNOWLEDGMENTS xv
CHAPTER 1 INTRODUCTION 1
CHAPTER 2 US RTO/ISO-MANAGED WHOLESALE POWER MARKETS: OVERVIEW 9
2.1 Chapter Preview 9
2.2 General Goals for Wholesale Power Market Design 9
2.3 US RTO/ISO-Managed Market Operations 10
2.4 Stresses Faced by Current US RTO/ISO-Managed Markets 14
CHAPTER 3 MOTIVATION FOR CURRENT STUDY 17
3.1 Chapter Preview 17
3.2 Problematic Design Aspects of US RTO/ISO-Managed Wholesale Power Markets 17
3.2.1 Artificial Distinction Between Energy and Reserve 17
3.2.2 Problematic use of Hedonic Pricing 18
3.2.3 Revenue Insufficiency and Incentive Problems 19
3.2.4 Computational Fragility of LMP Derivations 20
3.2.5 Performance Payment in Advance of Performance Delivery 22
3.2.6 Minimal Direct Representation of Retail Customer Interests 23
3.2.7 Reliance on Overly Simplistic Cost Conceptions 24
3.2.8 Use of Spot-Market Pricing for Forward Markets 26
3.3 Relation of Current Study to Previous Swing-Contract Work 26
CHAPTER 4 SWING CONTRACTS FOR ISO-MANAGED WHOLESALE POWERMARKETS 29
4.1 Swing Contract Overview 29
4.2 Swing Contracts: General Formulation 29
4.3 Swing Contracts in Firm or Option Form 31
CHAPTER 5 ILLUSTRATIVE SWING-CONTRACT RESERVE OFFERS 35
5.1 Chapter Preview 35
5.2 A Simple Energy-Block Swing Contract in Firm Form 37
vii
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:49 P.M. Page viii�
� �
�
viii CONTENTS
5.3 An Energy-Block Swing Contract in Option Form 40
5.4 Swing-Contract Implementation of Standard Supply Offers 41
5.5 A Swing Contract Offering Continuous Swing (Flexibility) in Power and Ramp 47
5.6 A Swing Contract Offering Battery Services 49
5.7 Swing-Contract Facilitation of Private Bilateral Contracting 52
CHAPTER 6 SWING-CONTRACT MARKET DESIGN 55
6.1 Chapter Preview 55
6.2 General Swing-Contract Market Formulation 55
6.3 Financial and Physical Feasibility of Swing-Contract Offers 58
6.4 Reserve Bids 58
6.5 Handling of Fixed Reserve Bids and Non-Dispatched Power 60
6.6 Performance Penalties and Incentives 60
6.7 ISO Cost Allocation 61
CHAPTER 7 SWING-CONTRACT MARKET OPTIMIZATION: BASE-CASE MILPFORMULATION 67
7.1 Chapter Preview 67
7.2 General Assumptions and Notation 68
7.3 Discretization of the ISO’s Optimization Problem 69
7.4 ISO Objective Function 73
7.5 Complete Analytical MILP Formulation 74
7.6 Additional Discussion of Optimization Aspects 76
7.7 Five-Bus Test Case 78
7.8 Thirty Bus Test Case with Adaptive Reserve Zones 81
CHAPTER 8 INCLUSION OF RESERVE OFFERS WITH PRICE SWING 85
8.1 Chapter Preview 85
8.2 Cost Function Preliminaries 86
8.3 MILP Tractable form of Reserve Offers with Price Swing 87
CHAPTER 9 INCLUSION OF PRICE-SENSITIVE RESERVE BIDS 93
9.1 Chapter Preview 93
9.2 Incorporation of Benefits 94
9.3 Modeling of Price-Sensitive Reserve Bids 96
9.3.1 Standard Demand Function Formulation 96
9.3.2 Reserve Bids with Time-of-Use Pricing 97
9.3.3 Reserve Bids with Price Swing 97
9.3.4 Reserve Bids Directly Expressed as Benefit Functions 99
9.4 MILP Tractable Approximation of Benefit Functions 100
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:49 P.M. Page ix�
� �
�
CONTENTS ix
CHAPTER 10 THE LINKED SWING-CONTRACT MARKET DESIGN 105
10.1 Chapter Preview 105
10.2 Multistage Optimization and Time Inconsistency 107
10.3 Settlement Time-Consistency of Swing-Contract Markets 109
10.4 Swing-Contract Long-Term Forward Markets 111
10.5 Swing-Contract Short-Term Forward Markets 112
10.6 Swing-Contract Very Short-Term Forward Markets 113
10.7 Swing-Contract Deployment in Real-Time Operations 114
CHAPTER 11 ILLUSTRATION: LINKED DAY-AHEAD AND HOUR-AHEADSWING-CONTRACT MARKETS 117
11.1 Chapter Preview 117
11.2 Hour-Ahead Market with Reserve Offers Consisting of Swing-Contract Portfolios 117
