markets for carbon and power

Markets for Carbon and Power Pricing inEurope

NEW HORIZONS IN ENVIRONMENTAL ECONOMICS

Series Editors: Wallace E. Oates, Professor of Economics, University of Maryland,College Park and University Fellow. Resources for the Future,USA and HenkFolmer, Professor of Research Methodology, Groningen University and Professor ofGeneral Economics, Wageningen University, The Netherlands

This important series is designed to make a significant contribution to thedevelopment of the principles and practices of environmental economics. Itincludes both theoretical and empirical work. International in scope, it addressesissues of current and future concern in both East and West and in developed anddeveloping countries.

The main purpose of the series is to create a forum for the publication of highquality work and to show how economic analysis can make a contribution tounderstanding and resolving the environmental problems confronting the worldin the twenty-first century.

Recent titles in the series include:The Impact of Climate Change on Regional Systems A Comprehensive Analysis of California Edited by Joel Smith and Robert Mendelsohn Explorations in Environmental and Natural Resource Economics Essays in Honor of Gardner M. Brown, Jr.Edited by Robert Halvorsen and David Layton Using Experimental Methods in Environmental and Resource Economics Edited by John A. List Economic Modelling of Climate Change and Energy Policies Carlos de Miguel, Xavier Labandeira and Baltasar Manzano The Economics of Global Environmental Change International Cooperation for Sustainability Edited by Mario Cogoy and Karl W. Steininger Redesigning Environmental Valuation Mixing Methods within Slated Preference Techniques Neil A. Powe Economic Valuation of River Systems Edited by Fred J. Hitzhusen Scarcity, Entitlements and the Economics of Water in Developing Countries P.B. Anand Technological Change and Environmental Policy A Study of Depletion in the Oil and Gas Industry Shunsuke Managi Environmental Governance and Decentralisation Edited by Albert Breton,Giorgio Brosio, Silvana Dalmazzone and Giovanna GarroneChoice Experiments Informing Environmental Policy A European Perspective Edited by Ekin Birol and Phoebe Koundouri Markets for Carbon and Power Pricing in Europe Theoretical Issues and Empirical Analyses Edited by Francesco Gull

Markets for Carbonand Power Pricing inEuropeTheoretical Issues and Empirical Analyses

Edited by

Francesco Gull

Universit Bocconi, Milan, Italy

NEW HORIZONS IN ENVIRONMENTAL ECONOMICS

Edward ElgarCheltenham, UK Northampton, MA, USA

Francesco Gull 2008 OECD/IEA 2008 for Chapter 4 by Julia Reinaud

All rights reserved. No part of this publication may be reproduced, stored in aretrieval system or transmitted in any form or by any means, electronic,mechanical or photocopying, recording, or otherwise without the priorpermission of the publisher.

Published byEdward Elgar Publishing LimitedThe Lypiatts15 Lansdown RoadCheltenhamGlos GL50 2JAUK

Edward Elgar Publishing, Inc.William Pratt House9 Dewey CourtNorthamptonMassachusetts 01060USA

A catalogue record for this bookis available from the British Library

Library of Congress Control Number: 2008927677

ISBN 978 1 84720 809 5 (cased)

Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents

List of contributors vii

1 Introduction 1Francesco Gull

PART I OVERVIEW AND THEORETICAL ISSUES

2 The European Emissions Trading Scheme: overview, lessonsand perspectives 13Liliya Chernyavska

3 Modelling the short-run impact of carbon trading on the electricity sector 36Francesco Gull

4 From electricity prices to electricity costs: impact of emissions trading on industrys electricity purchasing strategies 80Julia Reinaud

PART II EMPIRICAL ANALYSES

5 Options to address concerns regarding EU ETS-inducedincreases in power prices and generators prots: the case ofcarbon cost pass-through in Germany and the Netherlands 101Jos Sijm, Sebastiaan Hers and Bas Wetzelaer

6 A vector error correction model of the interactions among gas,electricity and carbon prices: an application to the cases ofGermany and the United Kingdom 145Derek W. Bunn and Carlo Fezzi

7 Impacts of the European Emissions Trading Scheme on Finnishwholesale electricity prices 160Juha Honkatukia, Ville Mlknen and Adriaan Perrels

8 The impact of the European Emissions Trading Scheme on power prices in Italy: the load duration curve approach 193Liliya Chernyavska and Francesco Gull

v

9 The joint impact of carbon emissions trading and tradablegreen certicates on the evolution of liberalized electricitymarkets: the Spanish case 213Pedro Linares and Francisco J. Santos

Index 235

vi Markets for carbon and power pricing in Europe

Contributors

Derek W. Bunn London Business School, UK.

Liliya Chernyavska IEFE (Instituto di Economia e Politica dellEnergia edellAmbiente), Bocconi University, Milan and DIEM (Departimento diEconomia e Metodi Quantitativi), Universit di Genova, Italy.

Carlo Fezzi CSERGE (Centre for Social and Economic Research on theGlobal Environment), School of Environmental Science, University ofEast Anglia, Norwich, UK.

Francesco Gull Bocconi University, Milan, Italy.

Sebastiaan Hers ECN, Energy Research Centre of the Netherlands.

Juha Honkatukia VATT, Government Institute for Economic Research,Finland.

Pedro Linares Instituto de Investigacin Tecnolgica, UniversidadPonticia Comillas, Spain.

Ville Mlknen VATT, Government Institute for Economic Research,Finland.

Adriaan Perrels VATT, Government Institute for Economic Research,Finland.

Julia Reinaud IEA (International Energy Agency), Paris, France.

Francisco J. Santos Instituto de Investigacin Tecnolgica, UniversidadPonticia Comillas, Spain.

Jos Sijm ECN, Energy Research Centre of the Netherlands.

Bas Wetzelaer ECN, Energy Research Centre of the Netherlands.

vii

1. IntroductionFrancesco Gull

The European Emissions Trading Scheme (EU ETS) is the rst interna-tional trading scheme for carbon dioxide (CO2) emissions in the world. Itsaim is to reduce the cost of meeting the Kyoto Protocol requirements bycreating an explicit and common price for carbon.

The EU ETS covers several industry sectors of which power generationis the largest. Therefore, on the one hand, the performance of the tradingscheme largely depends on the eicacy in inducing the power industry toreduce CO2 emissions signicantly in the short and long runs. On the otherhand, it might have a considerable impact on consumer surplus and rmsprots and competitiveness. Either the performance of the EU ETS or itsimpact on social welfare depends on how and to what extent the CO2 priceis passed through into power prices.

Given this premise, it is easy to understand why researchers, operatorsand regulators are so interested in the interaction between carbon andpower prices. Nevertheless, the current literature on this topic is quite con-troversial, from both the theoretical and the empirical points of view.

On the theoretical side, an open question is how the CO2 cost pass-through is correlated with the structural features of the output market,namely the power market. In fact, while it is quite clear what can occurunder perfectly competitive scenarios, opinions diverge when the assump-tion of perfect competition is abandoned. Some authors state that underimperfect competition electricity prices increase more than in a competitivescenario, while others argue in favour of the opposite result (electricityprices increase less under market power). This controversial frameworkemerges because each contribution is based on specic and dierenthypotheses on the strategic behaviour of the rms, the technology mix andthe regulation of both power and allowance markets. Consequently theresults of each study cannot be generalized as they oer only a partial viewof the problem.

The empirical literature is even more controversial. Some authors esti-mate full or almost full pass-through rates (PTRs). Others nd that powerprices do not internalize carbon costs or that there is limited evidence thatCO2 is factored into wholesale price. Indeed, most empirical analyses are

1

simply trying to check whether there is a full pass-through and genericallyattribute the deviation from this rule to various structural factors (amongwhich the exercise of market power in the output market) without attempt-ing to measure their specic eects.

In addition, these studies dier signicantly in their methodologicalapproaches. Some rely on forward markets whereas others analyse spotprices. Some are based on the econometric elaboration of time series,whereas others examine the change in the price and/or spread durationcurves (the load duration curve approach) and so on. Thus, no one empir-ical study, if considered separately, is useful to check the robustness of thetheoretical models and, once again, is able to provide results which can begeneralized.

The question of how the ETS can aect emissions is no less disputed. Allauthors agree that emissions are highest under the most-competitive sce-narios. Nevertheless some contributions show that emission reductions arelowest under perfect competition, while others state that generally higheremission reductions are achieved in the case without market power.

Moreover policy makers, regulators and operators declarations andposition papers do not help to clarify the overall framework, proving thatopinions are not convergent even outside the scientic community. Forinstance, in Germany throughout 2005, the energy-intensive industriesdebated with power rms the question of responsibility for the carbon pricepass-through into product prices. In other countries, such as Italy, thepower industry ensured that there was no CO2 cost pass-through in thewholesale power prices and, consequently, no windfall prot (mostly dueto the pass-through of the opportunity costs of freely allocated CO2 emis-sion allowances).

