• No results found

Integration of energy storage in power systems

N/A
N/A
Protected

Academic year: 2021

Share "Integration of energy storage in power systems"

Copied!
192
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Integration of energy storage in power systems

Citation for published version (APA):

Lopes Ferreira, H. M. (2017). Integration of energy storage in power systems. Technische Universiteit Eindhoven.

Document status and date: Published: 11/12/2017 Document Version:

Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers) Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be important differences between the submitted version and the official published version of record. People interested in the research are advised to contact the author for the final version of the publication, or visit the DOI to the publisher's website.

• The final author version and the galley proof are versions of the publication after peer review.

• The final published version features the final layout of the paper including the volume, issue and page numbers.

Link to publication

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal.

If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please follow below link for the End User Agreement:

www.tue.nl/taverne

Take down policy

If you believe that this document breaches copyright please contact us at:

openaccess@tue.nl

providing details and we will investigate your claim.

(2)
(3)
(4)

Integration of Energy Storage in Power

Systems

PROEFSCHRIFT

ter verkrijging van de graad van doctor aan de Technische Universiteit Eindhoven, op gezag van de rector magnicus, prof.dr.ir. F.P.T. Baaijens, voor een

commissie aangewezen door het College voor Promoties in het openbaar te verdedigen

op 11 december 2017 om 16.00 uur

door

Hélder Miguel Lopes Ferreira

(5)

Dit proefschrift is goedgekeurd door de promotoren: d prof.ir. W.L. Kling

prof.dr.ir. J.G. Slootweg prof.dr.ir. J. Peças Lopes

Printed by Ipskamp Printing, Enschede Cover design by Stefania Vulpi, Torino, Italy

A catalogue record is available from the Eindhoven University of Technology Library. ISBN: 978-90-386-4343-4

Copyright © 2017 H. M. Lopes Ferreira, Pombal, Portugal

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic, mechanical, including photocopy, recording, or any information storage and retrieval system, without the prior written permission of the copyright owner.

(6)
(7)

Promotors:

prof.dr.ir. J.G. Slootweg, Technische Universiteit Eindhoven prof.dr.ir. J.A. Peças Lopes, Universidade do Porto

Committee:

prof.dr.ir. B. Smolders (chairman), Technische Universiteit Eindhoven prof.dr.ir. J. Driesen, Katholiek Universiteit Leuven

prof.dr. I.G. Kamphuis, Technische Universiteit Eindhoven prof.dr.ir. F. Mulder, Technische Universiteit Delft

(8)

Summary

Sustainability, energy security and aordability are the three major pillars of electricity supply. With the ongoing decarbonization of the electricity sector, renewable energy sources are seen by many as the major option to solve current energy challenges. Their integration in the electricity networks in general, and in smart grids in particular, allows the electricity sector to reduce greenhouse gases and dependency on fossil fuels. However, even if these technologies seem to be obviously respecting the rst pillar, the impact on the other two pillars should also be veried. Furthermore, the current electricity networks were not originally designed for a high integration of variable generation technologies. They were designed based on the possibility of adapting the generation to the load. In such an evolving status quo, the aspects of energy security and aordability become vital. Therefore, the electricity networks need to include exibility (energy security) in a cost-eective manner (aordability).

Among the options to provide energy exibility one can nd energy storage technologies. In this dissertation an example is presented of how energy storage systems can be cost-eective and thus, form a solution for power systems and energy markets associated.

This dissertation rstly analyses the status quo of electricity networks markets around the globe, analysing the background behind the remaining research. The European Union (EU) regulatory framework is presented. Also, the current situation in some of the European Union Member States (MS) is highlighted. Additionally, dierent market structures and mechanisms are presented and examined. Ancillary services and balancing markets are particularly focused, with several examples of integration of these markets. Reserve capacity mechanisms and types of reserves are then examined. The electricity market of The Netherlands, as the case study for this dissertation is analysed in greater detail. Aspects such as Dutch regulation, network congestion, balancing market are focused.

(9)

ii Summary Secondly, a review of modelling energy storage integration is performed. The role of energy storage in electricity networks is focused. Also, regulations and their impact on energy storage systems deployment, here including aspects such as non-market related regulation, power non-market design, storage ownership and right of dispatch, as well as direct nancial support are highlighted. Models concerning the market integration of energy storage are then under the spotlight. Many possible classications, types of problems solved and solving techniques are described and analysed. Energy storage models can be classied by taking several aspects into consideration, e.g., level of detail, perspective, level of complexity, services evaluated, solving techniques, linearity. The dierentiation between system and engineering models and their subtypes are then further studied. Price taker approaches with and without perfect forecast are then scrutinized. Furthermore, hybrid storage system and mutualisation (provision of two or more services) are also focused.

Thirdly, a review of the existing energy storage technologies was performed, taking advantage of the available literature and also of a manufacturer's survey carried out as part of the research. In this survey, key worldwide energy storage technologies manufacturers were invited to provide information on their technologies and devices. Energy storage technologies are described and relevant characteristics are analysed and used to classify the technologies under the spotlight.

Energy storage technologies are capable of providing several services to the network, such as : load levelling, peak shaving / valley lling, load following, spinning reserve, power quality, investment deferral, intermittency mitigation, end-use applications, demand side management, loss reduction, contingency service, black start, and area control. Furthermore, sixteen energy storage technologies are described and analysed. The technologies are compared considering power rating, discharge duration, energy density in terms of weight and volume, power density, eciency, durability in terms of time and cycles, availability and maturity. Categorisation of energy storage technologies is also performed. From this analysis, a journal paper was written.

As mentioned, several services can be provided to the electricity grid by energy storage technologies. These can be divided in two big families of services, which are power market arbitrage on the one hand and ancillary services and balancing markets on the other hand. The hypothesis analysed in this doctoral research was that a system combining arbitrage and ancillary services will expectedly allow a higher level of revenues and present a better cost-benet relation than arbitrage only. As opposed to the existing literature, the model presented includes a dual technology energy storage system, acting in two dierent markets.

A mathematical formulation for the model applied to two Dutch electricity markets is presented. Furthermore, this approach is developed in a way that can be applied in a practical manner to a realistic situation. Also, the model uses a price taker approach and its usage is justied using the literature review. Additionally, the impact of dierent strategies is evaluated in terms of the benets potentially

(10)

Summary iii provided, by using a pattern search optimization algorithm. The strategies analysed include dierentiation between weekend and working days, perfect price forecast versus non-perfect price forecast, and dierent connection strategies between the two devices. In total, three versions and ve variations of the model are presented. An in-depth market price analysis is performed for both day-ahead and balancing markets. This analysis included price histogram depiction and market volatility picturing. The two strategies developed initially using the bulk energy partial model are described and compared. Also are compared the possible four strategies for the high power partial model to act during mode 2 of the balancing market.

The results for the base version of the model are then presented. The device size impact on the revenues is carefully analysed. The results show that this dual approach provides better results than the single technology, as it is possible to observe a revenues increase in 256% between a situation of using only the bulk device or the two devices system described. The eciency benet impact is also veried, and is shown to have an impact up to almost 70% of revenues increase in the cases analysed. Also, the impact of yearly price variation is break down, where it is possible to see that the base year under analysis, 2014 was a below average year, and that the years presenting previous year prices allow higher level of revenues.

The results for the version analysing the impact of the dierentiation between days of the week (workdays and weekends) on the outcome are also presented. It is shown that the result improvement using this version is limited.

The third version is used to analyse the impact of imperfect information on the results. There it is shown that the reduction in the revenues using non-perfect prediction is also limited.