11.3 SCED Solution for Hour-Ahead Swing-Contract Market 122
11.3.1 Overview 122
11.3.2 Power Balance 122
11.3.3 Coverage of the ISO’s Uncertainty Set 123
11.3.4 Constrained Minimization of Expected Cost 125
11.4 Linked Day-Ahead and Hour-Ahead Markets 126
CHAPTER 12 STANDARD MODELING OF A COMPETITIVE MARKET 131
12.1 Chapter Preview 131
12.2 Key Definitions 131
12.3 Standard Competitive Market Assumptions 132
12.4 Law of One Price for Commodities 132
12.5 Competitive Market: Basic Formulation 133
12.6 Net Surplus Extraction 136
12.7 Market Efficiency Metric 137
12.8 Market Efficiency and Pricing Rules 139
12.9 Strategic Trade Behavior and Trader Market Power 140
CHAPTER 13 US RTO/ISO-MANAGED MARKETS: EFFICIENCY AND MARKETPOWER 143
13.1 Chapter Preview 143
13.2 Daily Market Operations 144
13.3 Illustrative Analytical DAM Formulation 146
13.4 Net Surplus Extraction in the Illustrative DAM 147
13.5 Market Power in the Illustrative DAM: Type-I Error 152
13.6 Market Power in the Illustrative DAM: Type-II Error 156
13.7 Market Inefficiency in the Illustrative DAM 160
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:49 P.M. Page x�
� �
�
x CONTENTS
13.8 DAM Performance: General Assessment 163
13.9 Scheduling of Bilateral Contracts 165
CHAPTER 14 COMPARISONS WITH SWING-CONTRACT MARKETS 167
14.1 Chapter Preview 167
14.2 Product Definition in US RTO/ISO-Managed Markets 168
14.3 Wholesale Power and the Law of One Price (Not) 170
14.4 Differential vs. Uniform Pricing 171
14.5 Comparison of SC and Current US DAM Designs 172
CHAPTER 15 ADVANTAGES OF THE LINKED SWING-CONTRACT MARKET DESIGN 175
15.1 Chapter Preview 175
15.2 SC Markets are Physically-Covered Insurance Markets 176
15.3 Longer-Term SC Markets Support New Investment 177
15.3.1 Energy-Only Market 179
15.3.2 Centrally Managed Capacity Market 181
15.3.3 LSE Bilateral Contract Obligations 182
15.4 SC Markets Ensure Revenue Sufficiency 183
15.5 SC Markets Ameliorate Merit-Order Concerns 184
15.6 SC Markets are Robust-Control Mechanisms 185
15.7 SC Markets Reduce Rule Complexity 186
15.8 SC Markets Reduce Gaming Opportunities 187
15.9 SC Markets have Smaller-Sized Optimizations 189
15.10 Additional Advantages of SC Markets 190
15.10.1 Ensure a Level Playing Field for Resource Participation 190
15.10.2 Permit Co-Optimization of Diverse Reserve 191
15.10.3 Appropriately Remunerate Diversity and Flexibility 191
15.10.4 Encourage Accurate Forecasting and Dispatch Following 191
15.10.5 Ensure Settlement Time-Consistency 191
CHAPTER 16 GRADUAL TRANSITION TO LINKED SWING-CONTRACT MARKETS 193
16.1 Chapter Preview 193
16.2 A DAM Formulation Permitting Gradual Transition 195
16.3 Cost Function Preliminaries for the Transitional DAM 197
16.4 MILP SCUC/SCED Optimization for the Transitional DAM 201
CHAPTER 17 SWING-CONTRACT SUPPORT FOR INTEGRATED TRANSMISSIONAND DISTRIBUTION SYSTEMS 209
17.1 Chapter Preview 209
17.2 Transactive Energy System Design for ITD Systems 211
17.3 Role of Distribution Utilities 215
17.4 An IDSO-Managed Bid-Based TES Design for Households 216
17.5 IDSOs as Grid-Edge Resource Aggregators 219
17.6 Swing-Contract Support for IDSO Participation in Wholesale Power Markets 220
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:49 P.M. Page xi�
� �
�
CONTENTS xi
CHAPTER 18 DESIGN EVALUATION VIA THE ITD TES PLATFORM 221
18.1 Chapter Preview 221
18.2 Design Readiness Levels 222
18.3 An ITD TES Platform Permitting TES Design Evaluation 223
18.4 Illustrative Test Cases: Overview 226
18.5 Illustrative Test Cases: Report 229
18.5.1 IDSO Peak-Load Reduction Capabilities 229
18.5.2 IDSO Load-Matching Capabilities 229
18.5.3 Household ITD Test Cases: Discussion 233