These positions, however, became very diicult to defend when thecarbon price collapsed at the end of April 2006. Since this fall was imme-diately followed by a drop in wholesale power prices in some Europeanmarkets, it appeared unequivocally clear that there was an interactionbetween carbon and power prices. In other words, the empirical evidenceconrmed what the theoretical analysis suggested, and thereafter thedebate moved from the question of whether power companies gain wind-fall prots or not to more interesting arguments which can be summed upby means of the following questions.

Why do power prices seem to be correlated with the carbon price in somemarkets and not in others? Why do power rms pass through into powerprices in some cases more and in others less than the carbon opportunitycost? What is the relationship between the pass-through and the structureof the power market? What is the inuence of the technology mix and ofthe available generation capacity? How can market power in the power

2 Markets for carbon and power pricing in Europe

market aect the pass-through and vice versa? Can carbon trading deter-mine a rise rather than a decline in carbon emissions of the power system,at least in the short and medium runs? Should eorts be focused on bring-ing the electricity price down? Should regulators change the method ofallocating the CO2 emission allowances in order to improve the eciencyand eectiveness of the ETS and eliminate the windfall prots? Shouldpolicy makers and regulators care about the redistribution of the cost ofthe EU ETS among industry and consumers?

It is dicult to answer these questions without addressing rst to whatextent and under which conditions the CO2 price is passed through intopower prices. This requires carrying out a robust simulation model, on thetheoretical side, and going beyond the results of a (country and method-ological) specic study, on the empirical side. In other words, it is necessaryto provide a comprehensive analysis that would be useful to understand howthe pass-through depends on market structural features, on the one hand,and how methodological approaches aect the empirical results, on theother.

This book attempts to set out such an analysis and to answer these ques-tions by bringing together and interpreting the contributions by the leadingexperts of each EU country or regional market. The volume is divided intotwo parts and has the following structure.

Part I includes an overview of the EU ETS, including its early results,and two chapters on the theoretical issues. These contributions are usefulto interpret the empirical analyses in accordance with the statements ofeconomic theory.

In particular, Chapter 2, by Chernyavska, is organized as follows. First,the author describes the EU directive on emissions trading, discussing theCO2 allowance market regulation (industrial sector covered, methods ofallocation, regulatory periods and so on.).Then the allocation plans and theearly results of the EU ETS are presented and discussed. Commentsabout the allowance price dynamic are proposed as well as considerationsabout the organization of the European CO2 markets. Finally, the authorexplores the possibility of learning some lessons for the future, mainly bylooking at the post-Kyoto and at the new EU targets (reducing emissionsof greenhouse gases by 20 per cent by 2020).

In Chapter 3, Gull examines the impact of trading of CO2 emissionallowances on electricity pricing in the short run. The author is mainlyinterested in exploring the role of electricity market structures. He uses asimple analytical model to verify whether (and under what conditions) theimpact of the ETS under market power could be lower (or higher) than thatunder perfect competition. Furthermore, the author examines howemissions trading impacts on carbon emissions in the short run, that is,

Introduction 3

whether the environmental regulation can determine an increase (ratherthan a decrease) in pollution.

In Chapter 4, Reinaud focuses on the situation of energy-intensive indus-tries going beyond the analysis of the spot electricity prices. She examinesthe way carbon price is passed through into end-user prices under dierentelectricity pricing mechanisms such as short- and long-term contracts, reg-ulated taris and so on.

Part II includes the empirical analyses of the main European markets.These studies are based on dierent methodological approaches and arecollected and ordered so as to oer a comprehensive and thorough view ofthe topic.

In particular, Chapter 5, by Sijm, Hers and Wetzelaer, explores a varietyof options to address EU ETS- induced increases in power prices and wind-fall prots, notably whether these options are eective and whether theyalso have other (adverse) eects. The study starts with the estimates of theaverage PTRs in Germany and the Netherlands. The authors nd thatempirical estimates of carbon cost PTRs on forward wholesale powermarkets varied from 0.4 to more than 1.0 in 200506.

Bunn and Fezzi (Chapter 6) address the economic impact of the EU ETSfor carbon on wholesale electricity and gas prices. Specically, the authorsanalyse the mutual relationship between electricity, gas and carbon pricesin the daily spot markets in the United Kingdom and Germany. Using astructural co-integrated VAR (Vector Autoregression) model, they showhow the prices of carbon and gas jointly inuence the equilibrium price ofelectricity. Furthermore, the analysis derives the dynamic pass-through ofcarbon into electricity price and the response of electricity and carbonprices to shocks in the gas price.

Honkatukia, Mlknen and Perrels (Chapter 7) examine the extent towhich the costs of the EU ETS are reected in the electricity prices in theFinnish power market. The study involves a simple theoretical illustrationand an empirical exercise of electricity price formation. The authors usetwo econometric methods (the vector error correction model and theautoregressive conditional hederoskedasticity model) and nd that about50100 per cent of a price change in the EU ETS is passed on to the FinnishNordPool spot price.

In Chapter 8, Chernyavska and Gull carry out an empirical analysis inthe Italian context (an emblematic case of an imperfectly competitivemarket), which can be split in four sub-markets (North, macro-South,macro-Sicily and Sardinia) with dierent structural features. The authorsuse the load duration curve approach, which allows them to investigatehow the pass-through depends on the degree of market power in powermarkets, other than on other structural features.


Finally, the contribution by Linares and Santos (Chapter 9) deals withthe interaction between carbon markets and the policies supporting renew-able energies. The authors look at how the EU ETS and a tradable greencerticate system for the promotion of renewable energies may aect theSpanish electricity sector in the long term. The analysis was carried outusing a generation-expansion model, which accounts for the possible oli-gopolistic behaviour of generation rms.

It is clearly beyond the scope of this introduction to enter into the detailsof each study, since the reader can nd a full description of the method-ologies and results in the corresponding chapters. The following provides asummary of the books ndings with regard to the theoretical and method-ological issues as well as the empirical results and their policy implications.

The theoretical analysis shows that under imperfect competition theextent to which the carbon cost is passed through into power prices dependson several structural factors of the power market, namely: (i) the degree ofmarket concentration; (ii) the technology mix; (iii) the available capacity;and (iv) the allowance prices. Depending on these factors, the marginalpass-through rate (MPTR) can be either above or below 1, that is, theincrease in price under market power can be either higher or lower than thatunder perfect competition. Also, under certain conditions, the impact onpower prices can even be slightly negative (at least, in principle), that is, theETS can involve a decrease rather than an increase in prices. Market power,therefore, determines a signicant deviation from the full pass-throughrule but we cannot know the sign of this deviation, a priori, that is, withoutrst taking into account the structural features of the power market.

Another interesting result emerges from the theoretical analysis. TheETS causes a change in the degree of market power in the sense that, afterthe introduction of the trading scheme, the time (the number of hours inthe year) over which power rms prefer to bid the maximum price (forexample, the price cap) increases or decreases depending (again) on thestructural factors of the power market. This means not only that the ETScan amplify or lessen the existing distortions in the output market but alsothat it might determine a rise rather than a decrease in carbon emissions,namely when the change in market power signicantly expands the pro-duction share of the most-polluting plants. However, this does not neces-sarily imply that perfect competition is preferable to market power from theenvironmental point of view, but simply that under imperfect competition,and provided that certain conditions are satised, it might be more dicultto achieve the environmental targets.

With regard to the empirical results, Table 1.1 presents a summary ofthe estimated PTRs on the power markets analysed in this book. As canbe noted, these values are signicant in all cases and suggest that the

Introduction 5

6Tab

le 1

.1S

umm

ary

ofth

e ca

rbon

cos

t pa

ss-t

hrou

gh r

ates

Met

hodo

logy

Pri

ceP

erio

dA

vera

geP

eak

O

-pea

k

Mid

-mer

itV

ery

o-p

eak

Fin

land

(C

hapt

er 7

)E

cono

met

ric

VE

CM

W

hole

sale

spo

t20

050

60.

51.