The connection between the two devices were also evaluated. There is shown that, for the market prices under analysis, the version charging the high power device to the maximum when this device is not been used in the market is the one that provides the best result.

The technologies selected for the two practical case studies take in consideration The Netherlands orography conditions. In order to achieve this, compressed air energy storage system (CAES) technologies were selected as bulk energy storage devices. This option envisages the usage of existing depleted gas reservoirs as well of existing underground salt deposits to build dedicated caverns.

For the cost-benet analysis of the two practical case studies are used internal rate of return and payback period calculations. Fixed and variable operation and maintenance costs are also considered. The nal results obtained (around 1027 years of payback time) indicate that systems similar to the one presented in this paper show a realistic potential to be cost eective.

Furthermore, the minimum payback period of 10 years is also within the expected life span of many energy storage technologies.

Also, when compared to the level of investment, the level of revenues potentially provided from the balancing market are comparably high.

(11)

iv Summary These studies show the importance of initial and O&M costs, as well as of eciency on the practical implementation of a dual energy storage system such as the one described. Additionally, the impact of the volume of energy sold on the overall cost-benet result, and the limited impact of the net revenues per kWh is also demonstrated.

The international objectives of RES integration are increasingly ambitious. As the biggest share of the new installations are expected to be of variable RES, electric networks have an impressive challenge ahead.

The increasing integration of variable renewable energy resources will likely increase price uctuations. That increased volatility will foster the protability of energy storage systems such as the one presented in this dissertation.

In the long run, the implementation of a dual energy storage system and strategies like the ones described in this dissertation, using simultaneously two electricity markets, will provide an increase in the exibility of the electric grid. Therefore, considering the increasing need of exibility in the grid, the prospects are positive for systems like the one described in this dissertation.

Furthermore, this increase in exibility, will provide to the countries fostering it, the possibility of increasing their national RES integration, thus decreasing their eventual dependency on fossil fuels while increasing their security of supply. Therefore, the integration of these type of systems will allow for an increase in the exibility in the market and thus the prosecution of the RES integration, as mentioned in international agreements.

(12)

Samenvatting

Duurzaamheid, leveringszekerheid en betaalbaarheid zijn de drie pijlers van de elektriciteitsvoorziening. In het licht van de op handen zijnde decarbonisatie van de elektriciteitssector worden hernieuwbare energiebronnen als de belangrijkste oplossing beschouwd voor de actuele uitdagingen op het vlak van de verduurzaming. De integratie in de elektriciteitsnetwerken, en speciek in smart grids, maakt het voor de elektriciteitssector mogelijk om de uitstoot van broeikasgassen en de afhankelijkheid van fossiele brandstoen te verminderen. Hoewel deze technologieën de duurzaamheid bevorderen, is het van groot belang hun invloed op de leveringszekerheid en de betaalbaarheid van elektriciteit nader te onderzoeken.

Elektriciteitsnetwerken zijn bovendien oorspronkelijk niet ontworpen om grote hoeveelheden hernieuwbare, uctuerende productiemiddelen te accomoderen. uitgangspunt is namelijk dat de elektriciteitsproductie aan de belasting aangepast kan worden. In het licht van dit alles is het des te belangrijker om aandacht te besteden aan de leveringszekerheid en de betaalbaarheid. Want het is noodzakelijk dat het elektriciteitssysteem op een kosteneectieve manier (met het oog op betaalbaarheid) exibiliseren (met het oog op leveringszekerheid).

Energieopslag is één van de mogelijkheden om het elektriciteitssysteem te exibliseren. In dit proefschrift wordt een voorbeeld gepresenteerd hoe ze dit op een kosteneectieve manier kan worden ingezet in de elektriciteitsvoorziening en de daaraan gekoppelde energiemarkten.

Dit proefschrift analyseert eerst de actuele status van verschillende elektriciteitsmarkten in de wereld, waarbij een breed overzicht wordt gegeven van bestaand onderzoek. Het reguleringskader van de Europese Unie (EU) wordt behandeld. Daarnaast wordt ook de huidige situatie in enkele van de lidstaten van de belicht. Verschillende marktstructuren en -mechanismen worden beschreven en onderzocht. De focus ligt daarbij speciek op markten voor ondersteunende diensten, de zogenaamde ancillary services en op balansmarkten, waarbij enkele voorbeelden van de integratie van deze markten voor het voetlicht komen. Daarna

(13)

vi Summary worden mechanismen voor het beschikbaar maken van verschillende categoriën reservecapaciteit behandeld. De elektriciteitsmarkt van Nederland, die gebruikt wordt als case study voor dit proefschrift, wordt diepgaand geanalyseerd. De focus ligt daarbij op zaken als de Nederlandse wet- en regelgeving, congestiemanagement en de (on)balansmarkt.

Daarna volgt een review van het modelleren van integratie van energieopslag in het energiesysteem. De nadruk ligt daarbij op de rol van energieopslag in elektriciteitsnetwerken. Daarnaast wordt ook gekeken naar regulering en de impact hiervan op de inzet van energieopslagsystemen waarbij aandacht is voor niet-markt-gerelateerde regulering, ontwerp van de elektriciteitsmarkt, eigendom van de opslag en recht van afroep, alsmede directe nanciële ondersteuning.

Daarna worden verschillende benaderingen voor het modelleren van de integratie van opslag in de markt onderzocht. Mogelijke classicaties, verschillende typen reeds opgeloste problemen en oplostechnieken worden beschreven en geanalyseerd. Energieopslagmodellen kunnen geclassiceerd aan de hand van een aantal aspecten, zoals detailniveau, perspectief, niveau van complexiteit, geëvalueerde diensten, oplostechniekenen lineariteit.

Als derde is een review van bestaande energieopslagtechnologieën uitgevoerd, gebruikmakend van de beschikbare literatuur en van een enquête onder fabrikanten, die is uitgevoerd als deel van het onderzoek. In deze enquête zijn belangrijke wereldwijde fabrikanten van energieopslag-technologieën uitgenodigd om informatie te leveren over hun technologieën en installaties. Energieopslagtechnologieën worden beschreven en de relevante karakteristieken zijn geanalyseerd en gebruikt om de onderzochte technologieën te classiceren.

Energieopslagtechnologieën kunnen het netwerk op verschillende manieren ondersteunen. Te denken valt aan load levelling, peak shaving / valley lling, load following, draaiende reserve, power quality, uitstellen van investeringen, vermindering van intermittentie, eindgebruiker applicaties, vraagsturing, verliesreductie, contingency services, black start en sturing van uitwisseling tussen verschillende regelzones. Zestien verschillende energieopslagtechnologieën zijn beschreven en geanalyseerd. Deze worden vergeleken op basis van vermogen, ontlaadduur, energiedichtheid in termen van gewicht en volume, vermogensdichtheid, eciëntie, levensduur in termen van tijd en cycli, beschikbaarheid, en technologische volwassenheid. Daarnaast is een categorisering van de onderzochte technologieën uitgevoerd. De resultaten van de analyse zijn verwerkt in een artikel in een wetenschappelijk tijdschrift.

Zoals eerder aangegeven kunnen energieopslagtechnologieën worden ingezet voor verschillende diensten. Daarbij kunnen twee hoofdcategorieën worden onderscheiden. Aan de ene kant arbitrage op de elektriciteitsmarkt, en aan de andere kant ondersteunende diensten en balansmarkten. De hypothese die in dit onderzoek centraal staat, is dat een systeem dat arbitrage en ondersteunende diensten combineert een hogere winst mogelijk maakt en een betere verhouding kent tussen kosten en baten dan een systeem dat alleen wordt ingezet voor arbitrage.