CHAPTER 19 POTENTIAL FUTURE RESEARCH DIRECTIONS 235
19.1 Effective use of Option Swing Contracts 235
19.2 Representation of Reserve Bids 236
19.3 Compensation for Storage Services 236
19.4 Compensation for Reliability Services 236
19.5 Representation of Power-Paths 237
19.6 Implementation of Contract-Clearing Optimizations for Swing-Contract Markets 237
19.7 Gradual Transition to a Swing-Contract Market 238
CHAPTER 20 CONCLUSION: THE DOTS KEEP CONNECTING 239
APPENDIX A APPENDICES 241
REFERENCES 249
INDEX 259
Trim Size: 6in x 9in Tesfatsion670124 f01.tex V2 - 11/20/2020 10:49 P.M. Page xii�
� �
�
Trim Size: 6in x 9in Tesfatsion670124 f03.tex V2 - 11/19/2020 11:42 P.M. Page xiii�
� �
�
PREFACE
THE NEED FOR FLEXIBLE DEPENDABLE RESERVE provision in electricpower systems has dramatically increased in recent years. Growing reliance onvariable energy resources and greater encouragement of demand-side participationhave led to greater uncertainty and volatility of net load. Consequently, systemoperators are finding it harder to secure reserve with sufficient flexibility to permit thecontinual balancing of net load, a basic requirement for power system reliability. Thisstudy reconsiders the design of wholesale power markets in light of these concerns.Four design principles are stressed: (i) Wholesale power markets must necessarilybe forward markets due to the speed of real-time operations; (ii) Only one type ofproduct can effectively be offered in a wholesale power market: namely, reserve,an insurance product offering availability of net load balancing services for futurereal-time operations; (iii) Net load balancing services offered into wholesale powermarkets primarily take the form of power paths that can be dispatched at specific gridlocations over time; (iv) All dispatchable resources should be permitted to competefor the provision of power-paths in wholesale power markets without regard forirrelevant underlying technological differences. If these four principles are accepted,current trade and settlement arrangements for wholesale power markets need tobe fundamentally altered. This study proposes a new linked swing-contract marketdesign, consistent with principles (i)–(iv), that could meet the needs of centrallymanaged wholesale power markets better than currently implemented designs.
xiii
Trim Size: 6in x 9in Tesfatsion670124 f03.tex V2 - 12/12/2020 9:42 P.M. Page xiv�
� �
�
AUTHOR BIOGRAPHY
LEIGH TESFATSION, PhD, is a Research Professor, and Professor Emeritaof Economics, Mathematics, and Electrical and Computer Engineering, at Iowa StateUniversity. She is a Senior Member of the IEEE whose work has been supported bygrants and contracts from the US Department of Energy, Pacific Northwest NationalLaboratory, Sandia National Laboratories, Los Alamos National Laboratory, andNSF. She has served as guest editor and associate editor for a number of journals,including the IEEE Transactions on Power Systems, the IEEE Transactions onEvolutionary Computation, the Journal of Energy Markets, and the Journal ofEconomic Dynamics and Control.
xiv
Trim Size: 6in x 9in Tesfatsion670124 flast.tex V2 - 11/19/2020 11:42 P.M. Page xv�
� �
�
ACKNOWLEDGMENTS
THIS STUDY HAS BEEN SUPPORTED by US Department of Energy(DOE) grants DE-AR0000214 and DE-OE0000839, by Contract No. 339051 withthe Pacific Northwest National Laboratory (PNNL) operated by Battelle for theUS DOE under Contract DE-AC05-76RL01830, by Contract No. 1163155 withSandia National Laboratories, by Project Award No. M-40 from the Power SystemsEngineering Research Center (PSERC), and by various grants from the Iowa StateUniversity Electric Power Research Center (ISU EPRC).