0an

d A

R-G

AR

CH

Ger

man

y (C

hapt

er 6

)E

cono

met

ric

VE

CM

Who

lesa

le s

pot

2005

06

0.52

Ger

man

y (C

hapt

er 5

)E

cono

met

ric

OL

SW

hole

sale

forw

ard

2005

0.60

0.41

2006

0.57

0.64

Ital

yN

orth

(C

hapt

er 8

)L

oad

dura

tion

cur

ve

Who

lesa

le s

pot

2006

1.5

2.1

0.8

1.3

0.9

1.1

appr

oach

Ital

ySo

uth

(Cha

pter

8)

Loa

d du

rati

on c

urve

Who

lesa

le s

pot

2006

0.0

0.5

1.7

2.1

0.9

1.1

appr

oach

Ital

yw

hole

(C

hapt

er 8

)L

oad

dura

tion

cur

ve

Who

lesa

le s

pot

2006

1.1

1.5

1.2

1.5

0.9

1.1

appr

oach

Net

herl

ands

(C

hapt

er 5

)E

cono

met

ric

OL

SW

hole

sale

forw

ard

2005

1.34

0.40

2006

1.10

0.38

UK

(C

hapt

er 6

)E

cono

met

ric

Who

lesa

le s

pot

2005

06

0.30

VE

CM

Not

e:V

EC

M

vect

or e

rror

cor

rect

ion

mod

el;A

R-G

AR

CH

au

tore

gres

sive

con

diti

onal

het

eros

keda

stic

ity

mod

el;O

LS

= o

rdin

ary

leas

tsq

uare

s.

bandwidth for pass-through can be quite large, varying between countriesand periods. The estimates range from 30 to 100 per cent if we refer to theaverage value, from 57 to 134 per cent in the peak hours and from 40 to 123per cent in the o-peak hours. Furthermore, the analysis of how the PTRis distributed over time (between peak and o-peak periods) would seem tosuggest that there is not the same behaviour everywhere (that is, no generalrule). The PTR is higher in the peak than in the o-peak hours in Germany(2005), the Netherlands and ItalySouth, and vice versa in Germany (2006)and ItalyNorth. The overall picture, therefore, would seem to support theconclusions of the theoretical analysis, that is, we cannot know how theETS impacts on power prices (whether the PTR is low or high, more or lessthan 1) without rst accounting for the structural features of the powermarkets under analysis.

Furthermore, the estimates have to be interpreted with due care as, tosome extent, they depend on (shortcomings of) the data and the method-ologies used or the assumptions made. With regard to the methodologicalissues, the contributors use many empirical methods to estimate carbonprice PTRs. Each method has its own strengths and weaknesses, so that oneapproach cannot be considered as denitively preferable to another.Nevertheless, there are some important dierences between the economet-ric and non-econometric techniques.

The econometric approach uses sophisticated statistical tools in order tomeasure carbon price PTRs. It is based on the statistical elaboration oftime-series of either forward or spot prices (of both electricity and carbon)and estimates the impact of the ETS on the average price, eventually dis-tinguishing between the peak and o-peak hours. The specications ofthese models are generally quite simple. The set of drivers commonlyincludes the fuel costs and the temperature (Sijm et al., Chapter 5; Bunnand Fezzi, Chapter 6). Only one model (Honkatukia et al., Chapter 7) usesadditional variables (namely, the production capacity and the utilization ofthe transmission capacity). Furthermore, models neither consider the realmarginal technology hour by hour nor are suited to capture (by usingappropriate drivers) the eect of market power. They assume that duringthe observation period power prices are set by a single (marginal) technol-ogy with a xed, generic fuel eciency. In other words, the econometricmodels are very useful for providing a precise (statistically signicant) valueof the carbon pass-through but they are not able to justify this value, thatis to explain why a PTR is high, low or zero.

The non-econometric approach consists of two steps: (i) estimate thePTR by means of an analytical model; and (ii) compare the estimateswith the observed data. It does not elaborate time series but carry out thepass-through distribution over time (the pass-through curve) starting from

Introduction 7

the load duration curve. This allows us to obtain the PTR hour by hour.The non-econometric approach has several advantages compared with theeconometric one. First, as pointed out before, it provides a detailed analy-sis of the pass-through over time, on an hour-by-hour basis. As a result, itseems to be well suited to describe the impact of market power whoseextent depends on the level of power demand (and hence on the time ofconsumption). Moreover, by using this approach, market power can beeectively simulated by means of a theoretical model assuming oligopolis-tic competition or (as in Chapter 3) a dominant rm framework. Second,the non-econometric approach allows us to take into account other impor-tant structural factors which cannot be included in econometric models, forexample, the technological mix and the available capacity in the market.However, unlike the statistical approach, it does not provide a precise valueof the pass-through but only a range of its variability. In this sense, the twoapproaches are complementary. The non-econometric one is useful toimprove the specication of the statistical models and to interpret theirresults.

Finally, concerning the policy implications, this work can help to shedlight on two issues: (i) the eectiveness and eciency of the ETS itself(regardless of its specic design); and (ii) how the current design of the EUETS should be changed in order to improve its performance.

With regard to the ETS itself, the theoretical and empirical analyseshighlight that under imperfect competition power prices may increase lessthan under perfectly competitive scenarios. Thus, provided that certainconditions are satised, imperfect competition can partly lessen the ETSeectiveness in terms of emission abatement. Also, under certain condi-tions, the ETS can determine a rise rather than a decrease in carbon emis-sions, at least in principle. At the same time, the change in emissionsbecomes high only when the carbon price is above the switching price, thatis, when the carbon cost internalization determines a switch in the meritorder of power plants, for example between coal and gas-red plants. Thus,in designing the ETS, regulators should mainly ensure that the allocationwill create an adequate scarcity of tradable permits. Furthermore, policymakers should also take into account the interaction of the scheme withother energy and environmental policies. For instance, a combination ofboth carbon reduction and renewable promotion policies may result inlower costs than carbon reduction policies alone, with the additionalbenet of increased deployment of the renewable technologies.

Looking at the design of the scheme, we have to be aware that theincrease in power prices due to the pass-through of the (opportunity) costsof (freely allocated) CO2 emission allowances is a rational and intendedeect from an ecient carbon abatement policy perspective. Nevertheless,


the supposed ETS-induced increases in power prices and generators prots(windfall) have raised several concerns about its impact on the internationalcompetitiveness of some power-intensive industries, the purchasing powerof electricity end-users such as small households or, more generally, the dis-tribution of social welfare among power producers and consumers. As aresult, in several countries, a variety of options have been suggested inorder to address these concerns, such as changing the ETS allocationsystem, taxing windfall prots or controlling the market prices of EUcarbon allowances, electricity or both. Some contributions of this book,and especially Chapter 5, examine these options in order to address theirperformance and suggest the best way of improving the design of thescheme.

Overall, this volume attempts to oer a comprehensive analysis of therelationship between carbon and power markets. As such we hope it willprovide a useful contribution to the debate on the perspective of the EUETS and to the literature on the interaction between environmental policyand the structure of the environmentally regulated markets.

Introduction 9

PART I

Overview and theoretical issues

2. The European Emissions TradingScheme: overview, lessons andperspectivesLiliya Chernyavska

2.1 INTRODUCTION

The increasing risk that radical/dramatic events caused by climate changecould occur, having a negative impact on all aspects of human activity,raises the question concerning reduction of greenhouse gas (GHG)emissions.

Now, in fact, most of the scientic community admits that there is a rela-tionship between the increase in concentration of GHGs in the atmosphereand the rise in the average temperature of the planet, even though we arestill far from explaining the exact mechanism of such a relationship.However, it is clear that the anthropogenic emissions have a strong negativerole in the persisting increase in the concentration of GHGs and thus inclimate change.

The Kyoto Protocol is the rst step towards the international commit-ment to mitigate the process of climate alteration. This agreement, signedin 1997, sets the overall amount of the reduction (5.2 per cent, with respectto the level of emissions registered in 1990) to be achieved by the parties.This commitment is shared out in dierent proportions among groups ofcountries and among countries within each group on the basis of their eco-nomic development and contribution to global emissions. For example, thetarget for the European Union (EU) was xed at 8 per cent. The reduc-tions are to be achieved in ve years, namely from 2008 to 2012.

The Protocol also provides the rules and mechanisms (the so-calledKyoto exible mechanisms) to reach this goal most cost-eectively andequitably without precluding the economic development of all the parties.

In December 2002, after lengthy discussions, the EU opted for a cap-and-trade scheme as one of the main instruments to ensure the success inattaining the Kyoto target.1 Such a decision had both an economic and apolitical basis. First, the European Commission (hereafter Commission)

13

was looking for an ecient tool to reduce the costs of compliance that wasalso able to capture the structural dierences among countries and sectors,creating as few distortions as possible. Second, on the political side, it wasnecessary to guarantee a univocal decision, taking into account the inter-ests of each member state. Finally, at the international level, the EU aimsto have a leadership position in the ght against global warming.

As a result, in October 2003 the European Council and Parliamentapproved Directive 2003/87/CE establishing a scheme for GHG emissionallowances within the EU.

2.2 EMISSIONS TRADING DIRECTIVE 2003/87/CE

Directive 2003/87/CE (European Commission, 2003a) establishes a cap-and-trade scheme for CO2 emission allowances in order to promote thereduction of carbon emissions in cost-eective and economically ecientways. It has a pan-European coverage, including all new member states.This latter provision is necessary to guarantee lower costs of compliance,reecting relatively low abatement costs in the Eastern European countries.