(14)

Summary vii In tegenstelling tot de bestaande literatuur, bestaat het in dit werk gepresenteerde model uit een duaal energieopslagsysteem, dat op twee verschillende markten actief is. Het voorgestelde systeem is wiskundig gemodelleerdwaarbij de Nederlandse markten voor elektriciteit en regelvermogen als uitgangspunt zijn genomen. Het model is op een dusdanige wijze opgezet dat dit praktisch toegepast kan worden in een realistische situatie. Het model gebruikt het principe van een prijsnemer, wat gerechtvaardigd wordt middels het literatuuronderzoek. Aanvullend is de impact van de toepassing van verschillende strategieën geëvalueerd in termen van potentiële baten met een pattern search optimalisatie algoritme. De geanalyseerde strategieën zijn: dierentiatie tussen werk- en weekenddagen, perfecte- versus imperfecte prijsvoorspelling, en verschillende koppelingsstrategieën tussen de twee opslagtechnologieën. In totaal worden drie versies en vijf variaties van het model gepresenteerd.

Een diepgaande analyse van de marktprijs is uitgevoerd voor de spotmarkt en balansmarkten. Onderdeel hiervan waren weergave van prijshistogrammen en visualisatie van de marktvolatiliteit. De twee initieel ontwikkelde strategieën op basis van het bulk energie deelmodel worden beschreven en vergeleken. Daarnaast wordt een vergelijking gemaakt tussen de vier mogelijke strategieën voor het hoog vermogen deelmodel.

Vervolgens worden de resultaten van de basis versie van het model gepresenteerd. De impact van de opslagcapaciteit van het apparaat op de inkomsten is zorgvuldig geanalyseerd. De resultaten laten zien dat deze duale aanpak beter presteert dan die met een enkele technologie; een groei in inkomsten van 256% kan worden waargenomen bij het beschreven twee-apparaten systeem ten opzichte van het systeem met alleen de bulk opslag. Ook is de positieve impact van de eciëntie geverieerd, welke tot bijna 70% groei van inkomsten laat zien in de geanalyseerd cases. Ook is de impact van de jaarlijkse variatie in prijs uiteengezet, wat laat zien dat het basisjaar in de analyse, 2014, wat dit betreft een benedengemiddeld jaar was en dat prijzen in de voorgaande jaren representeren hogere inkomsten mogelijk maken.

De resultaten van de versie die de impact van de dierentiatie tussen werk- en weekenddagen op de uitkomst analyseert worden ook gepresenteerd. Dit leidt tot een relatief beperkte verbetering van de resultaten.

De derde versie wordt gebruikt om de impact van imperfecte informatie op de resultaten te analyseren. Die laat zien dat de reductie van inkomsten met imperfecte voorspelling ook beperkt is.

De koppeling tussen de twee apparaten is ook geëvalueerd. Voor de geanalyseerde marktprijzen laat de versie waarin het hoog vermogen apparaat tot het maximum opgeladen wordt als het niet in de markt gebruikt wordt, de beste resultaten zien.

Voor de twee praktische case studies zijn technologieën geselecteerd, rekening houdend met de orograsche condities van Nederland. Om dit te bereiken zijn technologieën voor energieopslag door gecomprimeerde lucht (CAES) technologieën geselecteerd als bulk energieopslag apparaten. Deze optie voorziet het gebruik van

(15)

viii Summary bestaande uitgeputte gasvelden en ondergrondse zoutmijnen om de voor CAES vereiste grotten te creëren.

Voor de kosten-baten analyse van de twee praktische case studies zijn berekeningen van de ROI (return on investment) en de terugverdientijd gebruikt. Hierbij zijn zowel de vaste en variabele operationele kosten als de onderhoudskosten meegenomen. De uiteindelijke resultaten (terugverdientijd ca. 10-27 jaar) wijzen er op dat systemen, zoals die gepresenteerd in dit proefschrift, een realistisch potentieel hebben om kosteneectief te zijn.

De minimum terugverdientijd van 10 jaar is bovendien ook binnen de verwachte levensduur van veel energieopslagtechnologieën.

Vergeleken met de hoogte van de investering, zijn de potentiële inkomsten van de balansmarkt naar verhouding hoog.

Deze studies laten de relevantie zien van zowel de investering en de exploitatiekosten als de eciëntie voor de haalbaarheid van een duaal energieopslagsysteem zoals in dit proefschrift wordt beschreven. Daarbij worden ook de impact van het verkochte energievolume op de kosten en de baten en het beperkte eect van de netto opbrengsten per kWh inzichtelijk gemaakt.

De internationale doelstellingen aangaande de integratie van hernieuwbare energiebronnen worden steeds ambitieuzer. Vanwege de uctuerende en niet of nauwelijks stuurbare productie van hernieuwbare energiebronnen leidt dit tot een forse uitdaging voor het elektriciteitssysteem

De toenemende integratie van variabele hernieuwbare energiebronnen zal naar alle waarschijnlijkheid de prijsuctuaties vergroten. De vergrote volatiliteit zal de rentabiliteit van energieopslagsystemen zoals die in dit proefschrift gepresenteerd worden, bevorderen.

Op de lange termijn zal de implementatie van een duaal energieopslagsysteem, alsmede de in dit proefschrift beschreven strategieën die simultaan gebruik maken van twee elektriciteitsmarkten, de exibiliteit in het elektriciteitsnetwerk vergroten. Wanneer wij rekening houden met de groeiende behoefte aan exibiliteit in het netwerk, zijn de vooruitzichten voor systemen, zoals beschreven in dit proefschrift, positief.

De groei van exibiliteit biedt de mogelijkheid voor het integreren van meer hernieuwbare energiebronnen in het energiesysteem. Daarmee reduceren landen hun afhankelijkheid van fossiele brandstoen en verhogen zij de leveringszekerheid. Daarmee wordt ook invulling gegeven aan internationale verdragen ten aanzien van het vergroten van de bijdrage van hernieuwbare energiebronnen aan het energieverbruik.

(16)

Contents

Summary i Samenvatting v Samenvatting v Glossary xv 1 Introduction 1

1.1 Research objectives and questions . . . 2

1.2 Scope and boundary conditions . . . 2

1.3 Approach . . . 3

1.4 Dissertation outline . . . 3

2 Electricity networks and markets: background 5 2.1 Introduction . . . 5

2.2 European status . . . 6

2.2.1 European Union regulatory framework . . . 6

2.2.2 Current situation in the European Union . . . 9

2.3 Energy market structure . . . 12

2.3.1 Introduction . . . 12

2.3.2 Market design . . . 14

2.3.3 Ancillary service markets . . . 14

2.3.4 Balancing markets . . . 16

2.3.5 Reserve capacity mechanisms . . . 18

2.4 Electricity market structure of the Netherlands . . . 21

2.4.1 Regulation . . . 22

2.4.2 Network congestion . . . 23

(17)