I am especially grateful to Joe Dygert, Joe Eto, Marija Ilic, and Lorenzo Kris-tov for early support and encouragement of the swing-contract ideas expressed inthis study, and to Swathi Battula, Raymon H. Bryne, Ryan T. Elliott, James F. Elli-son, Ross Guttromson, Deung-Yong Heo, Wanning Li, Verne Loose, Shanshan Ma,César A. Silva-Monroy, and Zhaoyu Wang who have collaborated with me on pre-vious swing-contract research. I am indebted to Ross Baldick, Ramón Alberto LeónCandela, Kwok Cheung, Paul Gribig, Bill Hogan, Akshay Korad, Dick O’Neill, JimPrice, Fereidoon Sioshansi, Di Wu, Nanpeng Yu, and Tongxin Zheng for thoughtfulcomments on earlier draft versions. I thank Yonghong Chen, Hailong Hui, DanielKirschen, and Eugene Litvinov for patiently responding over the years to my manyquestions about current US RTO/ISO-managed wholesale power market operations.Finally, I thank five anonymous referees for helpful constructive comments on pre-liminary drafts of this monograph.
xv
Trim Size: 6in x 9in Tesfatsion670124 flast.tex V2 - 11/19/2020 11:42 P.M. Page xvi�
� �
�
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 1�
� �
�
CHAPTER 1INTRODUCTION
“Design to the mission, design as a system, keep it simple.”
—[109, p. 20]
CENTRALLY MANAGED WHOLESALE power markets operating overhigh-voltage transmission grids support the steady flow of electric power from bulkpower sellers to bulk power buyers, for ultimate resale and distribution to retailcustomers. This mission has been complicated in recent years by a dramatic surge inthe availability and use of variable energy resources (VERs).
A VER is a power source whose power injections into a transmission gridcannot be fully dispatched in a controlled manner to balance changes in power with-drawals or to meet other system requirements. Examples include solar panels andwind turbines that are not fully firmed by storage. The increased participation ofVERs in wholesale power markets, together with the increased encouragement ofactive demand-side participation, increases the uncertainty and volatility of grid netload, i.e. power withdrawal net of non-dispatched power injection.
In consequence, as discussed more fully in Chapters 2–3, US RTO/ISO-managed wholesale power markets1 are finding it harder to secure dependablereserve with sufficient flexibility to permit the continual balancing of net load, abasic requirement for power system reliability. Trade and settlement arrangementsin these markets are still largely based on rigid reserve definitions, eligibilityrequirements, and settlement processes that make it difficult to ensure adequateprovision and appropriate compensation of needed reserve from multiple types ofresources. Emphasis is placed on the designation and compensation of artificiallyseparated product concepts such as energy, ramping, and capacity, whereas value
1The US Federal Energy Regulatory Commission [58] defines an RTO/ISO-managed wholesale powermarket to be the collection of all capacity, energy, and/or ancillary service markets operated by a regionaltransmission organization (RTO) or an independent system operator (ISO). The key distinction betweenan RTO and an ISO is that RTOs have larger regional scope.
A New Swing-Contract Design for Wholesale Power Markets, First Edition, Leigh Tesfatsion.© 2021 The Institute of Electrical and Electronics Engineers, Inc. Published 2021 by John Wiley & Sons, Inc.
1
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 2�
� �
�
2 CHAPTER 1 INTRODUCTION
in power markets in fact principally arises from the dispatchable availability anddelivery of power-paths, i.e. flows of power into and out of a grid at specific gridlocations during designated operating periods.
This study reconsiders the design of US RTO/ISO-managed wholesale powermarkets in light of these concerns. Four market design principles are stressed:
[MD1:] All wholesale power markets must necessarily be forward markets2 due tothe speed of real-time operations.
[MD2:] Only one type of product can effectively be offered in a wholesale powermarket: namely, reserve, an insurance product offering availability of netload balancing services for future real-time operations.
[MD3:] Net load balancing services offered into wholesale power markets gener-ally take the form of power-paths that can be dispatched at specific gridlocations over time.3
[MD4:] All dispatchable resources should be permitted to compete for the provi-sion of power-paths in wholesale power markets without regard for irrele-vant underlying technological differences.