The European Emissions Trading Scheme (EU ETS) took eect on1 January 2005. It foresees mandatory participation for the sectors men-tioned in Annex I of the directive, and is planned to take place in two phases:Phase I, from 2005 to 2007, is a pilot period, necessary to try out the designof the trading framework and to provide the opportunity to accumulate nec-essary experience for the second phase; Phase II starts in 2008 and coincideswith the Kyoto commitment period (200812). According to the directive, thetradable allowances are allocated directly to industrial emitters, which canexchange/trade them freely. The volumes of permits are xed and determinedbefore the starting date of trading. This ensures that, at least in principle,major predictability and thus stability of the system should be guaranteed.The participants are obliged to surrender the amount of allowances equal totheir annual emissions after each calendar year. The eventual decit has tobe covered by purchasing the permits. If the operator does not surrender theemitted quantities, penalties are applied. Legislation stipulates two dierentpenalties for the rst and second periods: 40 /ton and 100 /ton, respectively.Moreover, payment of the penalty does not release the operator from theobligation to surrender the allowances. Such a framework further underlinesthe mandatory character of the participation and, from the theoretical pointof view, it could lead to a situation in which the carbon price could exceedthe penalties. We could also suppose that the increase in the sanction fromone period to the next reects in some way the expectations of the highercosts of abatement during the second period (Kyoto commitment).

14 Overview and theoretical issues

Table 2.1 summarizes the key features of the EU ETS. With regard to sec-toral coverage, participants in the scheme include energy activities (elec-tricity generators, oil reneries and coke ovens) and some manufacturingsectors (such as production and processing of ferrous metals, cementclinker, glass and ceramics, and pulp and paper). The scheme includes morethan 10,500 installations accounting for about 40 per cent of total GHGemissions of the EU. The directive also foresees the possibility for futureexpansion of the activities. The intention was to include the chemical andtransportation sectors, but the problems with monitoring as well as hightransaction costs meant that they had to be excluded from the actualcoverage. For the same reason, only CO2 was chosen to start with. Othergases could be included later on.

One of the main critical points of the entire system is the initial alloca-tion of allowances, as this may generate distortions among countries,sectors and single installations.2 The directive states that most of theallowances (95 per cent in Phase I and 90 per cent in Phase II) have to beallocated free of charge. This provision creates two problems: first, it maylessen rms incentives to reduce carbon emissions; and second, since rmsmay pass through the carbon opportunity cost into product prices, freeallocation may determine signicant windfall prots.

Furthermore, the directive does not indicate the technical mechanism forthe allocation (grandfathering, benchmarking, future emissions and so on).Almost all countries opted for grandfathering which, however, has certainweaknesses. First, it is necessary to evaluate the choice of the base periodin order not to penalize those industries/sectors that could suer a drop inproduction due to conjectural, atmospheric or other conditions and inorder to reward possible early actions. Second, grandfathering tends to bemore advantageous to the large emitters in comparison with less-pollutingrms, undermining the polluters pay rule. Thus the exploitation of cleantechnologies may not receive adequate incentives.

As a result of all this, the eciency and eectiveness of the tradingscheme could be undermined, as it could not ensure the abatement of emis-sions according to the least-cost principle. Moreover, we should take intoconsideration that the directive did not stipulate the exact amount ofreduction to be achieved by each member state, leaving this decision to thenational governments. This involved a real risk that further distortionscould derive from the dierent choices of the dierent countries in termsof relative allocation, that is, the ratio between the amount of allowanceallocated and the expected emissions at country, sector and participantlevels.

Overall, the text of the directive seems to be a compromise between theurgency to reach the Kyoto target and the interests of the dierent

The European Emissions Trading Scheme 15


Table 2.1 Framework of the EU ETS

Feature Description

Type of system Cap-and-trade system based on direct emissionsTiming Phase I: 200507, preliminary or try-out period

Phase II: 200812, coincides with Kyoto commitmentGeographical EU25, with possibility of expanding to new member states

coverageCoverage of Energy activities: all combustion installations with more thenactivities 20 MW thermal input, including power sector; oil reneries;

coke ovensProduction and processing of ferrous metalsMineral industry: cement clinker, glass and ceramicsOther activities: pulp and paper from timberOpt-out provisions: member states may apply to theCommission for installations to be temporarily excludeduntil 31 December 2007

Opt-in provisions: member states may voluntarily extend thescheme to other installations, starting from Phase II

Coverage of Only CO2 in Phase IGHG Other gases may be included in Phase II, provided that

adequate monitoring and reporting systems are available,and there is no damage to the environmental integrity of theETS or distortions to competition

Size of the Nearly 10,500 installationsmarket1 About 41% of total EU GHG emissions

Allocation Free allocation within national allocation plans (NAPs) basedon Annex III criteria and Commission Guidelines

Member states can auction up to 5% in Phase I and up to 10% in Phase II

NAPs have to be presented to the Commission to be evaluatedand approved. The Commission may demand any change inorder to make NAP comply with criteria of Annex III

Operational On 30 April of each year, participants have to surrender arules quantity of allowances equal to their emissions in the

preceding calendar yearParticipants are allowed to trade the permits amongthemselves

Participants are allowed to form an emissions pool bynominating a trustee who takes on the responsibility forsurrender and trading allowances on behalf of all membersof the pool

Banking Implicit banking across years within each compliance periodMember states can determine their own banking from Phase I(200507) to Phase II (200812)

member states. Such compromise probably would not be possible if thedocument were more precise in indicating the methods and rules of allo-cation, caps to set and so on. At the same time, a broad outline wouldinevitably lead to a multiplicity of interpretations and thus to a lack ofhomogeneity.

As a result, the Phase I NAPs of each country diered in almost everyrespect.


Table 2.1 (continued)

Feature Description

Penalties Non-complying participants must pay a penalty of 40 /ton of CO2 during Phase I and 100 /ton in Phase II

Payment of the emission penalties does not release theoperator from the obligation to surrender the allowances

Use/links with Participants may convert emission credits from jointother Kyoto implementation (JI) and clean development mechanismmechanisms2 (CDM) projects into EU allowances in order to full their

obligations under EU ETSAll types of credits are allowed for conversion, except thosefrom nuclear facilities and carbon sink enhancement projects

The amount of credits allowed for conversion must beindicated in NAPs for each installation (% of totalallowances allocated to the installation)

The main eorts, however, have to come from domesticactions

Links with Agreement with third parties may provide mutual recognitionother schemes of allowances between the EU ETS and other schemes

Monitoring and Common monitoring, verication and reporting obligationsreporting were elaborated and delivered3

Verication Verication through third-party or government authorityRegisters Linked and harmonized national registries with independent

transaction log can track all the allowances

Notes:1. On the basis of 2005 data provided by the European Commission (2008).2. These Provisions are mentioned in the so-called Linking Directive (European

Commission, 2004a).3. For further details, see European Commission (2004b, 2007).

Sources: European Commission (2003a, 2004a, 2004b, 2007, 2008a); Sijm (2004).

2.3 PHASE I: FIRST PROBLEMS

The NAP is an instrument to accomplish Directive 2003/87/CE, which xesthe quantity of allowances that member states allocate to installationscovered by the ETS (for each phase). This amount has to be determinedbefore the beginning of the trading period and cannot be changed once ithas been approved by the Commission. The rst draft of the NAP had tobe presented by 31 March 2004 for EU15 countries and by 1 May 2004 fornewcomers. Once received, the Commission had to evaluate it within threemonths.

Each NAP is evaluated against a set of 11 criteria listed in Annex III ofthe directive. In order to assist the member states in the preparation of theNAPs, in January 2004, the Commission published guidelines for the appli-cation of the criteria, describing their scope and meaning, and highlightingwhether they are compulsory or optional (European Commission, 2004c).These guidelines together with the Commissions The EU EmissionsTrading Scheme: How to Develop a National Allocation Plan (2003b)established the base for the NAPs. Particular attention was paid tothe problems of allocation and the treatment of new entrants as these arethe major critical points that could undermine the competitiveness of theEuropean industry.

As specied earlier, the deadlines for the presentation of the NAPs were31 March and 1 May 2004, for the core European countries and for newentrants, respectively. However only half of the countries had presented therequired document by the deadline. The process of evaluation was con-cluded in June 2005, six months after the beginning of trading. Thisincreased uncertainty during the rst year of Phase I, penalizing partici-pants with late allocation plans.

With regard to the evaluation process, the most frequent violationsnotied by the Commission were overallocation, possible introduction ofex post adjustments, discrimination/unequal treatment of new entrants andan incomplete list of installations.

Thirteen countries were required to reduce the amount of their allocatedallowances (up to a total of 292,04 Mt of CO2). This intervention can beinterpreted as a signal of the tendency for some countries to overallocatetheir allowance.

