EUROPEAN COMMISSION

Brussels, 15.7.2015

SWD(2015) 135 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Accompanying the document

Proposal for a Directive of the European Parliament and of the Council

amending Directive 2003/87/EC to enhance cost-effective emission reductions and low-carbon investments

{COM(2015) 337 final}
{SWD(2015) 136 final}

TABLE OF CONTENTS

1.    Procedural issues and consultation of interested parties    

1.1.    Identification    

1.2.    Organisation and timing    

1.3.    Consultation and expertise    

1.3.1.    Expertise used    

1.3.2.    Consultation    

1.4.    Subsidiarity    

1.5.    Scrutiny by the Commission Impact Assessment Board    

2.    Policy Context    

3.    General evaluation and lessons learnt    

4.    General problem definition    

5.    Objectives    

5.1.    General policy objectives    

5.2.    Specific policy objective    

5.3.    Consistency with other policies and objectives    

6.    Implementation of the EU's GHG emission reduction target    

6.1.    Linear reduction factor    

6.2.    Auction share    

7.    Free allocation and addressing the risk of carbon leakage    

7.1.    Problem definition    

7.1.1.    Underlying drivers of the problem    

7.2.    Operational policy objectives    

7.3.    Policy options for free allocation and other carbon leakage measures    

7.3.1.    Benchmark values    

7.3.2.    Production level and adjustments    

7.3.3.    Reserve for new entrants    

7.3.4.    Carbon leakage groups    

7.3.5.    Cross-sectoral correction factor    

7.3.6.    Compensation for indirect carbon costs    

7.4.    Option packages    

7.5.    Analysis of impacts    

7.5.1.    Direct and indirect effects    

7.5.2.    Environmental impacts    

7.5.3.    Economic impacts    

7.5.3.1.    Quantified impacts on competitiveness    

7.5.3.2.    Administrative burden    

7.5.4.    Social impacts    

7.5.4.1.    Employment    

7.5.4.2.    Energy prices for households    

7.5.5.    Comparing the options    

7.5.5.1.    Quantified environmental and social impacts    

7.5.5.2.    Reflect technological progress in industry sectors    

7.5.5.3.    Fully preserve incentives for industry to innovate    

7.5.5.4.    Most efficient installations do not face undue carbon costs leading to carbon leakage    

7.5.5.5.    Better alignment with production levels    

7.5.5.6.    Avoid windfall profits    

7.5.5.7.    No increased administrative complexity    

7.5.6.    Concluding remarks    

8.    Low-carbon funding mechanisms    

8.1.    Innovation fund    

8.1.1.    Problem definition    

8.1.1.1.    Underlying drivers of the problem    

8.1.2.    Operational policy objectives    

8.1.3.    Policy options    

8.1.3.1.    Screening of projects    

8.1.3.2.    Conditionality of Awards    

8.1.3.3.    Type of instrument    

8.1.3.4.    Maximum rate of funding    

8.1.3.5.    Parameters not varied in the options    

8.1.4.    Option packages    

8.1.5.    Analysis of impacts    

8.1.5.1.    General impacts    

8.1.5.2.    Economic impacts    

8.1.5.3.    Environmental impacts    

8.1.5.4.    Social impacts    

8.1.6.    Comparing the options    

8.2.    Modernisation Fund    

8.2.1.    Problem definition    

8.2.1.1.    Context    

8.2.1.2.    Underlying drivers of the problem    

8.2.1.3.    Lessons learnt    

8.2.2.    Policy Objectives    

8.2.3.    Development and screening of policy options    

8.2.3.1.    Policy options    

8.2.3.2.    Option packages    

8.2.4.    Analysis of Impacts    

8.2.4.1.    Effectiveness    

8.2.4.2.    Coherence    

8.2.4.3.    Market Distortion    

8.2.4.4.    Administrative Burden    

8.2.5.    Comparing the Options    

8.3.    Free allocation to promote investments for modernising the energy sector    

8.3.1.    Policy objective and problem definition    

8.3.2.    Conclusions on lessons learnt and policy context    

8.3.3.    Operational policy objectives    

8.3.4.    Development and screening of policy options    

8.3.4.1.    Aspects not varied between the options    

8.3.5.    Retained option packages    

8.3.6.    Assessment of options    

8.3.6.1.    General impacts - Effectiveness    

8.3.6.2.    General impacts - Complexity and administrative burden / flexibility to adjust to MS circumstances    

8.3.6.3.    General impacts - Transparency    

8.3.6.4.    General impacts - Potential distortion of EU energy market / complementarity with existing EU instruments    

8.3.6.5.    Timing and distribution of investments and volume and timing of allowances on market / auction revenues of investments    