x CONTENTS

2.4.3 Balancing market . . . 23

2.5 Summary . . . 24

3 Modelling energy storage integration - state of the art 25 3.1 Introduction . . . 25

3.2 Energy storage role . . . 26

3.3 Regulation and its impact on energy storage systems deployment . . 27

3.3.1 Non-market-related regulation . . . 28

3.3.2 Power market design . . . 28

3.3.3 Storage ownership and right of dispatch . . . 28

3.3.4 Direct nancial support . . . 29

3.4 Models concerning market integration of energy storage . . . 29

3.5 System models . . . 33

3.5.1 Energy system models . . . 33

3.5.2 Market models . . . 34

3.5.3 Simplied systems . . . 34

3.5.4 Network models . . . 34

3.5.5 Island systems . . . 35

3.6 Engineering models . . . 35

3.6.1 Price taker with perfect forecast approach . . . 35

3.6.2 Price taker approach without perfect forecast . . . 36

3.6.3 Hybrid storage systems modelling . . . 37

3.6.4 Providing two or more services (mutualisation) . . . 37

3.6.5 Categorisation of energy storage technologies applications . . 39

3.7 Summary and conclusions . . . 40

4 Characterisation of electrical energy storage technologies 41 4.1 Introduction . . . 41

4.2 Technologies . . . 41

4.2.1 Pumped Hydroelectric Energy Storage (PHES) . . . 41

4.2.2 Compressed Air Energy Storage (CAES) . . . 42

4.2.3 Chemical Batteries . . . 42

4.2.4 Flow Batteries . . . 44

4.2.5 Metal-Air Batteries . . . 45

4.2.6 Flywheels . . . 45

4.2.7 Superconducting Magnetic Energy Storage (SMES) . . . 45

4.2.8 Super Capacitors . . . 46

4.2.9 Hydrogen storage systems (with Fuel Cells) . . . 46

4.2.10 Thermal energy storage (TES) . . . 46

4.3 Comparison of technologies . . . 47

4.4 Characteristics of storage technologies . . . 50

4.4.1 Eciency . . . 50

(18)

CONTENTS xi

4.4.3 Energy and Power Density . . . 52

4.4.4 Reliability . . . 52

4.4.5 Response time . . . 53

4.4.6 Storage capability: power vs energy . . . 53

4.5 Categorisation of energy storage technologies . . . 53

4.6 Methods of connection . . . 54

4.7 Conclusions . . . 55

5 Mathematical formulation and implementation 59 5.1 Introduction . . . 59

5.2 Background . . . 59

5.3 Base version: single set of thresholds (SST) . . . 61

5.3.1 SST formulation . . . 61

5.3.2 SST implementation . . . 66

5.4 Working days and weekends dierentiation: dual set of thresholds version (DST) . . . 68

5.4.1 DST formulation . . . 68

5.4.2 DST implementation . . . 68

5.5 Dual modied version: including previous day balancing market prices (DST+) . . . 70

5.5.1 DST+ formulation . . . 71

5.5.2 DST+ implementation . . . 71

5.6 Energy transfer variants . . . 72

5.7 Partial models and optimisation . . . 74

5.7.1 Bulk partial model . . . 74

5.7.2 High power partial model . . . 74

5.7.3 Full model . . . 75

5.8 Optimisation algorithms . . . 75

5.8.1 Particle swarm optimisation . . . 76

5.8.2 Pattern search . . . 77

5.9 Summary . . . 79

6 Results and analysis 81 6.1 Introduction . . . 81

6.2 Markets price analysis . . . 81

6.3 Strategies developed initially . . . 87

6.3.1 Average based set of thresholds (ABST) . . . 87

6.3.2 Sorting strategy (SORT) . . . 87

6.3.3 Comparison of the two strategies . . . 90

6.4 High power partial model . . . 90

6.5 Full model results . . . 91

(19)

xii CONTENTS 6.5.2 Impact in the results of the dierentiation between the days

of the week (workdays and weekends) - DST version . . . 98

6.5.3 Impact of imperfect information - DST+ version . . . 100

6.6 Connection between the two devices: analysis of variants . . . 100

6.6.1 Comparison of the results of the dierent variants . . . 104

6.7 Cost-benet analysis . . . 104

6.7.1 Case study 1 . . . 109

6.7.2 Case study 2 . . . 109

6.7.3 Comparison . . . 111

6.8 Summary and Discussion . . . 112

7 Conclusions, contributions and further research 117 7.1 Conclusions . . . 117

7.2 Contributions . . . 120

7.3 Recommendations and further research . . . 121

A Variant results I

A.1 Variant B . . . I A.2 Variant D . . . IV A.3 Variant E . . . VI A.4 Variant Z . . . VIII

B Particle swarm optimization convergence XI

C Function study XV

C.1 Single set of thresholds (SST) . . . XV C.2 Dual set of thresholds (DST) . . . XV C.3 Dual set of thresholds strategy using previous day average for the

balancing market (DST+) . . . XV

D Additional results XXI

D.1 Cost benet analysis . . . XXII D.1.1 Case study 1 . . . XXII D.1.2 Case study 2 . . . XXII

References XXIII

List of publications XXXVII

Journal Papers . . . XXXVII Book Chapters . . . XXXVII Conference Papers . . . XXXVII

(20)

CONTENTS xiii

(21)
(22)

Glossary

(23)

xvi Glossary

List of abbreviations

The following abbreviations are used in this dissertation:

AACAES Adiabatic advanced compressed air energy storage

ABST Average based set of thresholds

AC Alternating current

ACER Agency of Coordination of Energy Regulators

APX-Endex Dutch day-ahead electricity market

DC Direct current

BRP Balancing responsible party

BSP Balancing service provider

CAES Compressed air energy storage

CENER Spanish National Renewable Energy Centre

DG Distributed generation

DSO Distribution system operator

DST dual set of thresholds

DST+ Dual set of thresholds including previous day balancing

market prices

D-CAES Diabatic compressed air energy storage

ES Energy storage

ESS Energy storage systems

EV Electric vehicles

EC European Commission

ENTSOs European Networks of Transmission System Operators

ENTSO-E European Network of Transmission System Operators

for Electricity

EU European Union

FCR Frequency containment reserves

FRR Frequency restoration reserves

FW Flywheel

GDP Gross domestic production

ICT Information and communication technologies

IRR Internal rate of return

ISO Independent System Operator

Li-ion Lithium-ion batteries

MTBF Mean time between failures

MTTR Mean Time To Recovery

MS EU Member States

NGET National Grid Electricity Transmission

NaS Sodium-sulphur batteries

Ni-Cd Nickel-cadmium batteries

(24)

Glossary xvii

PHES Pumped hydroelectric energy storage

PBP Payback period

PS Pattern search

PSB Polysulphide bromide ow batteries

PSO Particle swarm optimisation

PTU Programme time unit

PV Photovoltaic panels

RES Renewable energy sources

RR Replacement reserves

SMES Superconducting magnetic energy storage

SoC State of charge

SORT Sorting strategy

SST Single set of thresholds

TES Thermal energy storage

TSO Transmission System Operator

V2G Vehicle-to-grid

VRB Vanadium redox ow batteries

Zebra Sodium-nickel-chloride batteries

(25)

xviii Glossary

List of symbols and notations

Below list the main symbols used in this dissertation, their meaning, and units (if applicable).