The opposition to ex post adjustments derived from several sources: theneed to create a market based on simple and clear rules, at the same timereducing the uncertainty regarding the quantities that could be traded andthe intra-sector distortions; ensuring that the EU ETS would have asignicant impact on product prices (mainly on power prices) and conse-quently on emissions by means of the impact on (decrease in) output


demand. The ex post adjustment (updating), in fact, determines a reductionin the carbon opportunity cost (see Sijm et al., this volume, Chapter 5).

In line with earlier predictions, the NAPs were highly disparate. Theywere also too complex, especially those sections describing the allocationrules, and showed lack of transparency. Thus it is quite dicult to comparethem. Moreover, the allocation methodologies were extremely complex andsomewhat confusing. Most of the member states opted for grandfatheringbased on historical emissions. Only a few used the possibility to auction apart of the permits.

One of the most controversial and complex problems was the allocationto new entrants. About a hundred dierent rules and benchmarks were pro-posed, resulting in confusion and possible distortions. The need for aunique methodology that is able to provide incentives for the adoption ofthe cleanest and most ecient technologies became a clear issue for thefuture.

The analysis of the content of the allocation plans highlights anotherquestion to be answered as soon as possible. The denition of the combus-tion installation was generic and suitable for multiple interpretations, intro-ducing relativity into the coverage of the ETS. Roughly speaking there werethree main groups of interpretations: narrow (only the energy productionsector), medium (production of electricity, heat or steam for the purpose ofenergy production) and broad (production of electricity, heat or steam).Not one member state adopted the narrow interpretation. Austria, Belgium,Cyprus, Denmark, Hungary, Ireland, Latvia, Malta, the Netherlands,Portugal, Slovenia and Sweden chose the broad interpretation, while the restopted for the medium one. As a consequence of the adoption of the mediumdenition, some installations with high emissions remained outside the ETS.Some sectors, such as production and processing of ferrous metals, ceramicsand so on, had, therefore, dierent coverage in the dierent countries withpossible signicant consequences in terms of distortions of competition (forfurther details, see Ecofys, 2006).

As pointed out earlier, the EU ETS is the largest international scheme; itincludes more than 10,000 facilities of which about 16 per cent are instal-lations that together hold less then 0.2 per cent of allocated allowances(Ecofys, 2007). They are small and medium installations with a ratedthermal input exceeding 20 MW and a very low emission rate (less then5,000 tCO2/year). These have a marginal importance in terms of emissions,but they increase the administrative and transaction costs of the scheme.

The results of the rst years of the EU ETS are summarized in Table 2.2.Note that only six countries closed the rst year with a decit of allowanceswith respect to the average annual allocation. The UK is the leader with ashortfall of 33 Mt, followed by Spain with 19.5 Mt and Italy with 18.4 Mt.


20

Tab

le 2

.2O

verv

iew

of

allo

cate

d an

d su

rren

dere

d/ve

rifie

d al

low

ance

s,20

052

007

Mem

ber

stat

eT

otal

em

issi

onK

yoto

tar

get

(t)

Tot

al e

mis

sion

sE

TS

emis

sion

s E

TS

emis

sion

sC

O2

1990

(t)

CO

220

05 (

t)C

O2

2005

(t)

CO

220

06 (

t)

Aus

tria

78,9

00,0

0068

,700

,000

93,3

00,0

0033

,372

,841

32,3

82,8

19B

elgi

um14

6,80

0,00

013

5,90

0,00

014

3,80

0,00

055

,354

,096

54,7

84,4

62C

ypru

s6,

000,

000

n.t.

9,90

0,00

05,

078,

877

5,25

9,27

3C

zech

Rep

ublic

196,

300,

000

180,

600,

000

145,

600,

000

82,4

53,7

2783

,625

,869

Den

mar

k69

,300

,000

54,8

00,0

0063

,900

,000

26,0

90,9

1034

,584

,396

Est

onia

43,5

00,0

0039

,600

,000

20,7

00,0

0012

,621

,824

12,1

09,2

81F

inla

nd71

,100

,000

71,1

00,0

0069

,300

,000

33,0

72,6

3844

,648

,475

Fra

nce

564,

000,

000

563,

900,

000

553,

400,

000

126,

430,

451

128,

132,

861

Ger

man

y1,

231,

500,

000

973,

700,

000

1,00

1,50

0,00

047

3,71

5,87

247

8,44

8,31

4G

reec

e11

1,70

0,00

013

8,80

0,00

013

9,20

0,00

071

,033

,294

70,1

99,6

09H

unga

ry12

2,20

0,00

011

4,90

0,00

083

,100

,000

25,7

14,5

7426

,159

,149

Irel

and

55,8

00,0

0063

,000

,000

69,9

00,0

0022

,397

,678

21,7

46,1

17It

aly

519,

500,

000

485,

700,

000

582,

200,

000

225,

875,

412

227,

074,

462

Lat

via

25,3

00,0

0023

,800

,000

10,9

00,0

002,

854,

424

2,94

0,75

3L

ithu

ania

48,0

00,0

0044

,300

,000

22,6

00,0

006,

603,

869

6,51

6,91

1L

uxem

burg

12,7

00,0

009,

100,

000

12,7

00,0

002,

603,

349

2,71

2,97

2M

alta

2,20

0,00

0n.

t.3,

400,

000

1,97

1,25

81,

985,

765

Net

herl

ands

214,

600,

000

201,

700,

000

212,

100,

000

80,3

51,2

9276

,701

,187

Pola

nd58

6,90

0,00

055

1,70

0,00

039

9,00

0,00

020

2,58

8,89

120

8,31

4,95

0Po

rtug

al60

,900

,000

77,4

00,0

0085

,500

,000

36,4

13,0

0433

,096

,808

Slov

akia

73,4

00,0

0067

,500

,000

48,7

00,0

0025

,237

,739

25,5

37,2

73Sl

oven

ia20

,200

,000

18,6

00,0

0020

,300

,000

8,72

0,55

08,

842,

182

Spai

n28

9,40

0,00

033

2,80

0,00

044

0,60

0,00

018

1,57

4,30

218

0,64

9,54

0Sw

eden

72,3

00,0

0075

,200

,000

67,0

00,0

0019

,306

,761

19,9

55,6

32U

nite

d K

ingd

om77

9,90

0,00

068

2,40

0,00

065

7,40

0,00

024

2,39

6,03

925

1,21

5,42

1T

otal

5,40

2,40

0,00

04,

975,

200,

000

4,95

6,00

0,00

02,

003,

833,

672

2,03

7,62

4,48

1

21

Mem

ber

stat

eTo

tal E

TS

Di

eren

ce in

ET

SN

umbe

r of

Ave

rage

yea

rsA

vera

ge a

lloca

tion

Ave

rage

ann

ual

emis

sion

sem

issi

ons

2006

inst

alla

tion

allo

cati

onvs

.ET

S em

issi

ons

allo

cati

on a

tC

O2

2005

06

(t)

vs.2

005

(t)

cove

red

2005

07

(t)

CO

220

05 (

t)re

serv

e (t

)

Aus

tria

65,7

55,6

60

3.0

203

32,6

74,9

05

697,

936

330,

050

Bel

gium

110,

138,

558

1.

032

759

,853

,575

4,49

9,47

92,

545,

876

Cyp

rus

10,3

38,1

503.

613

n.a.

n.a.

n.a.

Cze

ch R

epub

lic16

6,07

9,59

61.

440

596

,907

,832

14,4

54,1

0534

8,02

0D

enm

ark

60,6

75,3

0632

.639

031

,039

,618

4,94

8,70

82,

460,

382

Est

onia

24,7

31,1

05

4.1

5018

,763

,471

6,14

1,64

718

9,52

9F

inla

nd77

,721

,113

35.0

602

44,5

87,0

3211

,514

,394

862,

952

Fra

nce

254,

563,

312

1.3

1,13

115

0,50

0,68

524

,070

,234

4,87

1,31

7G

erm

any

952,

164,

186

1.0

1,87

549

5,07

3,57

421

,357

,702

3,92

6,42

6G

reec

e14

1,23

2,90

3

1.2

168

71,1

35,0

3410

1,74

03,

286,

839

Hun

gary

51,8

73,7

231.

724

130

,236

,166

4,52

1,59

21,

424,

738

Irel

and

44,1

43,7

95

2.9

117

19,2

38,1

90

3,15

9,48

83,

081,

180

Ital

y45

2,94

9,87

40.

51,

002

207,

518,

860

18

,356

,552

15,5

51,5

75L

atvi

a5,

795,

177

3.0

103

4,05

4,43

11,

200,

007

505,

760

Lit

huan

ia13

,120

,780

1.

310

111

,468

,181

4,86

4,31

279

7,21

3L

uxem

burg

5,31

6,32

14.

215

3,35

6,66

775

3,31

838

7,00

6M

alta

3,95

7,02

30.