8.3.6.6.    Environmental impacts - Potential decarbonisation    

8.3.6.7.    Social impacts    

8.3.7.    Comparing the options    

8.4.    Inter-linkages between the low carbon funding mechanisms    

9.    Monitoring and evaluation    

Annexes    

1.    Glossary    

2.    Summary of relevant findings of the 2030 impact assessment    

3.    Summary of stakeholder consultations    

3.1.    General consultation on the ETS revision    

I.    Free allocation and addressing the risk of carbon leakage    

Industry stakeholders    

Public authorities    

Civil society    

II.    Innovation Fund    

Industry stakeholders    

Public authorities    

Civil society    

III.    Modernisation Fund    

Industry stakeholders    

Public authorities    

Civil society    

IV.    Free allocation to promote investments for modernising the energy sector    

Industry stakeholders    

Public authorities    

Civil society    

V.    SMEs/regulatory fees/other    

Industry stakeholders    

Public authorities    

Civil society    

VI.    General evaluation    

Industry stakeholders    

Public authorities    

Civil society    

3.2.    Consultation on carbon leakage provisions and innovation support    

I.    Business and trade representations of business interests    

II.    Government and regulatory authorities    

III.    Academic and research organisations, NGOs, citizens, trade unions (civil society)    

4.    Summary of evaluation    

4.1.    NER 300 - lessons learnt    

4.1.1.    Introduction    

4.1.2.    NER 300 principles    

4.1.2.1.    Definition    

4.1.2.2.    Legal basis    

4.1.2.3.    Selection of projects    

4.1.3.    State-of-play    

4.1.3.1.    Results of calls for proposals    

4.1.4.    Lessons learnt so far    

4.1.4.1.    Geographical coverage    

4.1.4.2.    Technological coverage    

4.1.4.3.    Support to innovation    

4.1.4.4.    Cost-effectiveness in the use of NER 300 funds    

4.1.4.5.    Administrative simplification    

4.1.4.6.    Monetisation of allowances for grants    

4.1.4.7.    No ring-fencing    

4.1.4.8.    Number and timing of calls    

4.1.4.9.    Length of preparatory phase    

4.1.4.10.    Impact of projects    

4.1.4.11.    NER 300 and other financing instruments    

4.1.5.    Main challenges encountered    

4.1.5.1.    Financial barriers    

4.1.5.2.    Management structure    

4.1.6.    Conclusions    

4.2.    Free allocation to the power sector    

4.3.    Exclusion of small emitters – lessons learnt    

4.4.    Guaranteeing a robust and secure EU ETS (registry fees) – lessons learnt    

5.    Detailed description and screening of options for free allocation to industry    

5.1.    Benchmarks    

5.1.1.    Lessons learnt from the current system in phase 3 (2013-2020)    

5.1.2.    Policy options for updating benchmark values for 2021-30    

5.1.3.    Screening    

5.2.    Carbon leakage groups and criteria    

5.2.1.    Lessons learnt from the current system in phase 3 (2013-20)    

5.2.2.    Policy options    

5.2.3.    Screening    

5.3.    Production level and adjustments    

5.3.1.    Lessons learnt from the current system in phase 3 (2013-20)    

5.3.2.    Policy options    

5.3.3.    Screening    

5.4.    Reserve for new entrants    

5.4.1.    Lessons learnt from the current system in phase 3 (2013-20)    

5.4.2.    Policy options    

5.4.3.    Screening    

5.5.    Compensation for indirect carbon costs    

5.5.1.    Lessons learnt from the current system in phase 3 (2013-20)    

5.5.2.    Policy options    

5.5.3.    Screening    

6.    Further improvements to the current set of rules    

6.1.    Validity of emission allowances    

6.1.1.    Problem definition    

6.1.2.    Operational policy objective    

6.1.3.    Policy options    

6.1.4.    Analysis of impacts    

6.1.5.    Comparing the options    

6.2.    Optional exclusion of small emitters    

6.2.1.    Problem definition and general objective    

6.2.2.    Operational policy objective    

6.2.3.    Policy options    

6.2.3.1.    Assessment of the options    

6.2.3.2.    Comparing the options    

6.3.    Guaranteeing a robust and secure EU ETS (registry fees)    

6.3.1.    Problem definition and general objective    

6.3.2.    Operational policy objectives    

6.3.3.    Policy options    

6.3.3.1.    Screening of the options    

6.3.3.2.    Analysis of impacts    

7.    Impacts of different option packages for free allocation on individual sectors    

7.1.    Effect of carbon leakage groups on sectors    

7.2.    Cross sectoral correction factor    

7.3.    Estimated level of free allocation for biggest sectors in phase 4    

7.4.    Estimated compliance cost per sector, taking into account cost pass through    

7.5.    Sensitivity analysis    

8.    Impacts of different option packages for free allocation: Administrative burden and social impacts    

8.1.    Analysis of administrative burden    

8.2.    Analysis of social impacts    

8.2.1.    Employment    

8.2.2.    Energy prices for households    

9.    Literature review on cost-pass through rates    

9.1.    Theoretical studies on cost pass-through    

9.2.    Ex-ante empirical studies on the EU ETS    

9.3.    Ex-post empirical studies for the EU ETS    

9.4.    Conclusions    

10.    Description of Financial Barriers for energy sector: power generation, energy efficiency and networks    

11.    Impacts of creating the Modernisation Fund compared to the Baseline    

12.    More comprehensive overview of impacts for different options for creating the Modernisation Fund    

13.    Timing and approach to the monetisation of allowances for the Innovation fund and Modernisation Fund    

14.    Sensitivity analysis for possible funding rates for the Innovation Fund    

15.    Options for unallocated allowances    

Policy options    

Comparison between the options    

List of tables

Table 1: Option packages for addressing the risk of carbon leakage    

Table 2: Estimated number of sectors and distribution of free allowances under 'Baseline', 'limited changes' and 'targeted 'options    

Table 3: Overview of assessment of policy option packages compared to 'Baseline B'    

Table 4: Option packages for Innovation Fund    

Table 5: Comparison of options for the innovation fund    

Table 6: Retained options for examining for the Modernisation Fund    

Table 7: Impact of policy options for the Modernisation Fund    

Table 8: Option packages for the free allocation to the power sector    

Table 9: Comparison of options for the free allocation to the power sector    

Table 10 Stakeholder consultation responses    

Table 11: Projects submitted and awarded    

Table 12: Distribution of awarded NER 300 funding    

Table 13: Screening of options for updating benchmark values    

Table 14: Overview of minimum cost pass-through values found in existing literature    

Table 15: Screening of options for carbon leakage groups and criteria    

Table 16: Screening of options for production level and adjustments    

Table 17: Screening of options for reserve for new entrants    

Table 18: Screening of options for compensation for indirect cost    

Table 19: Comparison of options    

Table 20: Comparison of the options in relation to relevant problems and objectives    

Table 21: Comparing the options in relation to relevant problems and objectives    

Table 22: Estimated number of sectors, share of gross free allocation and GVA in the different groups    

Table 23: Estimated carbon leakage status at sectoral level based on past data    

Table 24 Overview table on estimated compliance cost per sector, taking into account cost pass-through rates    

Table 25 Sensitivity analysis of compliance costs compared to sector turnover at different carbon prices    

Table 26: Estimated administrative costs for additional NIMs exercises:    

Table 27: Estimated additional annual administrative costs for production level adjustments    

Table 28: Estimated administrative costs for harmonised rules for indirect cost compensation (additional costs compared to national compensation currently applied in 6 MSs as baseline)    

Table 29: Estimated employment impacts of policy option packages compared to baseline B from the pass through of additional costs (annual average for 2021 – 2030, underlying assumptions in italics)    

Table 30: High estimate of employment impacts of policy option packages compared to baseline B from additional costs absorbed (annual average for 2021 – 2030, underlying assumptions in italics)    

Table 31: Analysis of impacts of policy option packages on the district heating sector 2021-30    

Table 32    Estimate of cost pass-through rates from ex-post econometric studies    

Table 33    Overview of the range of average cost pass-through in selected sectors from literature    

Table 34: Comparison under baseline and creation of the Modernisation Fund for allocation of 2% of total allowances, 2021-2030    

Table 35: Modernisation Fund allocation shares per beneficiary Member States, 2021-2030    

Table 36: Monetised health benefits in 2030 as a percentage of GDP    

Table 37: Reduced air pollution control costs by 2030 as a percentage of GDP    

Table 38: EU28 jobs associated with investments in the power sector and energy efficiency, 2011-2030    

Table 39: Overview of impacts for different options for creating the Modernisation Fund    

Table 40: Sensitivity of amount to raise by RES, CCS or industry project sponsor to funding rates and project size    

Table 41: Sensitivity of 15% to price of carbon    

Table 42: Comparison of options for further action with unallocated allowances    

List of figures

Figure 1 Respondents from industry    

Figure 2: Structure of the total quantity of allowances in phase 3    

Figure 3: Sectors on and off the carbon leakage list.    

Figure 4: Estimated combined effect of carbon leakage groups and the correction factor under different option packages    

Figure 5: Annual allowances from fund by type    

Figure 6: Combined free allocation trajectories    

Figure 7: Stakeholder consultation responses    

Figure 8: Industry stakeholders – shares of each sector replying    

Figure 9: Distribution of installations in terms of amount of emissions    

Figure 10: Annual Emissions from small installations (installations with annual emissions < 25 kt) and from opt-out installations    

Figure 11: Ranking of sectors in terms of emissions in EU ETS    

Figure 12: Indicative carbon leakage groups in the 'Limited changes' option package based on 2009-2011 data    

Figure 13: Indicative carbon leakage groups in the 'Targeted' option package based on 2009-2011 data    

Figure 14: Estimated combined effect of carbon leakage groups and the correction factor under different option packages    

Figure 15: Employment in selected energy-intensive industries    

Figure 16    Cost pass-through rates derived by Vivid Economics (2014) in sectors investigated in reduced and full detail (2020, € 15/tCO2)    

Executive Summary Sheet

1.Procedural issues and consultation of interested parties

1.1.Identification

Lead Directorate-General (DG): Climate Action

Other services involved: Secretariat-General; Legal Service; DG Budget; DG Competition; DG Economic and Financial Affairs; DG Employment; DG Energy; DG Environment; DG Internal Market, Industry, Entrepreneurship and SMEs; DG Mobility and Transport; DG Regional and Urban Policy; DG Taxation and Customs Union; DG Trade and DG Research, Science and Innovation.

Work Programme 2015 reference: Included under initiative no. 5 Strategic framework for the Energy Union 

Agenda Planning reference: 2015/CLIMA/001

1.2.Organisation and timing

The analysis of the policy framework for climate and energy in the period from 2020 to 2030 (hereafter 'the 2030 framework') 1 has played a central role in identifying the key elements for the revision of the European Union Emission Trading System (EU ETS) for the period after 2020. The current impact assessment is a follow-up exercise from the impact assessment for the 2030 framework 2  and focuses on certain ETS-specific methodological elements not already assessed.

The work for this impact assessment (IA) started in December 2014 with the launch of a 12-week online consultation on the revision of the EU ETS 3 .The work also builds on the results of the consultation on the 2030 framework 4 and a separate consultation on the post-2020 carbon leakage provisions 5 .

DG Climate Action invited the above-mentioned Commission services to be part of an Impact Assessment Steering Group. Three meetings took place (on 18 December 2014, 26 February 2015 and 16 April 2015) where comments were exchanged and taken into due account. The final draft impact assessment was submitted to the group on 13 April 2015. Another meeting of the Steering Group took place to discuss the revised draft on 23 June 2015.