γ1,δ,t1,m1,δ,t1 energy reserved by device 1 to be transferred to device 2 on day δ, time step t1and mode m1,δ,t1 (MWh)

δ day, element of set D

∆t time dierence between two time steps1

j energy rating of device j (kW)

ηjD discharging eciency of device j, ηjD∈ [0, 1]

ηjC charging eciency of device j, ηCj ∈ [0, 1]

ι internal rate of return

κj variable operation and maintenance costs for device j

(e/kWh)

λ2,δ,t2,m2,δ,t2 energy received by device 2, transferred from device 1, on day δ, time step t2and mode m2,δ,t2 (MWh)

µj cost per unit of power of device j (e/kW)

ξj cost per unit of energy of device j (e/kWh)

πj,δB minimum buying price for market j on day δ (e/MWh)

πj,δS minimum selling price for market j on day δ (e/MWh)

ρj power rating of device j (kW)

σj,δ,mδ j,δ,tj

t historical price volatility in market j for day δ, mode m

j,δ,tj and time dierence δt

φ1,δ,t1,m1,δ,t1 energy reserved on device 1 and not used to supply device 2on day δ, time step t1 and mode m1,δ,t1 (MWh)

ψj minimum payback period for the technology j under

analysis (years) ψmj,δ,tj

Tj total number of time steps on day δ for which market j is

in mode mj,δ,tj

D set of days under analysis

hj,k(δ)B buying threshold for market j and day type k(δ), hj,k(δ)B ∈ [0, 1]

hj,k(δ)S selling threshold for market j and day type k(δ), hj,k(δ)S ∈ [0, 1]

J set of electricity markets and of energy storage devices,

J = {1, 2}

j electricity market/storage device, element of set J

k(δ) indication of weekend/working days, k(δ) ∈ {1, 2}

1usual values: 24 time units for an hourly market, 96 time units for market with quarters of an

(26)

Glossary xix

Mj set of possible modes for market j, M1 = {0}, M2 =

{−1, 0, 1, 2}

mj,δ,tj mode of market j on day δ and time step t

j, element of set

Mj

nX number of elements of set X

oj yearly xed operation and maintenance costs for device j

(e)

pj,δ,tj,mj,δ,tj energy price in market j, day δ, time step t

j and mode

mj,δ,tj (e/MWh)

pj,δ,(tj−∆t),mj,δ,(tj −∆t) energy price in market j, day δ, time step (t

j− ∆t) and

mode mj,δ,(tj−∆t)(e/MWh) pj,δ,tj,mj,δ,tj

B buying energy price in market j, day δ, time step tj and

mode mj,δ,tj (e/MWh) pj,δ,tj,mj,δ,tj

S selling energy price in market j, mode m

j,δ,tj, day δ and time step tj (e/MWh)

qj,δ,tj,mj,δ,tj

B energy quantity bought in market j, mode m

j,δ,tj, day δ and time step tj (MWh)

qj,δ,tj,mj,δ,tj

C quantity of energy charged by device j, on day δ, time step

tj and mode mj,δ,tj (MWh)

qC,maxj maximum amount of energy that device j can charge in one time step (MWh)

qj,δ,tj,mj,δ,tj

D quantity of energy discharged by device j, on day δ, time

step tj and mode mj,δ,tj (MWh)

qD,maxj maximum amount of energy that device j can discharge in one time step (MWh)

qj,δ,tj,mj,δ,tj

S energy quantity sold in market j, mode mj,δ,tj, day δ and

time step tj (MWh)

Tj set of time steps for market j within a day2

tj element of Tj;time step for market j within a day

u vector of buying and selling price thresholds to be optimised

u∗ vector of optimal buying and selling price thresholds

vj,δ,tj,mj,δ,tj

∆t price return, ratio between market j prices on day δ, with

time step tj and with time step (tj− ∆t)

vj,δ,mj,δ,tj mean value of the price return in market j, day δ and mode

mj,δ,tj

xj,δ,tj,mj,δ,tj state of charge of device j on day δ, time step t

j and mode

mj,δ,tj (MWh)

2usual values: 24 time units for an hourly market, 96 time units for market with quarters of an

(27)

xx Glossary

xjmin minimum state of charge of device j (MWh)

xj

max maximum state of charge of device j (MWh)

(28)

Chapter

1

Introduction

Since the times of industrial revolution humanity has become more and more dependent on the availability of energy, in its several forms. This has increased during the last century, by making electricity a critical energy carrier for our civilization. Nowadays, electricity is a major commodity and used everywhere, from households appliances to industrial applications, including information and communication technologies (ICT), hospitals and schools. Its usage has increased productivity and comfort in all aspects of modern everyday life. Electricity is the corner stone for our current way of life and a driver for wealth. The relation between gross domestic production (GDP) and electricity consumption is well known and shows its importance for modern life.

In the last decades, humanity became increasingly aware of the environmental impact of the industrialized way of life, especially in what concerns energy usage and electricity generation. This awareness has lead to sustainability policy goals being brought up around the globe by most of the world's nations. These sustainability policy goals have triggered an increasing attention towards more ecient and cleaner generation and consumption as well a research on new ways of the power system to work.

Another dimension, not less relevant than sustainability is the energy security dimension, especially in what concerns the middle and the long term aspects. The increasing volatility of oil and gas prices, as well as the political instability of some of the oil and gas producing countries has increased political concerns concerning the dependency on fossil fuels. As a result, new power plants, which are more ecient and cleaner were built or the older ones retrotted. While, at the same time innovative renewable energy generation, such as wind or solar photovoltaic were developed and saw an increasing deployment.

Furthermore, after the Fukushima incident, nuclear electricity generation raised new political and safety concerns. These concerns have downgraded the expectations on this technology as a long term solution. This situation has directed more attention

(29)

2 Introduction towards renewable energy sources (RES) as not only a sustainable means of energy generation, but also as a safe one. Among RES, energy generation technologies can be classied according to their climate conditions dependency in two groups:

i. variable RES, directly depending on the weather conditions.

ii. slow-variable and non-variable RES, where this dependency is smoothed or indirect.

The technologies of the second type have reached a stage where new sites are not available or cost eective (e.g., hydro) or are not yet mature enough for massive deployment (e.g., concentrated solar power with thermal energy storage). Therefore, most of the new RES generation sites are based on technologies of the rst type, variable RES solutions, such as solar photovoltaic and wind generation, both onshore and oshore.

At the same type, due to the transportation decarbonization goals, an unprecedented level of deployment of electric vehicles (EV) is also expected for the coming years.

The electrical grids and the related markets where not originally designed for such developments. Thus, they have to be adapted and evolve to deal with this new situation. One of the important aspects of this evolution, is the provision of exibility in an cost eective manner.

Energy storage is one of the solutions to provide such exibility and will be the one studied in this dissertation.

1.1 Research objectives and questions

The main research objective is to verify if energy storage systems can be already cost-eective nowadays and thus facilitate the evolution towards a more sustainable electricity power system.

The research questions are the following:

ˆ Which are the most relevant energy storage technologies for the dierent power system applications?

ˆ Will a dual system (with 2 technologies) providing two services be more attractive than a traditional energy storage system? If so, in which conditions and which are the advantages?

ˆ Are there market and regulatory adjustments that would support the integration of energy storage technologies?

1.2 Scope and boundary conditions

The scope of this research was dened as the characterization of the existing energy storage technologies and evaluation of the integration of those technologies

(30)

1.3. Approach 3 in the system, in the current regulatory framework. Furthermore, the strengths, applications and limitations of the application of energy storage technologies will be highlighted. Out of scope are, for instance, future markets and their prices, as well as new technologies for storage which are still at experimental phase. Also out of scope are price maker approaches as well as end of life cycle costs.

1.3 Approach

The approach used to implement this research is the following:

ˆ Description of current developments in power systems and electricity markets; ˆ Characterization of existing energy storage technologies and of the case study

markets;

ˆ Evaluation of the state of the art concerning scientic analysis of integration of energy storage technologies and its modelling;

ˆ Formulation of the problem;

ˆ Modelling of the case studies and its markets (day-ahead, balancing); ˆ Analysis of the results and conclusions.