72

2,94

2,31

497

1,05

62,

288,

466

Net

herl

ands

157,

052,

479

4.

521

286

,439

,031

6,08

7,73

92,

503,

805

Pola

nd41

0,90

3,84

12.

884

323

9,10

0,00

036

,511

,109

2,47

2,40

5Po

rtug

al69

,509

,812

9.

125

436

,898

,516

485,

512

1,26

2,89

8Sl

ovak

ia50

,775

,012

1.2

178

30,3

64,8

485,

127,

109

7,18

0Sl

oven

ia17

,562

,732

1.4

988,

691,

990

28

,560

66,6

67Sp

ain

362,

223,

842

0.

51,

056

162,

111,

391

19

,462

,911

13,1

62,1

30Sw

eden

39,2

62,3

933.

475

322

,530

,831

3,22

4,07

067

8,14

9U

nite

d K

ingd

om49

3,61

1,46

03.

679

320

9,38

7,85

4

33,0

08,1

8515

,527

,484

Tota

l4,

041,

458,

153

1.7

10,8

342,

074,

874,

995

76,1

20,2

0078

,537

,547

All other member states have a net surplus of permits resulting in more then76 Mt of overallocation. There was no substantial change in the surren-dered emissions from 2005 to 2006.

Following this brief description of the main problems raised during thepresentation and evaluation of NAPs, we shall now discus price dynamicsand organization of the European market for tradable permits. Someimportant marketplaces (such as ECX: European Climate Exchange; EEX:European Energy Exchange; EXAA: Energy Exchange Austria; NordPooland Powernext) were created during the rst year of the ETS. Most of thesealready specialized in energy trading, while others (for example, ECX) hada particular focus on environmental goods. The dierence in the price ofpermits is negligible as the number of traders is very large and the good isalways 1 tonne of CO2. The price of carbon has been very volatile and itsdynamic reects all the uncertainties and rumours surrounding this market.

Figure 2.1, which shows the carbon price dynamic, highlights the factthat the market was heavily inuenced by the process of revision andapproval of the NAPs. The ocial statements (and rumours) from theCommission constituted one of the main price drivers. The interpretationof the NAPs (whether involving stringent allocation or not) caused themarket to uctuate, until the results of the rst year were announced. Later,when clear overallocations were conrmed, the price collapsed and reachedvalues close to zero by the middle of 2007.


0

5

10

15

20

25

30

35

01/01/05 01/04/05 01/07/05 01/10/05 01/01/06 01/04/06 01/07/06 01/10/06 01/01/07 01/04/07 01/07/07

/ton

CO2

Source: EEX (2008).

Figure 2.1 Evolution of CO2 prices, 20052007

By the end of the rst year of the emissions trading it was clear that thedirective had too many gaps and there was need for revision. The main openquestions could be divided into three groups: (i) key elements of the EUETS; (ii) expansion of the coverage of the system (unilateral inclusion ofsectors and gases); and (iii) the process of preparation, presentation andevaluation of NAPs.

With regard to the rst group, it is necessary:

to choose a clear and univocal denition of combustion installationin order to have uniform coverage of the scheme;

to identify clearly the entity/dimension (in terms of both power andemissions) of the installation covered by the system and to allowsmall and medium installations to opt out;

to dene specic reduction caps at national and sectoral levels(bearing in mind the eorts to assign reduction targets to the non-trading sectors); and

to indicate the method of allocation, dening allocation mechanism,applicable sectors, technological benchmarks, treatment of closuresand new entrants as well as combined heat and power plant (CHP)provisions.

The unilateral inclusion would be feasible only if the appropriate measuresregarding monitoring and verications could be implemented.

The rst experience of the NAPs was not totally positive. However, prob-lems with delays as well as lack of transparency in the evaluation processshould be solved in Phase II.

The reviewing process, undertaken by the European Commission in2005, provided an opportunity not only to discuss the problems but also toset up some feasible solutions for the short (Phase II) and medium (from2013 and beyond) terms. Such timing was unavoidable because for majorchanges it was necessary to modify the existing directive, a task that wasnot feasible in the short term. However, eorts to improve the functioningof the EU ETS within Phase II had been made.

Moreover, it is necessary to underline the importance of Phase I. Despiteall the problems, the rst international emissions trading scheme had beenstarted. Unique information had been gathered and experience had beengained, contributing to a better knowledge of the sectors covered by thescheme and an understanding of the functioning of important instrumentsof environmental policy. The price of atmospheric pollution had nallybeen set.


2.4 PHASE II

Before discussing the changes made in Phase II, and the results of the pre-sentation of the second tranche of NAPs, we shall recall the new provisionsconcerning the 200812 period that had already been included in Directive2003/87/CE.

In October 2004, the so-called Linking Directive 2004/101/CE was dis-cussed and approved. It was necessary to create the link between the EUETS and the exible mechanisms of the Kyoto Protocol such as JI andCDM. The proposed mechanism allows the use of credits from JI andCDM for compliance. In the current version of the directive there are norestrictions on the deployment of these credits at national and Europeanlevels. The quantities to be used should be indicated in the NAPs at thesingle installation level.

Moreover, the NAPs had to be presented well before the start of the newperiod of trading, namely 18 months before (in June 2006). And, as men-tioned earlier, the Commission had to start the process of the revision ofDirective 2003/87/CE by changing some of the key elements in order toincrease its eciency in reaching the new and more pressing target by 2020.

Phase II coincides with the Kyoto commitment (200812), thus memberstates should increase their eorts to reach the targets xed for both the EU(8 per cent) by the Kyoto Protocol and for individual countries by theBurden Sharing Agreement. Temporary exclusion of an installation is nolonger permitted, but the member states could include new activities/sectorsand new gases, thus expanding the coverage of the scheme. Finally, as men-tioned earlier, penalties were increased from 40 /t CO2 to 100 /t CO2.

In order to overcome the shortcomings of the existing directive and con-solidate positive experience, in December 2005 the Commission publishedits Further guidance on allocation plans for the 2008 to 2012 trading periodof the EU Emissions Trading Scheme (European Commission, 2005). Infact, the Commission recognized that the emissions trading directive wasrather broad and is subject to multiple interpretations, increasing the com-plexity of the system and making the process of evaluation extremelydicult and unclear.

This document is an attempt to provide a clear and univocal interpreta-tion of all the provisions of Directive 2003/87/CE, and also introduces aunique and compulsory format for the presentation of the relevant infor-mation within the NAPs. This requires a large amount of detailed infor-mation in order to assess all aspects of the present and future trends inGHG emissions. Once again, the legislator underlines the importance ofthe domestic actions, drawing up a list of criteria to be used to evaluate theconsistency of the national caps and their targets. Domestic measures


aimed at reducing the emissions have to be applicable to both trading andnon-trading sectors and would be integrated with credits from JI andCDM.

The guidelines did not stipulate specic national caps. However, theymentioned the reduction (of 6 per cent) to be made with respect of the rstperiod of trading. It is clear that this reduction would be distributed amongthe countries, taking into account how far they were from the Kyoto targetand the eciency of the domestic measures already in place.

The possibility of using the credits from the exible Kyoto mechanismsis of strategic importance, as it enables the Kyoto obligations to be fullled,thus minimizing the cost of compliance. In order to evaluate the correctusage of the mechanisms, detailed information on timing, entity of theproject and the expected volume of permits to be acquired have to beindicated.

All of the problems raised during Phase I are rather dicult to tackle.The treatment of new entrants, closures and small and medium installa-tions was not resolved to the satisfaction of all parties. The decision regard-ing the possibility of setting up a unique common reserve for new entrants,as well as creating common rules on closures was postponed until June2006, but even after that date no major changes took place. As for the allo-cation, the Commission encourages the deployment of allocation methodsother than grandfathering. In order to widen the coverage of the directive,the broad interpretation of the denition of the combustion installationwas chosen. The problem of windfall prots was not addressed in theguidelines.

Thus we could say that the problems revealed in Phase I of trading weresolved only partially. However, the unique denition of the combustioninstallation as well as other provisions will all help to increase the homo-geneity of the coverage, allocation rules and, consequently, the NAPs. Theadditional measures aimed at decreasing emissions are becoming highlyrelevant, as they will have a direct inuence on absolute caps of memberstates.

Moving from the rst to the second phase, the Commission is gradu-ally reducing the degrees of freedom of national governments in makingtheir decisions (for example, on the entity of caps or limits of JI andCDM). Also in this phase, the problem of windfall prots has remainedunsolved.