An evaluation of the existing ETS Directive is part of the impact assessment work and has fed into the assessment of the policy options. 6  

1.3.Consultation and expertise

1.3.1.Expertise used

This impact assessment builds on the impact assessment for the 2030 framework and analyses ETS-relevant aspects not covered in it.

In terms of external expertise, the Commission drew upon a study on evaluation of the ETS, commissioned in 2014 and carried out by a consortium led by ICF International 7 . Furthermore, in 2014, a study 8 was commissioned to assess the issue of costs being passed through from industrial sectors to their downstream customers and to determine the factors influencing such ability to pass through costs quantifying it for major energy intensive industry sectors. Another study 9 was commissioned to evaluate the experience gathered with the harmonised benchmark-based allocation process, and in particular to evaluate whether the benchmarks have achieved the intended objectives. In 2013, a study 10 was commissioned to assess the evidence for carbon leakage in the period 2005-2012 for ten major energy intensive industry sectors. The findings of these studies are discussed in chapter 7 .

In terms of data, in 2011-2013, Member States submitted to the Commission verified, detailed, confidential and commercially sensitive data on preliminary free allocation to industrial installations. 11  After checking their compliance with the harmonised allocation rules, 12 these data were used for the analysis of the amount of free allocation to industrial installations.

1.3.2.Consultation

Relevant stakeholders (Member States, industry representatives, NGOs, research and academic institutions, trade unions and wider public) were involved throughout the entire process. Complementing the consultation for the 2030 framework, an extensive follow-up stakeholder consultation was carried out on various technical aspects of the post-2020 carbon leakage provisions, as well as on innovation support. It included three stakeholder meetings (June, July and September 2014) 13 and a written consultation (May–July 2014) 14 .

An online consultation (December 2014-March 2015) also on other aspects (free allocation for the power sector, Innovation and Modernisation Funds, small and medium sized enterprises (SMEs), regulatory fees and general evaluation of the EU ETS) followed. The Commission did its best to accept also late submissions, and the consultation attracted a total of over 529 contributions from a broad spectrum of stakeholders. 15 The main findings of the public consultation are found in Box 1 and the summary report in Annex  3 .

Moreover, there have been bilateral meetings with many stakeholders, allowing them to express their specific views on the future of the system, as well as their dedicated efforts on certain aspects. A questionnaire on benchmarks for industry stakeholders was distributed in March 2015 and a conference on the Modernisation Fund was organised.

All of the above consultations have been published on the DG Climate Action website. The Commission minimum standards on stakeholder consultation have all been met.

The views of the stakeholders were taken into account to the extent possible in this impact assessment, given their number, the sensitivity and complexity of the issues and the diverging opinions expressed by different stakeholder groups. The respective sections indicate how the opinions have been reflected in the proposed options.

Box 1: Main findings of the general consultation on ETS revision

Figure 1 Respondents from industry

1.4.Subsidiarity

The EU ETS Directive 16 exists and will continue being in force post 2020. It is an EU policy instrument. A revision of the EU ETS can only be implemented through proposals by the Commission to amend the EU ETS Directive.  

Climate change is a trans-boundary problem. Therefore, coordination of climate action at European level and, where possible, at global level is necessary, and EU action is justified on grounds of subsidiarity. Articles 191 to 193 of the TFEU confirm and further specify EU competencies in the area of climate change.

Many of the policy elements have an important internal market dimension, in particular the options related to the indirect cost compensation for industry (see section  7.3.6 ), the Modernisation Fund (see section 8.2 ) and the free allocation to power (see section 8.3 ), and many of the required investments and infrastructures have an important European dimension, especially in the low-carbon funding mechanisms and the free allocation to industry.

The Modernisation Fund aims to modernise the energy systems and improve energy efficiency in the beneficiary Member States, but the fund also has important implications on the internal energy market as the investments supported would be made in the context of a liberalised European energy market.

Therefore, the objectives can be better achieved by an EU framework for action. Delegating the legislative powers to Member States would lead to partitioning, an uneven playing field and decreased efficiency.

1.5.Scrutiny by the Commission Impact Assessment Board

The Impact Assessment Board of the European Commission assessed a draft version of the present impact assessment and issued its opinion on 22 May 2015. The impact assessment was improved and resubmitted to the Board, which issued its final opinion on 17 June 2015. The Impact Assessment Board made several recommendations and, in the light of the latter, the final impact assessment report contains a clearer presentation and explanation of options and their impacts, including more assessment of impacts on particular sectors as added in section 7.5.3  and Annex  7 . The report also explains better the rationale behind the choices of values for thresholds and criteria in sections 7.3 and 8.1.3 . The coherence between various funding mechanisms is also addressed, as well as their options and impacts. Various methodological questions are clarified (use of carbon leakage criteria and relation to cost pass -through, modelling of employment effects and energy prices for households). Annex 15  was added to address the issue of using the unallocated allowances inter alia for carbon leakage and innovation as agreed in the context of the market stability reserve proposal. 17  A particular effort has been made throughout the text to simplify the language and improve readability for non-expert audience, including boxes briefly explaining ETS and carbon leakage.

2.Policy Context

Building on the Commission Communication on the 2030 framework and the accompanying impact assessment, in October 2014 the European Council 18  and with the endorsement of the European Parliament 19 agreed on the 2030 framework, which includes a binding domestic reduction of greenhouse gas (GHG) emissions of at least 40% in 2030 as compared to 1990. 20 To meet this target, the European Council also specifically agreed that the emissions in the EU ETS should be reduced by 43% compared to 2005. As a result of the measures needed to meet this emission reduction target for the 2030 framework and the Market Stability Reserve which is expected to become operational in 2019 as provisionally agreed by the co-legislators 21 , the EU ETS will deliver a meaningful carbon price and stimulate cost-efficient emission reductions. A reformed EU ETS will play an important role in giving the right investment signals to businesses and Member States to ensure emission reductions happen at the least cost.

The European Council also gave strategic guidance on several issues regarding the implementation of the emission reduction target, namely free allocation of allowances to industry, the establishment of a Modernisation Fund and an Innovation Fund, as well as optional free allocation of allowances to modernise electricity generation in some Member States. Similarly, the European Parliament highlighted the necessity of a revised and well-functioning ETS, including the continuation of provisions on the need to address the international competitiveness of industrial sectors deemed to be at risk of carbon leakage. This guidance is being translated into a legislative proposal to revise the EU ETS for the period post-2020 and this impact assessment focusses on choices to be taken to establish these rules.

The free allocation of allowances to industry is a key element in addressing industry competitiveness concerns and how to determine the rules for this allocation constitutes a key part of this impact assessment, addressed mainly in chapter 7 . The low-carbon funding mechanisms are analysed in chapter 8 .

This impact assessment and the legislative proposal for revision of the ETS Directive do not cover detailed methodological options to establish the technical measures to implement these revised rules. Respective legislative proposals and accompanying analysis will follow in due course.

The revision of the EU ETS makes an important contribution to the efforts to establish a resilient Energy Union 22 with a forward looking climate change policy at its core. The goal of the Energy Union is to give EU consumers – households and businesses – secure, sustainable, competitive and affordable energy.

The EU climate policy will continue contributing to a major shift away from expenditure on fuels towards innovative equipment with high added value that will stimulate investments for innovative products and services, create jobs and growth and improve the Union's trade balance 23 . As already shown, together with the 2020 targets for renewable energy and energy savings, the GHG emission reduction target and the ETS have played a key role in driving progress to low-carbon economy and sustaining the employment of more than 4.2 million people in various eco-industries 24 , with continuous growth during the crisis.