1.4 Dissertation outline

The dissertation structure was based on the procedure described above. The remaining chapters of this dissertation are as follows:

Chapter 2 - Electricity networks and markets: background Chapter 3 - Modelling energy storage integration: state of the art Chapter 4 - Characterization of electrical energy storage technologies Chapter 5 - Model formulation and Implementation

Chapter 6 - Results Chapter 7 - Conclusions

(31)

4 Introduction

(32)

Chapter

2

Electricity networks and markets:

background

2.1 Introduction

As mentioned in Chapter 1 this dissertation analyses the potential integration of energy storage (ES) in modern electricity networks and markets as a cost-eective exibility provider. Therefore, as an introduction to the topic, a general overview of electricity energy systems and the associated value chain is performed in this chapter. Aspects focused on include energy policy, sustainability goals, electricity systems regulation, integration of renewable energy sources (RES) and electricity markets organization.

Global energy policy goals include further integration of renewable generation technologies in energy markets. As an example of this integration, the European Union (EU) has set as a goal to have 20% of energy generated from RES by 2020. Furthermore, it expects this gure to increase to at least 27% by 2030 [1]. Objectives for 2050 are even more ambitious, with carbon emissions set to drop between 80% and 95% [2] compared to those of 1990. All around the world, including in countries such as China [3], Japan [4], New Zealand [5], the United States of America [6, 7], and Turkey [8], power systems are being prepared and adapted for an increasing deployment of renewable generation technology.

In conjunction with RES, the integration of other recent technologies such as electric vehicles (EV), as well as unbundling and other developments in regulation are changing the structure of the electricity sector. Unbundling is the separation of the dierent links of the energy value chain, in order to allow a free, liberalised market ensuring fair and competitive perspective for new entrants. All these changes mentioned call for adjustments in planning as well as operation of electrical power systems: they need to become more exible.

This exibility can be achieved through the adoption of several technologies

(33)

6 Electricity networks and markets: background which can be combined or used independently. The exibility options currently available include energy storage systems (ESS), cross-border interconnection capacity, RES curtailment, exible conventional generation, active demand response, and management of EV charging and even power delivery from vehicles'batteries to the grid - vehicle-to-grid (V2G) [9]. Furthermore, the deployment of these technologies and techniques has to be performed in agreement with the three pillars of electrical power systems, which are reliability, aordability, and socio-economic and environmental sustainability. The balance between these aspects is aected by regulatory and policy documents concerning electricity systems, which comprise both physical grids as well as electricity markets.

This chapter analyses those aspects. Firstly, the evolution of the EU regulatory framework is outlined. Secondly, several case studies are depicted. Thirdly, electricity energy markets structure design are discussed. Finally, the Dutch electricity market is analysed in more detail.

2.2 European status

2.2.1 European Union regulatory framework

As part of the European Union policy, the opening of the electricity market to competition has been the strategy used to achieve energy policy goals. The liberalisation of the European electricity markets was initiated by the Electricity Directive (1996) [10] and has been evolving since then. The central point of this legislation is the division of the electricity value chain in dierent links with specic roles. Some parts of the value chain are considered part of the competitive market, while others are seen as natural monopolies. In this Directive are considered both transmission and distribution networks as natural monopolies while generation, trade and supply are dened as competitive activities.

However, the analysis performed in 2006 and 2007 by the European Commission (EC) concerning the application of the second energy package found several problems in the European Energy markets. According to the speech of European Commissioner Neelie Kroes for Competition Policy in 19/09/2007 [11], ve main problems have been detected:

ˆ First, continuing high levels of concentration so incumbents maintain market power;

ˆ Second, vertical foreclosure, as the old monopolists continue to own the energy infrastructure;

ˆ Third, low levels of cross-border trade, due to insucient interconnector capacity land to contractual congestion since spare physical capacity is not always released;

(34)

2.2. European status 7 ˆ Fourth, lack of transparency about operations in the wholesale energy sector, which makes it dicult for new entrants to understand how the markets work in practice and the risks that they take on;

ˆ Finally, lack of condence that wholesale energy prices are the result of meaningful competition. [11]

To solve these problems, the Third Energy Package was enacted in 2007. This package contains several regulations that focus on both the electricity and gas sector. It includes features such as consumer choice, fairer prices, cleaner energy and security. The EC also presented several measures/strategies through which improvements can be made in these areas, such as:

ˆ Separation of production and supply from transmission networks; ˆ Facilitating cross-border trade in energy;

ˆ Increase responsibility of national regulators; ˆ Cross-border collaboration and investment;

ˆ Greater market transparency on network operation and supply.

Afterwards, the EC established in 2008 the Energy and Climate Package with 20-20-20 goals for the year 2020 [12]. These goals include:

ˆ Improvement of energy eciency by 20%;

ˆ Reduction of gas emissions by 20% from the 1990 level;

ˆ Integration of renewable energies so that they account for 20% of primary energy consumption.

In the electricity market, the EU's most signicant action is the issuing of Directive 2009/72/EC [13]. This Directive is also part of the Third Electricity and Gas Liberalisation Package, and focuses on common rules for the internal electricity market. It regulates the whole process to unbundle the electricity system, including non-discriminatory access and energy security as essential elements. Also, it denes ocially several concepts, such as energy eciency/demand side management and distributed generation. Additionally, it encourages EU Member States (MS) to modernise electricity networks, amongst others through the adoption of smart grids and smart meters.

Furthermore, this Directive denes several measures to protect consumers and increase transparency. EU regulation denes dierent types of unbundling, which are described in gure 2.1 [14]. Vertical integration (or as dened in the Directive, vertically integrated undertaking) is where a company controls the full value chain from generation to supply electricity to the nal consumer. This option was the most common solution in many MS before European legislation came into force. The other two approaches, network operation unbundling and ownership unbundling are

(35)

8 Electricity networks and markets: background

Figure 2.1: Comparison between dierent forms of unbundling

strategies to separate the activities of the energy chain, in accordance with current EU legislation.

In 2010 the European Commission (EC) adopted the Energy 2020-A strategy for competitive, sustainable and secure energy [15] which focuses on:

ˆ Achieving a higher energy eciency in Europe; ˆ Building a more integrated European energy market; ˆ Involving consumers as stakeholders of the power system;

ˆ Strengthening Europe's position as a leader in energy technology and innovation.

Other relevant legislative documents focus on: ˆ Energy Storage:

 Directive 2009/28/EC - Promotion of the use of energy from RES ˆ Demand Side Management:

 Directive 2005/89/EC - Support to energy eciency, demand side management, RES and DG;

 Directive 2006/32/EC - Energy end-use eciency and energy services; ˆ Electric Vehicles:

 Directive 2009/33/EC - Promotion of clean and energy-ecient road transport vehicles;

(36)

2.2. European status 9  Framework Directive 2007/46 - Framework for the approval of motor vehicles and their trailers, and of systems, components and separate technical units intended for such vehicles;

 Directive 2006/95/EC - Harmonisation of the laws of Member States relating to electrical equipment designed for use within certain voltage limits.

The Agency of Coordination of Energy Regulators (ACER) [16] was created to improve the regulation of energy by national regulators. ACER has authority over cross-border network capacity and regulates the European Networks of Transmission System Operators (ENTSOs). There are two ENTSOs, one for the electricity sector (ENTSO-E) and another for the gas sector (ENTSO-G). The European Network of Transmission System Operators for Electricity (ENTSO-E) represents the electricity transmission system operators (TSOs) in Europe [17]. The European distribution system operators are represented through several institutions, where one of the most relevant is the EDSO for Smart Grids" [18]. However, other associations represent or exist in the sector, such as Eurelectric, which represents the common interests of the electricity industry [19] and is also involved in many activities concerning DSO regulation. Furthermore, small DSOs are grouped under the umbrella of two sector associations: CEDEC [20] and GEODE [21].