A swing-contract market design is proposed that is in accordance withprinciples MD1–MD4. This design envisions an ISO-managed wholesale powermarket M(T) organized as a reserve market for some designated future operatingperiod T. Reserve consists of dispatchable power-paths for period T. As illustrated inFigure 1.1, a power-path for period T refers to a sequence of power injections and/orwithdrawals at a single designated grid location during period T.4 Dispatchableresources offer reserve (dispatchable power-paths) into M(T) by means of “swingcontracts.”
More precisely, as carefully explained in Chapter 4, a swing contract SCmissued by a dispatchable resource m is a reserve contract that m can offer into aswing-contract market M(T) in either firm or option form.5 SCm consists of fourcomponents, each specified by m: (i) an offer price 𝛼m; (ii) an exercise set 𝕋 ex
m ; (iii) a
2A forward market is a market involving the purchase and sale of a product for which the payment methodfor the product is contractually determined in advance of its delivery date. In contrast, in a spot market thedelivery and payment for a product are determined at the same time.3As discussed in [46, 59], primary frequency response is synchronized reserve capacity that autonomouslyresponds to changes in system frequency; consequently, it is not dispatched. The provision and compen-sation of primary frequency response is not considered in the current study.4Since a power-path refers to the injection and/or withdrawal of power at a single grid location overtime, a power-path is characterized without reference to spatial transmission. As illustrated in Figure 1.1,power-paths can be depicted in a time-power plane.5As explained more fully in Chapter 4, a firm contract is a noncontingent contract that imposes obligationson both the issuer and the holder. An option contract is a contingent contract that gives the holder the right,but not the obligation, to exercise the contract at one or more contractually specified exercise times. Theexercise of an option contract converts it into a firm contract.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 3�
� �
�
CHAPTER 1 INTRODUCTION 3
Power
(MW)
Time
T
Pmax
Pmin
ts te
RU
RD
Figure 1.1 One of many possible power-paths that a dispatchable resource with swing(flexibility) in down/up ramping and power amplitude could be signaled to deliver at its gridlocation during operating period T = [ts, te).
physically characterized set ℙℙm of power paths for period T, each of which m couldfeasibly deliver at a designated grid location during T in response to dispatch signals;and (iv) a performance payment method 𝜙m.
If SCm is cleared, the offer price 𝛼m (if positive) is paid to m either directlyor in amortized payment-schedule form. The offer price thus permits m to cover exante any cost that m would have to incur to ensure the availability of the power-pathsin ℙℙm. This availability cost could include capital investment cost, start-up cost,no-load cost, and opportunity cost. The exercise set 𝕋 ex
m consists of designated timesbetween the close of M(T) and the start of T at which the ISO can exercise SCm,assuming SCm has been cleared. The form of this exercise set determines whetherSCm is a firm contract or a type of option contract.6
The dispatchable power-paths in ℙℙm are characterized in terms of attributessuch as delivery location, start-time, minimum down/up time, active and reactivepower limits, ramp-rate limits, duration limits, and energy capacity. The precise spec-ification of these attributes determines the degree of swing (flexibility) in m’s offeredreserve. Finally, the performance payment method 𝜙m permits resource m to recoverex post any cost that m incurs for verified period-T service performance, i.e. forthe verified period-T delivery of a power-path in ℙℙm in response to dispatch sig-nals. This performance cost could include fuel cost, labor cost, transmission servicecharges, and machinery wear and tear caused by fast ramping.
Reserve offers submitted into M(T) take the form of portfolios of swing con-tracts offered by dispatchable resources for operating period T. These dispatchableresources can include generators, distributed-resource aggregators, and storagefacilities. Reserve offers in firm form effectively constitute regulation reserve,
6As will be clarified in Section 4.3, standard types of option contracts are distinguished by the number andpositioning of their exercise times.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 4�
� �
�
4 CHAPTER 1 INTRODUCTION
whereas reserve offers in option form effectively constitute contingency or planningreserve.
As demonstrated in Chapter 5, these reserve offers can take the stan-dard supply-offer forms required by current US RTO/ISO-managed wholesalepower markets. Examples include: must-run energy blocks; hourly step-functionpower supply schedules with a separate price designated for each power-step; andpower self-scheduled by power traders to secure needed transmission support for thepower outcomes of privately negotiated physically covered bilateral contracts.