3.General evaluation and lessons learnt

The main aim of the EU ETS is to incentivise cost-effective emission reductions in the sectors in its scope. As the emission reductions and thus the positive environmental outcome in the EU ETS is guaranteed by its absolute limit on emissions, i.e. the cap 25 , the EU is currently well on track to meet its 2020 targets for GHG emissions reduction not only for the sectors in the EU ETS, but for the EU as a whole. As decided in the 2008 climate and energy package, thanks to the emission reductions incentivised by the European carbon market and the measures taken in sectors out of the EU ETS, as well as the policies on renewables and energy efficiency, in 2020 the GHG emissions from the sectors in the EU ETS will be at least 21% lower than in 2005. Thus, the ETS in particular, and EU climate and energy policies in general, seem to be working well.

The regulatory framework of the EU ETS was largely unchanged during the first eight years of its operation (2005-2012). Based on the lessons learnt in these years, in phase 3 (2013-2020), a significant number of architectural and regulatory changes took effect, improving the functioning of the system.

The following fundamental changes have been applied:

an EU-wide cap on allowances was introduced, as opposed to individual Member State caps. This EU-wide cap decreases by 1.74% annually, up to and beyond 2020, providing much greater regulatory predictability and stability;

auctioning became the default system of allocation of emission allowances in phase 3, ceasing the previous practice of granting free allocation for electricity production. Revenues from these auctions are distributed to Member States (for more details see section 6.2 ). At least 50% of these revenues are to be used for climate action. As a transitional derogation, some Member States can still grant some of their auction volume as free allowances to their power sector (for more details see section 8.3 );

harmonised rules for transitional free allocation to industrial sectors were introduced, ensuring that all companies in the EU in one sector receive free allowances based on harmonised rules to address the risk of carbon leakage. The amount of free allocation for the installations is calculated based on emission performance benchmarks established prior to phase 3, production data and a carbon leakage factor measuring the exposure of the industrial sectors to international competition and carbon leakage risk (for more details see chapter  7 );

stricter rules on the type and quantity of international credits that are allowed for use in phase 3. These aspects are not analysed in the current impact assessment, as they have been covered by the impact assessment by the 2030 framework;

replacement of national registries by a single Union registry;

Emission allowances were classified as a financial instrument with the latest revision of the financial markets legislation in 2014, to be enforced in 2017 26 .

The revised EU ETS Directive can be considered as fairly recent legislation. However, it is already clear that the present institutional framework with auctioning and EU-wide harmonised free allocation rules constitutes a significant improvement and simplification compared to the previous trading periods that still had National Allocation Plans. Allowing for auctioning to fully function as the main driving principle of the ETS - as compared to having rules for free allocation to industry and to power sector in certain Member States - would bring further simplification to the system.

Research concludes that in periods when the carbon price is likely to increase, other beneficial impacts of the EU ETS are visible through a broad range of mechanisms –incentives for cost-cutting and attention, experimentation, learning and investment in low-carbon solutions outside business-as-usual for companies 27 . Studies show that due to the EU ETS a large proportion of firms pursued some measures to reduce GHG emissions and that CO2 has now become part of the investment appraisal in construction of power stations 28 .

However, at the start of phase 3 in 2013, the EU ETS was characterised by a large imbalance between supply and demand of allowances, resulting in a surplus of around 2 billion allowances expected to grow over the coming years to more than 2.6 billion allowances by 2020, and a correspondingly weak carbon price signal.

As a short term measure to mitigate the effects of the surplus on the carbon market it was decided to postpone (“back-load”) the auctioning of 900 million allowances in the early years of phase 3 29 . This was followed by a proposal for a long-term measure of establishing a Market Stability Reserve to make the auction supply of emission allowances more flexible and increase shock resilience. The reserve's architecture also captures changes in the demand of allowances due to renewables and improved energy efficiency and, if need be, adjusts the auction supply accordingly. Hence, the Market Stability Reserve will, once fully functional, also strengthen the coherence between the EU ETS and energy efficiency and renewables policies, which also lead to lowering of emissions. The operation of the Market Stability Reserve is open-ended and does not affect the total quantity of emission allowances (the cap). Therefore, the introduction of the Market Stability Reserve improves the functioning of the EU ETS, but does not change its nature as a market-based policy instrument. 

According to many energy-intensive industry stakeholders, the EU ETS corresponded well to the EU climate objectives, namely to cost-effectively reduce GHG emissions, in its initial architecture. However, they argue that with back-loading and the Market Stability Reserve 30 , the EU ETS sectors will be facing a stricter GHG-target compared to EU climate policy objectives 31 . In contrast, other business stakeholders, including from renewables sector 32 , believe that the EU ETS is currently giving neither a long-term price signal that impacts investment decisions nor a short-term signal for operating decisions. Hence, they advocate in favour of structural measures, such as earlier implementation of the Market Stability Reserve and additional measures to address the surplus in order to 'fix' the EU ETS in the short and mid-term.  

Some stakeholders responded that the EU ETS does not correspond to the EU's climate policy objectives, because no country in the world has accepted the EU ETS concept 33 . However, it should be stressed that there are actually 17 emission trading systems in operation across four continents, accounting for 40% of global GDP 34 .

The study to evaluate the existing ETS Directive is ongoing 35 . It analyses the EU ETS in terms of relevance, effectiveness, efficiency, EU-added value and coherence with other Union policies.

The preliminary findings 36 indicate that despite criticism on some of its details, the EU ETS as a policy tool combining environmental regulation with a market instrument is working in practice and delivering on its targets. It is highly relevant for meeting the EU's climate targets, as it represents a cost-effective way for emission reductions. Emissions in the covered sectors have decreased steadily, and even though not all emission reductions can be attributed to the ETS alone, evidence has been found that the system does contribute effectively to emission reductions.

At the same time, the study observes that the EU ETS has found its way to the board rooms of companies and thus facilitates the internalisation of CO2 costs. The ETS contributes to investments decisions, even though with the current low carbon price, the CO2 costs are often included in the general envelope of energy costs. Smaller improvements in terms of GHG efficiency have become regular practice, but larger investments in GHG efficiency still remain the exception.

Moreover, the EU ETS has a clear EU value added since different systems or other climate policies at Member State level would lead to a fragmented and costly situation for the regulated entities, as well as different ambition levels and carbon prices throughout the EU, leading also to unfeasibility and administrative complexity. The EU ETS with an EU-wide carbon price and its harmonised infrastructure takes advantages of the synergies that EU action can provide.

In terms of coherence with EU policies, the study observes that renewables and energy efficiency policies both overlap with the EU ETS and may affect the cost of achieving the ETS' target, but both policies fully support the environmental effectiveness of the EU ETS as they do not affect the cap. In terms of coherence with international climate policy, the study also finds that the EU ETS performs well. It is widely used as a model for emissions trading systems around the world, benefitting from the EU's learning effects.

Specific evaluation and lessons learnt on the different aspects of the EU ETS are presented in the corresponding chapters.

A comprehensive evaluation summary report on individual aspects can be found in Annex  4 .

4.General problem definition

The general problem analysis concerning EU climate policy targets for 2030 and the ETS has been done in the impact assessment on the 2030 climate and energy framework.

In brief, despite the fact that the EU is on track to meet its short-term emission reduction target of -20% by 2020, to ensure the long-term goal of a low-carbon economy and emission reductions of 80 to 95% by 2050, an intermediate step needs to be made with the binding EU target of at least 40% domestic reduction by 2030 compared to 1990. This target should be delivered collectively by the EU in the most cost-effective manner possible, with the reductions in the ETS sectors amounting to 43% and 30% in the non-ETS sectors compared to 2005. Concerning the implementation of the emission reduction target, setting the cap at the emission level leading to a reduction of 43% would require a change in the linear reduction factor 37 from 2021 onwards. The European Council conclusions already foresee certain methodological elements for the implementation of the EU's GHG emission reduction target, for which no options were hence developed. These include the change in the annual linear reduction factor reducing the EU ETS cap from 2021 onwards and the share of allowances to be auctioned. These elements are outlined in chapter 6 . 

Free allocation and carbon leakage, low-carbon funding mechanisms and further improvements to the current system need to be analysed and specific problem definitions and options are developed in the respective sections.