2.2.2 Current situation in the European Union

Currently, the full integration of the Third Electricity and Gas Liberalisation Package remains a challenge for the EU energy markets. The market is fragmented due to dierences in the liberalisation process among the various EU countries (range of starting points and development speeds, various market designs . . . ) which results in a low degree of transparency and limited competition. Furthermore, the lack of interconnection capacity between countries may cause signicant discrepancies in electricity prices. Figure 2.2 depicts the average baseload electricity prices for the third quarter of 2015. The aforementioned discrepancies in prices are evident. 2.2.2.1 Iberian Peninsula

Since 2006, the two Iberian countries, Portugal and Spain, have merged their markets. The diagram of the resulting markets and applicable regulations for the electricity sector is depicted in gure 2.3.

With the implementation of the third energy package, both countries chose to follow the TSO model. In Portugal, Redes Energéticas Nacionais (REN) is the TSO for both electricity and gas. In Spain, the electricity TSO is Red Eléctrica de España (REE), and the main gas TSO is Enagás.

(37)

10 Electricity networks and markets: background

Figure 2.2: Comparison of average wholesale baseload electricity prices in the third quarter of 2015, from [22].

2.2.2.2 Great Britain

England, Wales and Scotland's markets present a large number of electricity generation companies. The transmission operator (National Grid Electricity Transmission - NGET) owns and operates the electricity transmission network in England and Wales and operates the transmission network in Scotland. As part of the European unbundling, NGET regulated and commercial activities form dierent juridical entities. A similar situation happens in the gas sector, where National Grid Gas, another branch of the same company, is the transmission system operator [24].

(38)

2.2. European status 11

Figure 2.3: MIBEL regulations diagram, based on [23].

2.2.2.3 France

In France, the state controls both electricity and gas transmission system operators. The government's dominance over the electricity sector is largely driven by the fact that most of the country's power is generated by nuclear power plants. The French transmission network is unbundled at the minimum level dened by the Directives. Therefore, the network is separated administratively, so the network operator (RTE) is part of the production side (EdF). The distribution network operators are also unbundled administratively [24].

2.2.2.4 Belgium

In Belgium, there is a similar situation to that of France, with the state controlling both electricity and gas transmission system operatos. In what concerns energy generation, Electrabel (part of Suez Group) is the largest electricity producer in

(39)

12 Electricity networks and markets: background Belgium and the Netherlands, controlling up to 20% of the Dutch generation capacity. In Belgium, the network operator, Elia, was created as a result of the legal unbundling in the electricity market. In 2005, Elia set up the power exchange Belpex, which is listed in APX, the Amsterdam power spot exchange. In Belgium, both the TSO and the DSO are legally unbundled from commercial activities [24]. 2.2.2.5 Germany

Historically, Germany's energy supply has been under private control or is managed by municipalities, which has resulted in a high degree of autonomy. The German electricity market is dominated by four TSOs (RWE, TenneT, 50 Herz and EnBW), each responsible for a network area. This structure, in combination with the congestion prevailing at all German borders with the exception of Austria, limits the development of competition. Although the market was opened formally in 1998, the introduction of eective competition was not stimulated, with the network taris being established by companies. After 2005, liberalisation had a boost with the introduction of a specic regulatory authority. This has supported the emergence of new suppliers in the market [24].

2.3 Energy market structure

2.3.1 Introduction

The electric energy system can be seen as composed of two dierent layers:

ˆ Physical layers where electricity ows from generation, through transmission and, usually, distribution networks to the consumers.

ˆ Energy market layers, where energy is traded between the main players operating in the system.

Concerning the physical side, the traditional linear model from generation to consumption is displayed in gure 2.4.

(40)

2.3. Energy market structure 13 With the deployment of distributed generation, it was modied as can be seen in gure 2.5.

Figure 2.5: Energy ows in the electric system considering the traditional network and distributed generation (DG), adapted from [14]

Considering energy storage integration at dierent levels, the electric system would change to something close to what is depicted in gure 2.6.

Figure 2.6: Energy ows in the electric system considering the traditional network, distributed generation and the several levels of energy storage, from [14]

(41)

14 Electricity networks and markets: background The economic layer includes several markets, stakeholders and their interactions. Here are included electricity generation companies, energy service companies, transmission companies, distribution and supply companies, as well as the primary fuels markets (such as coal, gas and oil), electricity markets in dierent timeframes, and other markets such as the transmission rights market, balancing market, reactive power market, and other ancillary service markets [25].

2.3.2 Market design

Most electricity markets have been restructured in the last decades in order to foster competition and sustainability. However, there are several constraints to the reforms made, such as market size, degree of insulation, and presence of indigenous energy sources or lack thereof.

Furthermore, dierent variables must be taken into account, depending on the type of market selected. Therefore, a choice between an integrated market, where the physical and economic aspects are strongly interconnected or a decentralised market (preferred in Europe) is the rst point. Another variable is the degree of privatisation, ranging from a government monopoly to full privatisation.

Regarding transmission and distribution networks, the regulation of network taris, access to third parties and the degree of unbundling of networks are essential for competitiveness in the market. Another point is the setting up of a regulator and the degree to which regulatory problems can be solved in detail. Finally, the ex-ante situation is of decisive importance in the energy market restructuring process (e.g. the number of generation companies, the degree to which they are integrated, their ownership structure, the generation and network structure, the interconnection capacity with neighbouring countries, etc.) [26, 27].

2.3.3 Ancillary service markets

Ancillary services facilitate and support the continuous ow of electricity so that supply and demand are balanced and kept within the required bandwidths [28]. They include a variety of operations beyond generation and transmission that are required to maintain grid reliability [29],such as reactive power and voltage control, frequency control, energy balancing and maintaining spinning reserve and operating reserve.

Traditionally, ancillary services have been provided by generators. However, the development of technologies such as energy storage, demand side response and electric vehicles have led to a shift in the technologies that could potentially used to provide these services [29]. Furthermore, until recently, demand was inexible. However, with the advent of new approaches and technologies this is changing and further evolution is expected in this eld.

The liberalisation of the energy sector, along with an increasing integration of variable renewable energy generation has promoted the importance of ancillary

(42)

2.3. Energy market structure 15 services. They become even more signicant as higher goals are set for the integration of renewable energy generation.

Therefore, especially in markets with a high penetration of variable renewable energy generation or with a high level of liberalisation, the need for these supporting services has increased. This situation has led to the creation of specialised ancillary service markets. Even though there is the possibility of being aggregated to other markets, usually ancillary service markets are built around a single service.

In the simpler version of these markets, a single buyer/multiple sellers system exists. The system operator (SO) reveals the level of the service that it intends to acquire, and interested parties submit their tenders. Afterwards, according to the needs in every market time unit, the SO procure the services required by choosing the best oers.

This choice is usually performed using a bidding ladder such as the one in gure 2.8. The two possible methods of payment are pay-as-bid and marginal cost. In pay-as-bid, each selected oer is paid according to their tender. In marginal cost, all the selected oers are paid the same price, with the highest bid being accepted determining this price.

Figure 2.7: Example of bidding ladder for The Netherlands, (August 15th, 2013, 10th hour), from [30].