However, as is also demonstrated in Chapter 5, the general formulation of aswing contract can accommodate reserve offers with a much broader range of offeredattributes than envisioned in these standard supply offer forms. Moreover, the issuerm of a swing contract SCm can use the performance payment method 𝜙m included inSCm to specify m’s required compensation ex post for dynamic aspects of a deliveredpower-path, such as ramping, duration, and reactive power support, as well as staticaspects such as total delivered energy.
Reserve bids submitted into a swing-contract market M(T) take the form ofprice-sensitive and/or fixed demands for power-path delivery during operating periodT. Reserve bids can be submitted by load-serving entities to service the forecastedloads of their customers during T, and by power traders who need to self-schedulethe power outcomes of privately negotiated physically covered bilateral contracts inorder to secure needed transmission support.
As detailed in Chapters 6–9, an ISO managing a swing-contract market M(T)solves a contract-clearing optimization problem to determine which reserve offersand price-sensitive reserve bids to clear for operating period T. The objective of theISO is to maximize the expected total net benefit of the market participants, condi-tional on initial state conditions and subject to system constraints.
Total net benefit consists of total benefit net of total avoidable cost. The systemconstraints include power balance, transmission line, and reserve constraints. Theseconstraints incorporate, as exogenous inputs: (i) all fixed demands; (ii) all forecastsfor non-dispatched power injection; (iii) all of the power-path attributes included bydispatchable resources in their reserve offers; and (iv) system-wide and zonal reserverequirements set by the ISO to ensure coverage of net load uncertainty sets as a robustmeans of protection against net load forecast errors.
The ISO functions as a clearing house for M(T), collecting payments and over-seeing payouts to market participants. However, the ISO does not have any financialstake in market operations. To maintain this independent status, all net reserve cost7
and transmission service cost incurred through market operations are passed throughto market participants. Net reserve cost is allocated across market participants basedon the relative volatility and size of their net must-service load.8 Transmission service
7Net reserve cost is reserve procurement cost net of any price payments for cleared price-sensitive reservebids and net of any penalty payments for real-time deviations from dispatch signals.8The net must-service load of a market participant at a particular grid location is the amount of itsnon-dispatched power withdrawal at that location, if any, minus the amount of its non-dispatched powerinjection at that location, if any.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 5�
� �
�
CHAPTER 1 INTRODUCTION 5
cost is allocated across market participants based on the power imbalance9 at theirgrid locations.
More generally, Chapter 10 proposes a linked collection of swing-contract mar-kets whose look-ahead horizons for designated future operating periods can range induration from multiple years to minutes. The linkage among these markets is achievedby having the reserve offers and price-sensitive reserve bids cleared in earlier mar-kets be carried forward on the books of the ISO as a portfolio of contracts that can beadaptively updated in subsequent markets. This linkage facilitates reserve procure-ment by permitting a successively refined understanding of resource availability andsystem conditions for future operating periods.
The key features of this Linked Swing-Contract Market Design in comparisonwith current US RTO/ISO-managed wholesale power market designs, elaborated inChapters 10–15, are summarized below:
• permits the robust-control management of uncertain net load
• handles uncertain net load by ensuring flexible dependable reserve supply
• eliminates the need for detailed net load scenario specifications
• facilitates a level playing field for resource participation
• recognizes the forward nature of wholesale power markets
• recognizes all offered product in these forward markets is a form of reserve
• identifies reserve as dispatchable power-paths available for future operations
• requires resources to internally manage commitment and capacity constraints
• permits co-optimization across a wide range of reserve attributes
• ensures settlement time-consistency through two-part pricing
• compensates reserve availability ex ante and reserve deployment ex post
• permits resource owners to cover ex ante their full costs of availability
• permits resource owners to recover ex post their full real-time performancecosts
• eliminates the need for out-of-market payment adjustments
• provides system operators with real-time flexibility for net load balancing
• encourages close following of dispatch signals through performance incentives
• reduces the complexity of market rules.
Chapter 16 considers how current US RTO/ISO-managed day-ahead marketscould gradually transition to a swing-contract market design. As shown, a swing con-tract submitted by a dispatchable resource into a day-ahead market can in principle
9Power imbalance is said to occur at a particular bus in a transmission grid if there is a nonzero net powerinjection at this bus that requires the transmission of power to or from other buses in order to ensure powerbalance across the transmission grid as a whole.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 6�
� �
�
6 CHAPTER 1 INTRODUCTION
be incorporated as follows. First, the swing-contract’s offer price and performancepayment method can be incorporated into the objective function for the optimizationused by the RTO/ISO to solve for generator unit commitments and scheduled dispatchlevels for next-day operations. Second, the power-path attributes designated by thisswing contract can be incorporated into the system constraints for this optimization.