Free allocation and carbon leakage

While current policies to prevent carbon leakage such as the allocation of free allowances in the ETS have been successful 38 , these do not automatically apply beyond 2020. In its strategic guidance the European Council has been clear that free allocation should not expire. Existing measures should continue after 2020 to prevent the risk of carbon leakage due to climate policy, as long as no comparable efforts are undertaken in other major economies, with the objective of providing appropriate levels of support for sectors at risk of losing international competitiveness. The problem to be addressed is how to adapt the respective ETS Directive provisions to make them suitable post-2020.

Low-carbon funding mechanisms: Modernisation Fund and Innovation Fund

The impact assessment of the 2030 climate and energy framework indicated that significant investments will be needed in the EU in the period through 2030 related to energy system modernisation and to reach the objectives of the 2030 climate and energy framework. In this context, the European Council has also given clear guidance that Member States with a GDP per capita below 60% of the EU average may opt to continue giving free allowances to modernise their energy sector up to 2030 (see section 8.3 ). In addition, a new reserve of 2% of the total quantity of allowances should be set aside to address particularly high additional investment needs in these Member States (so-called Modernisation Fund). The specific problems to be addressed with the setting up of the Modernisation Fund are analysed in section 8.2

The European Union will have to step up its efforts in research and innovation to support the post-2020 climate and energy framework. The European Council conclusions provide clear guidance that the existing NER 300 facility should be renewed, including for carbon capture and storage (CCS) and renewables, with the scope extended to low-carbon innovation in industrial sectors and the initial endowment increased to 400 million allowances (so-called Innovation Fund). The specific problems to be addressed with the setting up of the Innovation Fund are analysed in section 8.1

While the current architecture of the EU ETS is relatively recent, based on experience gathered, certain additional technical changes to the current set of rules in the ETS Directive should also be considered for the period post-2020. Since these changes are of more technical nature, options for these are developed in Annex 5 . 

This impact assessment explicitly does not address issues related to aviation emissions as covered under the ETS. As indicated in Regulation (EU) No 421/2014, following the 2016 ICAO Assembly, the Commission shall report to the European Parliament and to the Council on actions to implement an international agreement on a global market-based measure from 2020, that will reduce greenhouse gas emissions from aviation in a non-discriminatory manner.

5.Objectives

5.1.General policy objectives

The general objective of climate action policy, and of EU ETS as its key instrument, is to contribute to achieving the EU climate goal of limiting global average temperature increase to not more than 2 degrees Celsius above pre-industrial level. EU action against climate change was translated into a GHG emission reduction target of 20% compared to 1990 as adopted in the 2020 Climate and Energy Package. For the period 2021-2030, the GHG emission reduction target of at least 40% domestic reductions compared to 1990 by 2030, ensures the EU is on the path of low-carbon transition to emission reductions of 80-95% by 2050.

5.2.Specific policy objective

The specific policy objective is to align the EU ETS architecture with the 2030 emission reduction target and refine and improve the EU ETS post-2020 framework. This needs to be achieved in the light of the lessons learnt in a context where:

Fully comparable climate policy measures may not yet be undertaken by all other major economies;

Reinforced research, development and innovation efforts should take place in order to maintain Europe's industrial base and competence, and support the 2030 climate and energy framework as well as the long-term goal of a low-carbon economy;

Reinforced investment efforts should take place in order to modernise the energy system, and support the 2030 climate and energy framework;

Experience gathered during the first years of phase 3 (from 2013) suggests that there is still potential to reinforce efficiency of the system.

These specific objectives for the revision of the ETS Directive are further developed in the operational objectives of each of the chapters and sections. The operational objectives are tailored to the specifics of each cluster of issues, but also contribute to the achievement of the specific and general policy objectives of the ETS revision.

5.3.Consistency with other policies and objectives

As outlined in the 2030 impact assessment, the policy aim of the 2030 framework was to set climate and energy targets up to 2030, which are consistent and mutually reinforcing. 39  A well-functioning EU ETS is a key instrument to achieve the GHG emission reduction target and cornerstone of Europe's climate policy, consistent also with emission reduction efforts in non-ETS sectors. In line with the Energy Union strategy 40 , by putting a price on carbon at EU level the EU ETS reinforces the functioning of the internal energy market and stimulates the uptake of renewables and other low-carbon and energy-efficient technologies. Through low carbon funding mechanisms, it financially supports power sector, industry and Member states in the low-carbon transition.

In general, consistency of a revised EU ETS with any related policies is addressed in this impact assessment when discussing specific issues, such as for instance a forward-looking approach to CCS for the power and industrial sectors and the Innovation Fund.

There is a need to continue driving progress towards a low-carbon economy, as this ensures competitive and affordable energy for all consumers, creates new opportunities for growth and jobs and provides greater security of energy supplies and reduced import dependence for the European Union as a whole. This ETS revision initiative is coherent with these objectives in the field of climate and energy policy.

In terms of international competitiveness, previous Commission analyses of energy prices and costs have shown that there has been little impact on the EU's relative competitiveness which could be directly attributed to the ETS carbon price in the context of energy prices, although in the future the carbon price is assumed to be reflected in electricity retail prices 41 . However, as long as there are no comparable efforts undertaken in other major economies, policy measures (including a system of free allocation of allowances) are appropriate after 2020 to address the competitiveness of Europe's energy-intensive industries.

6.Implementation of the EU's GHG emission reduction target

6.1.Linear reduction factor

The outcome of the ETS in terms of emissions is determined by its cap on the total number of allowances. According to the ETS Directive and the present target of -21% by 2020 compared to 2005, the ETS cap for stationary installations declines linearly, by an annual amount equal to 1.74% of the average annual allocation during phase 2 (2008-2012), referred to as the linear reduction factor. Setting a cap at the 2030 emission level of -43% compared to 2005 requires a change in the linear reduction factor from 2021 onwards. This change is also needed in line with the EU's longer-term GHG emission reductions objectives. According to the analysis in the impact assessment accompanying the 2030 framework 42 , a revised linear reduction factor of 2.2% from 2021 onwards is required to ensure coherence with a 2030 cap equal to 43% emission reductions. The change from 1.74% to 2.2% reduces the supply of allowances by around 556 million in the period 2021-2030.

The European Parliament has called for legislation to be proposed at the earliest appropriate date with a view to adjusting the 1.74% annual linear reduction requirement so as to meet the requirements of the 2050 emission reduction target of 80 to 95% emission reductions. The European Council also explicitly endorsed the linear reduction factor of 2.2%. As the Commission proposal for the ETS revision has to achieve the objective of emission reductions in the ETS sectors of 43%, meeting the specific requirements of the European Council conclusions, no diverging policy options for the linear reduction factor post-2020 could be developed.

The impact assessment for the 2030 framework provides a comprehensive analysis of the impacts of emission reductions in the ETS of 43% by 2030 compared to 2005, and hence of the required linear reduction factor of 2.2% 43 .

Although the Market Stability Reserve was not explicitly included in the modelling work for the impact assessment for the 2030 framework, its effects can be considered as reflected: in the analysis the emission reductions required to reach the 40% GHG target in 2030 and the 80% in 2050 and the respective carbon prices were determined so as to achieve these emission reductions cost-efficiently and assuming rational behaviour 44 . In reality, however, businesses seem to base their abatement decisions on a shorter outlook for their industry than the 2030 reduction target would imply. 45 This is likely to mean that the current large supply-demand imbalance in the ETS reduces the incentives for low-carbon investment and thereby negatively affects the cost-efficiency of the system and of the achievement of EU emission reduction goals. Hence, by simulating the achievement of the 40% target, the economic model actually acts as capturing in a simplified way the expected impact of the Market Stability Reserve of addressing the surplus, increasing the confidence of market participants, and in turn delivering a meaningful price on carbon emissions and ensuring the emissions evolve in line with the cap by stimulating cost-efficient emission reductions. 

Concerning changes in the underlying assumptions in the work for the impact assessment for the 2030 framework, notably the recently lower oil prices, it can be concluded that this price drop is not expected to lead to any major changes in the modelling results, even if it is sustained. It has had no major impact on the carbon price because of the reduced supply due to back-loading and the increasing expectation of the Market Stability Reserve being established. According to the International Energy Agency, the recent developments have made non-OPEC production more responsive to price swings than previously, which would likely set the stage for a relatively swift recovery 46 . In addition, as relatively little oil is used as a fuel by installations in the ETS, the direct impact of changes in the oil price to the carbon price between 2021 and 2030 are expected to be limited, and a separate sensitivity analysis would not affect the outcomes for this impact assessment.