(43)

16 Electricity networks and markets: background

2.3.4 Balancing markets

Balancing markets present dierent structures and regulations around the globe. This subsection introduces examples from Europe and Western USA as case studies in this eld.

2.3.4.1 Western USA

Since 2014, there is in place an innovative energy balancing market in the Pacic Coast of the United States of America (USA). It is called the energy imbalance market (EIM) and was created in cooperation with several system operators. It is managed by the California Independent System Operator (CAISO), which also manages other national markets. Currently, EIM is serving parts of the West Coast states of Oregon, Washington, California, Utah, Wyoming and Idaho, but this list could be expanded in the future, as several neighbouring system operators intend to join this market [31].

In this balancing market, supply and demand are balanced every 15 minutes, with the resources having the lowest costs being dispatched automatically every ve minutes [32].

Several aspects of this market are unique:

ˆ It runs in two dierent network subregions of the USA, which may result in some regulatory dierences.

ˆ The grid in this market spreads through quite a wide area and presents a high share of renewable energy generation.

ˆ The balancing authorities (a type of system operator) jointly aggregated their services so as to optimise the usage of the balancing resources under their supervision.

ˆ The balancing authorities can take advantage of the most cost-benecial resource available in the whole market and within the limits set by congestion [33].

2.3.4.2 Europe

In Europe, a big step towards harmonisation of the dierent national regulations is being taken with the drafting of the Network Code on Grid Balancing (NCEB) [34]. However, this document is still under revision as shown in [35]. The NCEB intends to establish common principles for the procurement and settlement of Frequency Containment Reserves (FCR) [referred as primary reserves in the past], Frequency Restoration Reserves (FRR) [referred as secondary reserves in the past] and Replacement Reserves (RR) [referred as tertiary reserves in the past] and common methodology for the activation of Frequency Restoration Reserves and Replacement Reserves [34].

(44)

2.3. Energy market structure 17 2.3.4.3 Germany

In Germany the settlement period for balancing is 15 minutes. A net balance is performed for the generation and consumption balances, allowing each balance responsible party (BRP) to net their generation and consumption variations. Even though the closing time of the initial schedule is at 2:30 p.m. of the previous day, the nal closure is 45 minutes before delivery and it can be extended to 4:00 p.m. on the day after delivery in case of intra-area exchanges.

Furthemore, each BRP can transfer the responsibility of the imbalance to another BRP. Therefore, in an intra-area situation, BRPs may trade between themselves and minimise their imbalances ex-post delivery time [36]. An exception to this procedure is the balancing responsibility for wind and solar generation, which is to the responsability of TSOs. This exception is granted by the German Renewable Energy Law, which precribes a socialisation of the country s renewable energy integration costs [36].

2.3.4.4 Iberian Peninsula

In the Iberian market there are two levels of balancing services with dierent time units. Secondary regulation, which is the rst level, has a ve minute timeslot and regulation reserves a timeslot of one hour. These services are contracted under the Iberian electricity market (MIBEL) [37, 38]. Each TSO manages its own reserve procurement and that only at second level there can be a reserve exchange between the TSOs. The cross border interconnections between Portugal and Spain, and between Spain and France are managed by a platform called BALIT. (It is possible that Italy and UK join this platform in the near future.) The future goal of MIBEL will be to proceed with market integration, especially at reserve sharing level.

For secondary regulation, every day at 1:00 pm, the TSO publishes the needs for the next day. Each BRP may submit their tenders for the next day between 6:00 p.m. and 6:45 p.m. The thermal generators scheduled for regulation should respect the minimum of 10% of their rated power, and for hydro power plants this value is of 30% of their rated power [38].

In what concerns regulation reserve, each BRP may submit their tenders for the services to be supplied the next day until 8:00 p.m [38].

2.3.4.5 Nordic

The Nordic region considered here is composed of Denmark, Finland, Sweden and Norway. The three last countries jointly with Eastern Denmark form the Nordel synchronous zone. This synchronous zone is separated from the former UCTE synchronous zone, which includes Western Denmark and most Continental Europe. Since 2002, a common regulating power market is in eect in the Nordic region [39]. This market has an hourly time step. Integrated in it is a balancing market where the cheapest regulating bids in the control bid ladder are activated as needed

(45)

18 Electricity networks and markets: background to balance the demand and supply in the entire region. The price for the settlement of all the selected bids is the one of the last activated bid. This approach is called marginal pricing. When no congestions exist between dierent areas, a single regional regulation price is used. However, when congestion occurs, the Nordic region is divided into two or more price zones, with dierent regulation prices [36].

Even though the market is shared, rules and regulations on the balancing service provision diered between the Nordic countries until 2009. That year, part of these rules were harmonised. The generalised rules encompass a minimum bid size of 10 MW, a lower price limit equal to the day-ahead price, usage of pay-as-bid for congestion management, and a gate closure time of 45 minutes before the hour of delivery [40]. Furthermore, this regulation entails that consumers are rewarded when compared with producers, as contrary to consumers, producers do not receive a benet for having an imbalance in the opposite direction to the system imbalance [36, 41].

2.3.5 Reserve capacity mechanisms

In the current European electricity markets, there are concerns if the market is able to provide a sucient and stable reserve capacity, assuming large volumes of intermittent generation [42]. Periods of scarcity and hence, higher prices, often leads to requests for price caps and regulatory intervention.

The introduction of additional instruments, such as capacity mechanisms could be a market solution to meet supply and demand, potentially allowing the integration of technologies such as energy storage and demand side exibility. Furthermore, the presence of generation capacity in excess of the capacity which is contracted by market parties (reserve capacity) provides an additional benet for all consumers [43].

With the liberalisation, the European markets adopted an energy-only market design in which generators are paid only for the energy delivered. With this solution, the spot price drives investment decisions. Ideally, the market requires a demand that can react quickly to price changes. Therefore, during periods of shortage and consequently high prices, consumers should theoretically reduce their consumption. The question is whether demand is responsive enough to prices and how this is going to happen in practice.

The introduction of a capacity mechanism may encourage consumers and other market stakeholders to be more responsive to markets'stimuli. Such mechanism requires both an appropriate level of capacity and incentives so that it is delivered. Therefore, several aspects have to be taken in consideration:

ˆ Generators and energy storage operators receive revenues for both the energy delivered and the capacity installed.

ˆ When the capacity margin is small or when there is a shortage, a small reduction in the supply of electricity may lead to a high price increase due

Referenties

GERELATEERDE DOCUMENTEN

As inward FATS is an assembled dataset that tracks the business performance of all registered foreign affiliates in the Netherlands, the impact of Chinese

ACM analyzes this reasonable return using the Weighted Average Cost of Capital (WACC). In this report, ACM calculates the WACC for the regulated electricity and drinking-water

On page 5 of the draft WACC method, it is stated that it is justified to determine the reference capital market based on internationally active investors that are interested

Before estimating our four variable vector autoregressive model, the time series of the Ardour Global Alternative Energy Index – Europe (AGIEM), the physical electricity index

Nestlé appears to solve its human rights violations in Thailand mainly through a deliberative approach and shows that liberal democratic values are relatively less

What becomes clear from the pre-acquisition phase is that the fit on strategic, cultural and structural level between the two companies was either missing or very low. This implies

This hypothesis claimed that different practices of Lean Management have a positive significant linear relationship with performance outcomes for companies operating

• One of the tasks of the engineering department is to help solve problems at suppliers and help them improve • The company tries to use capacity of the supplier as good as