However, in order for this incorporation to result in accurate merit-orderdispatch for next-day operations, the optimization would have to account fully for theexpected total net benefit associated with each possible configuration of generatorunit commitments and scheduled dispatch levels. At present this is not the case. Forexample, the unit commitment costs appearing in the objective function typicallycover (at most) the start-up, no-load, and minimum-run costs of generators, nottheir full availability costs. Also, voltage limits are typically not included among thesystem constraints, thus preventing consideration of the benefits provided by offeredvoltage-support services. In consequence, swing contracts offering diverse dispatch-able power-paths, with explicit offer prices and performance payment methodsensuring full coverage of availability and performance costs, could be incorrectlyomitted from the merit-order dispatch stack on the grounds they are too costly.
To illustrate what might be done to address this issue, Chapter 16 presents anextended day-ahead market optimization in complete analytical form that permits afuller range of costs to be incorporated in the objective function. It is shown, explic-itly, how swing contracts offering dispatchable power-paths with swing (flexibility)in power amplitude and ramp rate can be incorporated into this extended optimization
Market processes Data and signal flows
Wholesale generation and storage
Distributed generation and storage
Contracts
Co
ntracts
Co
ntracts
Co
ntracts
Distributed consumers
and prosumers
Distribution gridDistribution
utilities
Transmission
utilities
High voltage
transmission gridIndependent
system operator
Grid-edge resourceaggregators
Power flows
Figure 1.2 Illustration of an integrated transmission and distribution system.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 7�
� �
�
CHAPTER 1 INTRODUCTION 7
along with standard types of supply offers while still retaining a mixed-integer linearprogramming formulation. The solution of this extended optimization results in anaccurate merit-order dispatch stack.
Chapters 17–18 explore swing-contract support for integrated transmission anddistribution system operations; see Figure 1.2. Special attention is focused on thepossibility that independent distribution system operators (IDSOs), functioning indistribution systems as grid-edge resource aggregators,10 could use swing contractsto facilitate their participation in transmission systems as reserve providers as wellas load-serving entities. The reserve provision of an IDSO could take the form ofswing contracts offering the availability of dispatchable power-paths harnessed fromgrid-edge resources in return for appropriate compensation. This IDSO participationwould permit retail customer interests to be more directly and completely representedat the wholesale power market table.
Potential future research directions are outlined in Chapter 19, and concludingremarks are given in Chapter 20. Glossaries and nomenclature tables for terms usedto describe market operations in both standard and swing-contract forms are providedin Appendix A.
10In this study, a grid-edge resource is defined to be any entity capable of power usage and/or power outputthat is directly connected to a distribution grid. A grid-edge resource aggregator is any entity that managespower usage, power supply, and/or ancillary service provision for a collection of grid-edge resources.
Trim Size: 6in x 9in Tesfatsion670124 c01.tex V2 - 12/07/2020 8:55 P.M. Page 8�
� �
�
Trim Size: 6in x 9in Tesfatsion670124 c02.tex V2 - 11/19/2020 11:37 P.M. Page 9�
� �
�
CHAPTER 2US RTO/ISO-MANAGEDWHOLESALE POWERMARKETS: OVERVIEW
2.1 CHAPTER PREVIEW
The decision to propose a major reformulation of current US RTO/ISO-managedwholesale power markets in this study has been prompted by a careful considera-tion of eight broadly accepted goals for wholesale power market design. A summarystatement of these eight design goals is provided in Section 2.2.
A brief overview of daily operations in current US RTO/ISO-managed whole-sale power markets is provided in Section 2.3. Increasing stresses faced by thesemarkets are discussed in Section 2.4.
As will be elaborated below in Chapter 3, problematic aspects of current USRTO/ISO-managed wholesale power market operations are hindering the ability ofthese markets to deal with the stresses outlined in Section 2.4, hence also their abilityto adhere to the eight design goals outlined in Section 2.2.
2.2 GENERAL GOALS FOR WHOLESALE POWER MARKETDESIGN
Eight broadly accepted goals for wholesale power market design have motivated thisstudy. These goals are based in large part on the goals set forth by Oren [141, SectionII.A] and Tesfatsion et al. [188, Section 2]. Briefly summarized, these goals are asfollows:
Goal-1. Incentive alignment: The market design should be well-aligned with theprivate objectives and constraints of the market participants, thus ensuring theirvoluntary participation.