6.2.Auction share

In phase 2 (2008-2012), the overall cap that limits the amount of emissions and thereby sets the ambition level of the ETS was equal to the sum of national caps determined by the Member States in their National Allocation Plans (NAPs). These national plans determined different allocations at sector and installation level among the Member States. Allocating allowances on the basis of historical emissions (grandfathering) was the general rule and the auctioning of allowances the exception, leading to non-optimal investment and undesired distributional effects. As part of the EU's climate and energy package for 2020, it was thus agreed that as of 2013 a single EU-wide cap would be set and that auctioning would become the rule with transitional declining free allocation of allowances.

For phase 3, the EU-wide total cap 47  amounted to 2.084 billion allowances in 2013 and this annually decreases by the linear reduction factor of 1.74%.

The total cap is divided into a part that is made available to installations for free and a part that is auctioned. While the rule is that everything that is not allocated for free is auctioned and the volume of auctioned allowances increases over time, the maximum amount for free allocation is currently a fixed share of the total cap 48 . Setting a maximum to the free allocation constitutes a backstop to ensure long-term environmental integrity of the system, effectively implementing the polluter-pays-principle, while recognising the need for maintaining the international competitiveness of industrial sectors exposed to the risk of carbon leakage.

In 2013, the maximum amount available for free allocation was around 809 million allowances. 49 The determination of the maximum amount for free allocation has been a lengthy process (involving the Commission, Member States' authorities and companies included in the ETS), in particular due to the amount and nature of the necessary data. The wider public generally perceived the procedures as complex and insufficiently transparent. 

Over the period 2013 to 2020, free allocation is provided on the basis of EU-wide harmonised rules and product benchmarks. Because the aggregate amount of gross free allocation 50 calculated by Member States on the basis of these rules exceeded the maximum amount of free allocation available to industry, the allocation for all installations over the period up to 2020 is reduced by the same proportion through the application of the cross-sectoral correction factor (Article 10a(5) of the ETS Directive) (for more details, see section 7.1 ).

As an exception to the general rule that electricity generators should not receive any free allocation, additional allocation is provided in case these installations produce heat. Since this allocation benefits the power generation sector, this free allocation is not part of the maximum amount for free allocation to industry, but of the quantity of allowances auctioned by Member States which is reduced accordingly.

Additional free allocation is made available from the new entrants' reserve (NER) for newly built plants or in case installations extend their capacity by more than 10%. The new entrants' reserve is constituted from 5% of the total cap, amounting to a total of around 780 million allowances for the period 2013 to 2020. 300 million of the allowances in the new entrants' reserve have, however, been earmarked and used to support carbon capture and storage (CCS) or innovative renewable energy projects under the NER300 facility.

Since, according to the ETS Directive, everything that is not allocated for free is auctioned, some allowances are foreseen to be added to the auction volume at the end of phase 3 in 2020. This concerns in particular, allowances that remain unused in the new entrants' reserve and allowances that are not handed out to installations because they stop operations (closures) or reduce their production (partial cessations) will be auctioned at the end of the period. In the context of the legislative discussions on the Market Stability Reserve, the co-legislators decided that these allowances (so called "unallocated allowances") should be transferred into the Market Stability Reserve in 2020 to avoid that auctioning them would create another supply peak adversely affecting the market balance in 2020. The ETS revision should also analyse whether these unallocated allowances can be used inter alia for innovation and to address the risk of carbon leakage. This analysis is provided in Annex 15 .

Certain Member States have the option to provide free allocation to the power sector in return for investments modernising power generation. This option constitutes a derogation from the general principle that no free allowances are available to installations generating electricity. However, considering the investment needs in the power sector in certain Member States, the ETS Directive foresees this derogation. While this amount is qualified as free allocation, the allowances given to the power companies are provided from the Member States' auction volumes if, and to the extent, Member States make use of this option. If not used to the full extent, the allowances are auctioned on behalf of the Member State concerned. For more details, see section 8.3 .

Figure 2: Structure of the total quantity of allowances in phase 3

Figure 2 illustrates the auction share over phase 3 (2013-2020). It translates the starting point of the ETS Directive, according to which, in principle, all allowances should be auctioned, but free allocation is granted transitionally. Over phase 3, 39% of the total quantity of allowances available will be allocated for free to industry and electricity-generating installations for the heat they produce. This share of free allocation will be further increased due to allocations to new entrants until 2020. Today, it is not yet known how much of the NER will be used in the coming years. However, up until January 2015, 247 installations have received allocation from the NER, which will amount to 69.6 million allowances to be allocated from the NER by 2020, representing 14% of the total of 480.2 million allowances. Based on these trends, it is thus expected that no more than 2% of the cap may additionally be allocated for free from the NER. In total, the free allocation over phase 3 is thus expected to be around 41% of the total amount.

To the same extent that the allocations from the NER are not yet known today, it will also only be known at the end of the phase how many installations that currently receive an allocation will stop operation or reduce their capacity or production. Where installations no longer receive free allowances, the legal default is that these allowances will be auctioned on behalf of the Member States. In 2013 and 2014, the allocations to around 1100 installations were revised downwards by 85.7 million allowances. In general, the uptake from the NER and the return from closures and reductions are expected to be within the same order of magnitude 51 , so that another 2% of total allowances are expected to remain within the auction share.

Another some 2% of the total cap is used to fund the NER300 programme for CCS and innovative renewables projects.

As explained above, free allowances provided to the power sector in return for investments modernising power generation are deducted from or added to the auction share of the relevant Member State and, as such, remain part of the auctioned volumes even though they are given for free to the energy sector 52 .

Taking into account these different elements, the average auction share over phase 3 amounts to 57%.

For phase 4, these elements should not fundamentally change. As was the case for the NER300 programme, 400 million allowances should be made available for innovation support. Free allocation provided to the power sector in return for investments modernising power generation should continue and allowances used for this purpose would remain part of the auction volumes which individual Member States can decide to allocate for free in return for investments. In line with the starting point in the ETS Directive, that in principle all allowances should be auctioned, the allowances that according to the European Council should be auctioned for a new fund for the modernisation of the energy systems in certain low-income Member States are part of the auction share in phase 4 53 .

The "backloading", i.e. postponing auction volumes into the latter part of phase 3, and the functioning of the Market Stability Reserve to address the structural surplus of allowances in the EU carbon market both relate to the amount of allowances auctioned by Member States. For example, in case of the Market Stability Reserve, auction volumes will be reduced when allowances are transferred into the reserve and increased at the time allowances are released from the reserve. Both these mechanisms have a neutral effect on the overall auction share. 54  

The European Council agreed that the share of allowances to be auctioned under the EU ETS post-2020 should not be reduced 55 . This principle of not reducing the auction share was an important and integral part of the agreement to which Member States attach particular importance. Any change to the auction share would have distributional implications and adversely affect the balance of the European Council agreement 56 , and have implications on the low-carbon funding mechanisms, in particular the free allocation to power, and on the re-distribution for solidarity purposes of allowances to be auctioned.  

The system post-2020 builds on phase 3. The shares of auctioned allowances between Member States are already set in the Directive and, to meet the specific requirement of the European Council conclusions, it is appropriate to express the auction share as a percentage figure in the legislation. The environmental integrity of the system guaranteed by the cap would remain entirely preserved, while providing a percentage figure for the auction share in the legislation would have considerable positive impacts on transparency, predictability and the functioning of the carbon market. It would enhance planning certainty of investment decisions and transparency for market participants inside and outside the system, as well as for the wider public. It would render the system simpler, more transparent, more easily understandable and thus positively impact the confidence in the EU ETS.