With regard to the current situation, all the NAPs have been presentedand evaluated, without signicant delays, and their main features aredescribed in Table 2.3. The homogeneity of the plans increased remarkably.At the same time the coverage of the EU ETS increased, including manyof the installations of the sectors not covered directly by Directive


26

Tab

le 2

.3M

ain

feat

ures

rel

ated

to

the

allo

cati

on o

fP

hase

II

Mem

ber

stat

eTo

tal e

mis

sion

sK

yoto

tar

get

(t)

Tota

l pro

ject

edTo

tal e

mis

sion

sA

vera

ge e

mis

sion

sC

O2

1990

(t)

emis

sion

s 20

10 (

t) B

AU

CO

2 20

05 (

t)E

TS

2005

06

(t)

Aus

tria

78,9

00,0

0068

,700

,000

92,5

00,0

0093

,300

,000

32,8

77,8

30B

elgi

um14

6,80

0,00

013

5,90

0,00

014

1,60

0,00

014

3,80

0,00

055

,069

,279

Cyp

rus

6,00

0,00

0n.

t.12

,200

,000

9,90

0,00

05,

169,

075

Cze

ch R

epub

lic19

6,30

0,00

018

0,60

0,00

014

5,70

0,00

014

5,60

0,00

083

,039

,798

Den

mar

k69

,300

,000

54,8

00,0

0067

,800

,000

63,9

00,0

0030

,337

,653

Est

onia

43,5

00,0

0039

,600

,000

18,9

00,0

0020

,700

,000

12,3

65,5

53F

inla

nd71

,100

,000

71,1

00,0

0085

,000

,000

69,3

00,0

0038

,860

,557

Fra

nce

564,

000,

000

563,

900,

000

569,

000,

000

553,

400,

000

127,

281,

656

Ger

man

y1,

231,

500,

000

973,

700,

000

955,

400,

000

1,00

1,50

0,00

047

6,08

2,09

3G

reec

e11

1,70

0,00

013

8,80

0,00

015

0,40

0,00

013

9,20

0,00

070

,616

,452

Hun

gary

122,

200,

000

114,

900,

000

87,4

00,0

0083

,100

,000

25,9

36,8

62Ir

elan

d55

,800

,000

63,0

00,0

0068

,400

,000

69,9

00,0

0022

,071

,898

Ital

y51

9,50

0,00

048

5,70

0,00

058

7,30

0,00

058

2,20

0,00

022

6,47

4,93

7L

atvi

a25

,300

,000

23,8

00,0

0013

,600

,000

10,9

00,0

002,

897,

589

Lit

huan

ia48

,000

,000

44,3

00,0

0033

,500

,000

22,6

00,0

006,

560,

390

Lux

embu

rg12

,700

,000

9,10

0,00

014

,200

,000

12,7

00,0

002,

658,

161

Mal

ta2,

200,

000

n.t.

2,20

0,00

03,

400,

000

1,97

8,51

2N

ethe

rlan

ds21

4,60

0,00

020

1,70

0,00

021

8,00

0,00

021

2,10

0,00

078

,526

,240

Pola

nd58

6,90

0,00

055

1,70

0,00

042

0,00

0,00

039

9,00

0,00

020

5,45

1,92

1Po

rtug

al60

,900

,000

77,4

00,0

0088

,000

,000

85,5

00,0

0034

,754

,906

Slov

akia

73,4

00,0

0067

,500

,000

58,3

00,0

0048

,700

,000

25,3

87,5

06Sl

oven

ia20

,200

,000

18,6

00,0

0021

,600

,000

20,3

00,0

008,

781,

366

Spai

n28

9,40

0,00

033

2,80

0,00

033

2,80

0,00

044

0,60

0,00

018

1,11

1,92

1Sw

eden

72,3

00,0

0075

,200

,000

70,3

00,0

0067

,000

,000

19,6

31,1

97U

nite

d K

ingd

om77

9,90

0,00

068

2,40

0,00

062

4,90

0,00

065

7,40

0,00

024

6,80

5,73

0To

tal

5,40

2,40

0,00

04,

975,

200,

000

4,87

9,00

0,00

04,

956,

000,

000

2,02

0,72

9,07

7

27

Mem

ber

stat

eA

vera

ge y

ears

Ave

rage

ann

ual

Red

ucti

onV

iola

ted

crit

eria

JI &

CD

MK

yoto

allo

cati

onal

loca

tion

at

requ

ired

by

(%)

mec

hani

sms

(t)

2008

12

(t)

rese

rve

(t)

EU

(t)

Aus

tria

30,7

29,9

0633

0,00

02,

070,

094

1,5,

1210

.00

9,00

0,00

0B

elgi

um58

,507

,703

5,00

0,00

04,

820,

532

1,2,

3,6,

10n.

a.7,

000,

000

Cyp

rus

5,47

9,78

096

0,00

01,

641,

718

1,2,

3,6,

10n.

a.n.

a.C

zech

Rep

ublic

86,8

36,2

64n.

a.15

,064

,736

n.

a.n.

a.

Den

mar

k24

,500

,000

500,

000

12

17.0

14,

200,

000

Est

onia

12,7

17,0

581,

840,

000

11,6

57,9

871,

2,3,

5,6

n.a.

n.a.

Fin

land

37,5

57,8

911,

400,

000

2,04

2,10

91,

2,3,

10,1

210

.00

2,40

0,00

0F

ranc

e13

2,80

0,00

03,

940,

000

6

(1,2

,3)

n.a.

n.a.

Ger

man

y45

3,07

0,17

5n.

a.28

,929

,825

1,2,

3,5,

10n.

a.n.

a.G

reec

e69

,087

,549

6,15

6,09

26,

414,

060

1,2,

3,6,

10n.

a.n.

a.H

unga

ry26

,908

,852

n.a.

3,82

4,46

11,

2,3,

10n.

a.n.

a.Ir

elan

d21

,151

,244

1,13

8,80

01,

486,

756

1,5,

6,10

,12

21.9

13,

600,

000

Ital

y20

1,57

2,00

015

,650

,000

13,2

53,5

141,

6,10

,12

14.9

919

,000

,000

Lat

via

3,28

3,30

33,

492,

000

4,48

0,58

01,

2,3

n.a.

n.a.

Lit

huan

ia8,

851,

304

n.a.

7,73

8,69

61,

2,3,

6,10

n.a.

n.a.

Lux

embu

rg49

7,66

1n.

a.1,

259,

094

1,2,

3n.

a.4,

700,

000

Mal

ta2,

143,

061

780,

531

812,

539

1,2,

3,10

,12

n.a.

n.a.

Net

herl

ands

85,8

13,4

586,

200,

000

3,90

2,65

41,

2,3,

5,10

,12

10.0

020

,000

,000

Pola

nd20

8,51

5,39

5n.

a.76

,132

,937

1,2,

3,5,

6,10

,12

10.0

0n.

a.Po

rtug

al34

,810

,329

5,08

0,00

01,

089,

671

1,6,

1210

.00

5,80

0,00

0Sl

ovak

ia30

,912

,261

1,80

4,22

210

,387

,739

1,2,

3,5,

6n.

a.n.

a.Sl

oven

ia8,

298,

937

n.a.

6,

1215

.76

600,

000

Spai

n15

2,25

0,72

97,

825,

000

422,

271

1,6,

10,1

2n.

a.31

,800

,000

Swed

en22

,802

,439

n.a.

2,39

7,56

11,

2,3,

10,1

210

.00

1,20

0,00

0U

nite

d K

ingd

om24

6,20

0,00

016

,320

,250

10

n.a.

n.a.

Tota

l1,

965,

269,

299

72,1

26,8

9519

9,82

9,53

4

n.a.

109,

300,

000

2003/87/CE, such as chemicals and petrochemicals. Almost all thecountries, except the UK, Slovenia, France and Denmark, were asked toreduce their amounts, altogether totalling 199.8 Mt of CO2. These reduc-tions are aimed at ensuring the consistency of the price of the permit andto bring the member states on track towards the Kyoto commitment.

Before dealing with future prospects, we shall discuss the costs of the EUETS. Unfortunately exact gures are extremely dicult, if not impossible,to obtain. However in its working document (2008a), the EuropeanCommission estimate that administrative costs for operators vary from2,000 to 15,000 per year and for authorities from 3,000 to 10,000 persite per year.

As stated earlier, there are still a number of issues that have to beresolved in order to improve the EU ETS and make it more ecient after2012. The experience that will be gained during the 200812 period shouldprovide clearer and more convincing answers to the following questions:the relation of the EU ETS with other similar initiatives and with theglobal carbon market for JI and CDM credits; the possibility of expand-ing the coverage to other sectors; the permits price formation mechanismsand so on.

2.5 2013 AND BEYOND

In March 2007, the European Council approved an EU objective of 30 percent reduction of GHG emissions by 2020 provided that developed coun-tries would commit themselves to comparable emission reductions and thateconomically advanced developing countries should contribute appropri-ately. In a stand-alone action, the EUs commitment will decrease to 20 percent by 2020. In the longer term, by 2050, the European Council hasrearmed that developed countries should reduce their emissions by 60 to80 per cent compared to 1990.