A New Swing-Contract Design for Wholesale Power Markets, First Edition, Leigh Tesfatsion.© 2021 The Institute of Electrical and Electronics Engineers, Inc. Published 2021 by John Wiley & Sons, Inc.
9
Trim Size: 6in x 9in Tesfatsion670124 c02.tex V2 - 11/19/2020 11:37 P.M. Page 10�
� �
�
10 CHAPTER 2 US RTO/ISO-MANAGED WHOLESALE POWER MARKETS: OVERVIEW
Goal-2. Resource adequacy: The market design should provide incentives for newresources to enter in sufficient quantity to accommodate retirements, de-ratings,and increases in power demand over time while maintaining adequate reserveto address uncertainty and volatility in net load.
Goal-3. Efficiency: The market design should be efficient, i.e. it should not wasteresources. To achieve short-run efficiency, i.e. the non-wastage of existingresources, the design should require the bid/offer-based participation of bothbuyers and sellers. To achieve long-run efficiency, the design should encouragethe development and adoption of new technologies permitting increasedbenefits from power usage as well as lower costs for power production andtransmission.
Goal-4. Reliability and resiliency: The market design should ensure the continualbalancing of net load during normal power system operations, despite weatherevents and other types of anticipated disturbances. The design should also sup-port rapid recovery and return to net load balancing following sudden majordisruptions, such as the loss of a line or a generation unit.
Goal-5. Open access: The market design should be fair, providing an even play-ing field for all potential and actual market participants. It should permit andencourage resources to compete for the provision of reserve and the procure-ment of power-path deliveries, and it should discourage the exercise of marketpower.
Goal-6. Conceptual coherency and transparency: The market design should beconceptually coherent, and market rules and operations under the design shouldbe as transparent as possible.
Goal-7. Minimum administrative intervention: The market design should discour-age ad-hoc rule-making and decision-making by administrators. To further thisgoal, market rules and operations should be based on service requirementsrather than on the physical and operational attributes of resources, to an extentcompatible with the attainment of other design goals. Wherever possible, mech-anisms should be instituted to permit and encourage transition to a design withlimited administrative control.
Goal-8. Supportive of previous reform efforts: The market design should be inaccordance with recent FERC and RTO/ISO initiatives promoting increasedmarket access, pay for performance, demand-side participation, and encour-agement of private initiative.
2.3 US RTO/ISO-MANAGED MARKET OPERATIONS
Bulk power trading in all US electric power regions includes bilateral transactionsmanaged by exchanges (centralized trading platforms), over-the-counter tradingmanaged by dealers, and privately negotiated bilateral contracting managed by
Trim Size: 6in x 9in Tesfatsion670124 c02.tex V2 - 11/19/2020 11:37 P.M. Page 11�
� �
�
2.3 US RTO/ISO-MANAGED MARKET OPERATIONS 11
Figure 2.1 North American RTO/ISO-managed wholesale power markets. Publicdomain: [61].
brokers [55, Chapters 3]. Traditionally, the transmission services supporting thisbulk power trading were provided by regulated high-voltage transmission gridsowned and operated by vertically integrated utilities or other types of entities.
However, in a series of notices culminating in a 2003 White Paper [52],1 theUS Federal Energy Regulatory Commission (FERC) recommended that US elec-tric power regions be re-organized as RTO/ISO-managed wholesale power marketsoperating over RTO/ISO-managed transmission grids. Power traders would thus beprovided an additional means for conducting their bulk power trades: namely, submis-sion of bids (to buy) and offers (to sell) into a wholesale power market. Cleared bidsand offers would automatically be scheduled to ensure needed transmission support.For all other forms of physically covered bulk power trades, transmission supportwould have to be obtained by the self-scheduling of these trades in the wholesalepower market.
Seven US electric power regions are currently operating in accordance withthis FERC design: California Independent System Operator (CAISO), Electric Reli-ability Council of Texas (ERCOT), Independent System Operator for New England(ISO-NE), Midcontinent Independent System Operator (MISO), Independent SystemOperator for New York (NYISO), PJM Interconnection (PJM), and Southwest PowerPool (SPP); see Figure 2.1. Nearly two thirds of US electric power usage is currentlyserviced by these seven regions.
1This white paper is based on seminal work by Schweppe et al. [28, 159, 160] and Hogan [79, 80].