7.Free allocation and addressing the risk of carbon leakage

The main policy tool to address the risk of carbon leakage is the provision of free allowances to industry. Free allowances lower the effective carbon cost for industry as otherwise allowances would have to be purchased on the carbon market. The total amount of allowances available to be handed out for free to industry for the period under assessment (2021 to 2030) is in the order of 6.3 billion allowances. The value of these allowances depends on the market price at the time and could be in the order of €160 billion.

If not handed out for free to industry, these allowances could be auctioned by Member States and provide them with revenues. Free allowances are thus a public resource. This section of the impact assessment covers different options on how to distribute this public resource in a way that gives optimal impact in terms of effectively addressing the potential risk of carbon leakage.

7.1.Problem definition

Free allocation of allowances to industry is designed to address the potential risk of carbon leakage (increase in greenhouse gas emissions in third countries where industry is not subject to comparable carbon constraints) until an international agreement is concluded or comparable climate policy measures are undertaken by other major economies. This gives the justification to deviate from the general principles of EU environmental policy, i.e. the "polluter pays principle" and "internalisation of external cost" principle, meaning that the costs of measures to deal with pollution should be borne by the polluter and that all costs associated with the protection of the environment should be included in the companies' production costs 57 .

Since carbon leakage could lead to an increase in worldwide emissions, thereby undermining the EU climate policy objective, free allocation was chosen as a transitional measure to address this risk.

In the first two phases of the ETS (2005-2007 and 2008-2012), free allocation was based on historic emissions and decided nationally. Basing the allocation on historic emissions implied that installations that emitted relatively more received more free allowances, compared to the more efficient ones.

The revision of the ETS Directive ensured that in the third phase (2013-2020), the approach for allocation of allowances was harmonised 58   across the EU to ensure a level-playing field across the internal market and to address concerns about potential market distortion. 

From 2013, the amount of free allowances that each installation receives is determined mainly by performance benchmarks instead of historic emissions. Generally speaking, benchmarks represent a value in CO2 emissions per tonne of product, reflecting the greenhouse gas emission performance of the best installations in the EU producing a specific product. Allocation to each industrial installation is calculated by multiplying the benchmark for the product it makes with its production data (for more details see Box 4 ).

For allocation purposes, each industrial sector is classified according to its exposure to the risk of carbon leakage that is assessed combining two parameters: trade intensity (imports and exports) and carbon intensity (CO2 emissions per tonne of product). Those sectors with a high combination of these two parameters are in general deemed as more impacted by the EU ETS, and therefore get a higher share of free allowances. They are put on the carbon leakage list and as such receive 100% of the amount of free allocation determined according to the benchmark-based methodology. The percentage of free allocation for sectors not on the carbon leakage list decreases in a linear manner from 80% in 2013 to 30% in 2020.

In line with the EU's climate policy to reduce GHG emissions, the ETS cap (maximum amount of allowances available) decreases gradually over time. The total amount of allowances available for industry is a share of the total cap and therefore decreases correspondingly. For more details see chapter  6 .

According to the current legal provisions, higher free allocation for sectors deemed to be exposed to the risk of carbon leakage ends 59  in 2020. Thereafter free allocation for all industrial sectors would correspond to 30% of the amount determined based on the benchmarks.

A further element in reducing costs for industries affected by the ETS is that installations in certain sectors can be compensated by Member States for higher electricity prices due to the ETS. These so called "indirect costs" are due to the power sector passing through the costs of the allowances (needed to cover their emissions) to their customers. This compensation is provided through state aid schemes determined at national level under the relevant State Aid Guidelines 60 .

The ETS is a flexible tool as it allows companies to choose the most cost-effective options to contribute to the overall emission reduction target, either by purchasing allowances on the market, or investing in less-carbon intensive technologies.

A crucial input to the work on revising the EU ETS came from the European Council, which, in its Conclusions of October 2014 61 gave strategic guidance on the 2030 climate and energy framework in general, but also on the design of the ETS post-2020. As long as no comparable efforts are undertaken by other international partners, the European Council underlined the need to continue free allocation to industry after 2020. This is meant to ensure appropriate level of support for sectors at risk of losing international competitiveness due to the carbon costs when operating in the EU. It can be noted that industry stakeholders also hold the view that free allocation is an adequate instrument. 62  

The European Council also stated that the most efficient installations in sectors should face no undue carbon costs, that there should be no windfall profits due to free allocation, and that existing measures to compensate for indirect emissions should remain also in phase 4.

Considering that the amount of free allowances is limited and will continue to decline in line with the necessary emission reductions, the future system to address the risk of carbon leakage needs to be better targeted. The total amount of allowances available for free allocation in phase 3 (2013-2020) is some 6.6 billion allowances, amounting to some € 50 billion (based on current carbon price), and the total amount available for free allocation in phase 4 (2021-30) is some 6.3 billion allowances, amounting to as much as €160 billion.

The future system also needs to ensure that incentives for industry to innovate will be fully preserved and windfall profits (i.e. situation when industry receive allowances to cover carbon costs, even though they are able to pass on these costs to their consumers) avoided, without reducing the share of allowances to be auctioned.

The problem to be addressed in this section concerns the future rules for free allocation to industry, and notably based on the principles defined by the European Council to establish how to optimally allocate the roughly 6.3 billion free allowances available.

7.1.1.Underlying drivers of the problem

The starting point for the analysis is the existing system.

Some general observations can be made. The harmonised rules applicable as of phase 3 represent an improvement compared to the previous situation. The harmonised approach is functioning well (an OECD study finds that the EU ETS has stimulated substantial emissions abatement of up to 28% compared to business-as-usual, while, at the same time, not causing disadvantages for the competitive position of the EU ETS firms 63 ) and the risk of carbon leakage seems to be properly addressed, while emission reduction incentives are also preserved 64 . A study concluded in 2013 65 that no conclusive evidence of carbon leakage occurrence can be found. 

However, when determining the amount of free allowances to be handed out each year for phase 3 (2013-2020) the amount calculated based on the agreed rules exceeded the amount available. Already in 2013, the demand for allowances ("gross free allocation", i.e. the amount of free allocation determined by multiplying the benchmark and production before the application of the carbon leakage and the possible correction factor) exceeded the amount available by about 6%. This triggered, as foreseen in the ETS Directive, the application of a uniform cross-sectoral correction factor 66 . Since the amount of allowances available decreases each year, and since allocation for phase 3 was determined for the entire period 2013-20, the cross-sectoral correction factor increases each year to 2020, when it will be 18%.

While the cross-sectoral correction factor is effective in ensuring that free allocation stays within the overall allocation limit and the amount of allowances available for auctioning and delivering fiscal revenue to Member States remains predictable, it does not differentiate among sectors: i.e. does not account for differences in terms of ability to decarbonise over time, exposure to the risk of carbon leakage, the pass through of costs in product prices, etc. Given the shrinking amount of allowances it is self-evident that unchanged carbon leakage rules would result in an increasing correction factor beyond 2020 reaching potentially a level of 35 % in 2030.

It should be noted that the cross-sectoral correction factor does not determine which sectors receive a certain level of free allocation: its purpose is only to ensure that the overall limit is respected and Member States auctioning revenues guaranteed.

The correction factor is subject to strong criticism by industry stakeholders 67 , which constitute the majority of the stakeholders replying to the consultation, mainly because it applies equally to all sectors and therefore the system does not guarantee that the most efficient installations in each sector do not face undue carbon costs. It has also has been criticised for having been determined too shortly before the start of phase 3.

Since many industry stakeholders considered that the cross-sectoral correction factor represents high unpredictability in the system, one element of this impact assessment is to analyse different ways that the current allocation system can be improved to ensure a fair and efficient allocation of allowances, while avoiding or minimising the likelihood that a cross-sectoral correction factor would apply in phase 4.

For the analysis, it is necessary to look at the different elements that determine free allocation. These are:

the updating of the benchmark values;

the classification of sectors, based on market situation including the possibility to pass through carbon costs into product prices;

the production data used.

On these three elements in the allocation system, the following can be noted:

1) The starting point for determining the benchmark values in phase 3, in accordance with the ETS Directive, was the average emission performance of the 10% most efficient installations in a sector in the years 2007-08. Since the benchmark values were set based on data from 2007-08, any technological progress since then is not taken into account. The ambition level of the benchmark values therefore erodes over time and in 2030 benchmarks would reflect the state of technology of more than two decades ago.