Bearing in mind these objectives, the Commission published its proposalfor a new emissions trading directive, amending Directive 2003/87/EC inorder to improve and extend the GHG allowance trading system (EuropeanCommission, 2008b).

In this section we shall present the most important elements of this pro-posal, to discover whether the persisting problems of the rst emissionstrading directive will be solved by this new draft.

Addressing one of the critical points, namely the scope of the directive,the proposal gives an explicit denition of the combustion installation anda clear list of activities and gases covered by the scheme. Thus the follow-ing activities were added in order to expand the coverage:


production and processing of ferrous metals (CO2); production of aluminium (CO2 and polychlorinated uorocarbons

PFCs); production and processing of non-ferrous metals (CO2); installation for manufacture of rock wool or stone wool (CO2); installations for the drying or calcination of gypsum or for the pro-

duction of plaster board and other gypsum products (CO2); Chemical industry:

production of carbon black (CO2);production of nitric acid (CO2 and N2O);production of adipic acid (CO2 and N2O);production of glyoxal and glyoxylic acid (CO2 and N2O);production of ammonia (CO2);production of basic organic chemicals (CO2);production of H2 and synthesis gas (CO2);production of soda ash and sodium bicarbonate (CO2);

capture, transport and geological storage of GHG emissions:installations to capture GHGs for the purpose of transportand geological storage in storage site (all GHGs listed inAnnex II);pipelines for transport of GHG for geological storage in astorage site (all GHGs listed in Annex II);storage sites for the geological storage of GHGs (all GHGs listedin Annex II).

The emissions from all these industries can be measured and veried withsucient accuracy and their inclusion would increase the volume coveredby the EU ETS by an additional 100 Mt of CO2. Such an expansion shouldlead to a rise in the eciency of trading. However, the road transport andshipping are not included in the proposal.

Another unresolved problem was the treatment of small installations. Itseems that in the current draft the solution has been found. The combus-tion installations with a rated thermal input below 25 MW and annualemissions less than 10,000 tons of CO2 in each of the three years precedingthe year of application may be excluded. The Commission expects thatabout 4,200 (0.70 per cent) installations would opt out. This provision isnecessary to reduce the administrative costs of the system.

The penalties remain the same, but should be indexed to the annualination rate of the Eurozone.

Moreover, the auction will become the basic method of allocation after2012. In the case of the power sector, all the allowances will be auctioned.For the other sectors, a period of transition is foreseen. This period will


start in 2013, beginning with the allocation of 80 per cent for free, subse-quently achieving zero free allocation in 2020. The Commission foreseesthat in 2013 about two-thirds of the total quantity will be auctioned.Ninety per cent of this amount will be distributed among the member statesaccording to their relative share of 2005 emissions in the EU ETS. Theremaining 10 per cent should be redistributed, taking into account fairnessand the common interest. Some of the revenues from the auctioning shouldbe used to support environmental policies within the EU.

Some rms that are particularly exposed to international competitionand thus to carbon leakage, as well as heat and steam delivered to districtheating, will continue to receive the allowances for free. However, the freeallocation will be subject to Community-wide rules.

A pan-European reserve should be created for new entrants. Allocationfrom this reserve will follow the general rules for existing installations.

The total amount of allowances should continue to decrease by a linearfactor of 1.74 per cent.

Also, further harmonization of use of credits from the Kyoto exiblemechanisms is necessary. The EU ETS should be able to link to othernational and international emissions trading schemes.

This analysis of the proposal of the new emissions trading directiveindicates that several problems which emerged during the rst two phaseshave been successfully addressed (among them, the problem of windfallprots). Some important aspects (for example, free allocation rules) are stillto be dened, but there is clear progress towards harmonization. The EUETS is also becoming more centralized, that is, national governments aregradually losing their autonomy. However, the common gain is clear thereis greater eciency and fewer distortions.

2.6 SOME FINAL REMARKS ANDCONSIDERATIONS

The EU ETS was launched on 1 January 2005. This event, which only afew years ago seemed a remote possibility, heralded the EU as a leader inthe climate change mitigation process. But, even more important, itopened a new era, highlighting the change in priorities of energy and envi-ronmental policies. Climate change has become a focal/hot point of theeconomic and political agendas, both at national and international levels.

Much positive and negative experience has been gained by all actorstaking part in the EU ETS (industry, the Commission, nancial institutionsand so on) since 2005 and many steps/challenges should be undertaken inthe future, the most important of which we shall outline here.


First, the rst largest international emissions trading system has beenlaunched and has completed its rst period of trading. This fact proves thatit is feasible. Despite many misgivings, such global trading pollutantsystems can be put in place and can potentially bring desirable results.

Initial problems relating to the installation of national and internationalauthorities involved in this complex trading scheme were overcome. Therehave been problems in coordination and the sharing of responsibilities. Thecontinual delays with respect to the Phase I timetable (especially within therst year of trading) are a reection of this. However, the system is nowestablished and the process is continuing without major problems (at leaston the administrative side).

At rst, there was a serious lack of precise and reliable information onmany important aspects for the future and overall design of the scheme. Forexample, there was no recent information on emissions at the installationlevel and sometimes at the sector/activity level. Together with a high degreeof information asymmetry, this led to signicant distortions in the initialallocation the overallocation experienced in Phase I is a clear consequenceof this. During this phase, this problem was largely solved, and a reliablesystem of information, collection, monitoring and verication was created.Future allocations will be based on these new data, reducing the risk ofdistortions.

It was immediately clear that the initial allocation is of fundamentalimportance for the success of the ETS. Many studies have assessed alloca-tion methodologies in order to determine the weak and strong points and,consequently, to indicate the most appropriate of them (see Harrison andRadov, 2002; Hepburn et al., 2006; Ellerman et al., 2007 and Harrisonet al., 2007).

Directive 2003/87/CE left the choice of allocation method to nationalstates. Thus a number of dierent methodologies/approaches have beenproposed. The preferred methods was free allocation based on grand-fathering which, together with other factors, resulted in signicant problemsof equity (for example, windfall prots) and distortions in competitiveness(McKinsey and Ecofys, 2006; Neuho et al., 2006; Reinaud, 2005; Sijmet al., 2006).

Thanks to the experience gained and to the theoretical and empirical evi-dence demonstrated by dierent studies, auctioning now seems to be themost suitable method of allocation. It will allow the resolution of manyimportant problems (which were still unsolved under Phase II), such aswindfall prots, overallocations (possibility of), creation/application ofcomplex allocation formulae and so on. It is robust, simple and decreasesthe complexity of the system. Of course it has its negative aspects, such asthe possibility of so-called carbon leakage, (due to the decrease in


competitiveness of EU industries) and it raises a huge redistributionproblem (as auctions will generate considerable revenues).

Particular attention should be paid to the design of auctions in order notto create distortions due to, for example, information asymmetry (regard-ing abatement and other costs) or the dierent nancial situations of par-ticipants. And, once again, economic theory could make an importantcontribution, indicating the most feasible solutions.

The rise in consumer prices due to ETS is one of the most debated topicsof recent years. There was and still continues to be a debate among indus-tries, stakeholders (institutional) and consumers regarding their responsi-bilities and to nd out what should or should not be done to reduce theseeects. We should emphasize that this increase is a normal and desired con-sequence of emissions trading. Product prices have to change (increase)with the price of environmental commodities (CO2), because it is the rightway to send signals to consumers in order to change their preferences andconsequently their level of consumption.

Unfortunately it has been extremely dicult to verify the extent of thiseect, as well as the eect on emissions abatement, in the early years of theEU ETS. In fact in Phase I, the carbon price was generally too low. The factthat surrendered emissions are substantially the same in 2005 and 2006 (seeTable 2.2) conrms this statement. In any case, we should not underesti-mate the fact that a price was put on CO2, even if it fell to very low valuesdue to overallocation (after April 2006).

Furthermore, the trading system was put in place and has been tested.Markets, trading rules and dierent nancial products related to this specicmarket were established, enabling the installations to exchange allowances.We are witnessing the creation of an environmental market with a completeset of actors. Further development of this market is desirable and is of vitalimportance, in particular greater articulation of services and products tohedge the dierent risks related to emissions trading. This will help to over-come the problems of price volatility, market liquidity and price formation,and consequently could have a positive impact on investment patterns.

Links to other systems, for example, the inow of JI and CDM, are ofvital importance. This is one of the main challenges of the 200812 period,although it is still quite unclear how the system will work. It is no secret thatmost operators still have no experience in this eld. The possibility of intro-ducing credits into the ETS system could potentially have a signicantimpact on the liquidity of the market and thus on the carbon price, alsoreducing the cost of compliance. We have yet to see what eect these mea-sures will have.

As discussed in previous sections, the proposed new emissions tradingdirective foresees further expansion of the sectors and activities covered by

32 Overv

markets for carbon and power

Documents