2) Currently based on criteria defined in the ETS Directive 68 , almost all industrial sectors, i.e. those responsible for more than 97% of industrial emissions under the ETS are on the 'carbon leakage list'. The system is thus not targeted. 

3) The third element in the allocation formula is the production data used to multiply the benchmark values. According to the current implementing rules, operators were allowed to choose between two historic production baseline periods (either 2005 – 2008 or 2009 – 2010). This led to a significantly higher demand for free allocation than if the same baseline period had been used for all. Furthermore, in general these years had particularly high production levels for most installations, but due to the economic crisis and slow recovery, actual production levels in 2013 and 2014 were significantly lower in certain sectors compared to the baseline production. This latter fact, however, is not always reflected in lower amounts of free allocation, resulting in a situation where some installations with low production levels receive allocation that exceeds the amount needed to cover their emissions.

All the above features must thus be assessed in order to improve the existing system with a view to ensuring the optimal allocation of allowances and targeting the measures against carbon leakage to those sectors most exposed. A more general goal is to ensure predictability for companies covered by the system (which is a key request from a considerable number of industry stakeholders).

4) A further element important to consider is how new and/or growing installations i.e. "new entrants" will be treated in the system post-2020. The current system provides specific rules and a specific amount of allowances is set aside (new entrants' reserve) for this purpose.

5) Due to requests from the electricity intensive sectors, which argue that the current system of addressing indirect carbon costs passed through in electricity prices via state aid causes distortions in the internal market, the issue of compensation for these costs will also be analysed.

Since the correction factor is a less efficient way of directing free allowances to those sectors most exposed to the risk of carbon leakage, an objective of this impact assessment is to analyse how a more focused system, resulting in a reduced likelihood of triggering a significant correction factor post-2020, should be designed.

7.2.Operational policy objectives

As mentioned above, the overall objective of the free allocation system is to address the risk of carbon leakage by providing appropriate levels of support to sectors at risk of losing international competitiveness, as long as no comparable efforts are undertaken in other major economies. In line with the European Council Conclusions of October 2014 69 , this general objective is operationalized in the following operational objectives:

Reflect technological progress in industry sectors;

Fully preserve incentives for industry to innovate;

Most efficient installations do not face undue carbon costs leading to carbon leakage;

Better alignment with production levels;

Avoid windfall profits;

No increased administrative complexity.

The options for each element to be assessed will be screened preliminarily for the effectiveness, efficiency and feasibility in achieving the operational objectives. Not increasing administrative complexity reflects the efficiency and feasibility criteria, while all the other operational objectives relate to the effectiveness. Some operational objectives are particularly relevant for some aspects of the free allocation rules, but not for others, so when screening and comparing options, preference will be given to the most relevant operational objective(s).

The degree of fulfilling an objective is indicated with pluses and minuses. Their meanings are: "--" much worse than baseline, "-" worse than baseline, 0 equal to baseline, "+" better than baseline, "++" much better than baseline.

As mentioned above, the likelihood of triggering a correction factor post-2020 will be analysed for the option packages as a whole to assess the cumulative impacts of the elements composing them.

It should be underlined that the development and assessment of options for the future system of free allocation and measures addressing the risk of carbon leakage are based on the assumption that no comparable climate policy measures are undertaken by other major economies.

However, there are a number of countries and regions that have or are intending to develop different emission trading systems (e.g. China, South Korea, California, Switzerland, some regions in Canada). Their approaches, design features and levels of ambition are heterogeneous, making it difficult to quantify the impacts. The policy options therefore assume that third countries fulfil Copenhagen/Cancun pledges, but there is no assumption on any further significant climate action in third countries.

Should an international climate agreement be concluded, it will be necessary to first analyse the concrete measures taken as a consequence and to determine their implications for the EU climate policy in general, and impact on the competitiveness of EU industry in particular, before any revision of the existing/proposed rules.

7.3.Policy options for free allocation and other carbon leakage measures

As explained above, the carbon leakage and free allocation system contains a number of elements. The level of free allocation per installation is the mathematical product of a benchmark value, a production level value, a carbon leakage factor and, if needed, a correction factor (see Box 4 ). The guiding principles adopted by the European Council address these elements and also compensation for indirect costs. Depending on the policy choices made, there will be differences in terms of the degree of realising individual operational objectives. The choices may also have different distributional effects between industrial sectors.

This section summarises the underlying rationale and range of policy options which have been used to design option packages. A more detailed discussion of the options can be found in Annex 5 .

7.3.1.Benchmark values

The first element in the allocation formula is the benchmark value. It determines tonnes of carbon emissions per tonne of product and is independent of the mode of production in terms of technology, fuel, the size of an installation, or its geographical location.

The currently used benchmark values were determined based on performance data of the most carbon efficient installations throughout the EU in each sector based on real-life historic industrial production in the years 2007 and 2008.

The European Council has requested a regular review of the benchmark values in order to reflect technological progress. The benchmark values should thus be lowered in phase 4, leading ceteris paribus to reduced allocation. However, this reduced allocation reflects lower emissions thanks to the ongoing technological progress and does therefore not as such make the allocation system less generous.

There are various options how and how frequently to update the benchmark values. The full range of these options is presented in Annex 5.1 , and out of the screened options two are used in the option packages.

The baseline option is to update all benchmark values based on new data reported by operators of ETS installations on mandatory basis and to keep the benchmarks constant over phase 4 (2021-2030). This baseline option is also used in the "Baseline Bbis" and "Limited changes" option packages.

One alternative option is the single flat-rate update (option 1 from Annex 5.1 ), updating all benchmark values by the same percentage once before the start of phase 4 and keeping them constant in the whole phase 4. This option is used in the "Simple" option package.

The other option is the full data collection update (option 4 from the Annex 5.1 ), updating all benchmark values based on new full data collection, and revising them every five years, i.e. twice during phase 4. This option is used in the "Targeted" option package.

In summary, the analysis shows that single flat rate is administratively simple, but may not sufficiently capture technological differences across sectors. However, using multiple flat rates (e.g. high, medium, low) can address this downside. On the other hand, full data collection update is administratively highly complex, creates prolonged regulatory uncertainty for companies and compromises the incentives to innovate, as realised profits from lowering emissions are taken away from fast movers. For further details see Annex 5.1 .

7.3.2.Production level and adjustments

The second element in the allocation formula is the production level. The determination of the production value is important for the functioning of the system. Linked to this are the rules on how allocation can change over the phase, e.g. if production increases or decreases

Currently historic production data from 2005-2010 were used to determine allocation for the 8-year period 2013-2020. Each installation had thus certainty in 2012 how many allowances it would receive year by year until 2020.

From the point of view of an installation, the system is quite stable and predictable. The installation keeps its allocation unchanged until the production level is 50% below the historic level. If the installation invests in new capacity it can apply for additional allowances from the new entrants' reserve.

Many industrial stakeholders have commented on the current situation and proposed that allocation should be adjusted not only for capacity increases, but also for production increases.

These rules can have a considerable impact on the total amount of allowances to be handed out per year in different economic circumstances, i.e. during an economic slowdown or upturn. The rules therefore also have an impact on the likelihood for the need for the correction factor. Moreover, the rules have a large bearing on the degree of administrative complexity and costs.

There are various options to tackle production level and adjustments. The full range of these options is presented in Annex 5.3 , and out of the screened options two are used in the option packages.

The baseline option is to have the production levels determined once and keep the current rules for production changes. This baseline option is also used in the "Baseline Bbis" package.

Out of the screened options (see Annex 5.3 ), the ones used in the option packages are:

One alternative (option 1) has the historical production levels defined once and based on five years (2013-17) for the entire 10-year period. Significant production increases and decreases are addressed through symmetrical annual adjustments 70 , i.e. same thresholds for increased and decreased production. This option is used in the "Simple" option package. 

Another alternative (option 2) has the production levels defined for two five-year periods (2013-17, respectively 2018-22) 71 . Significant production increases and decreases are addressed through symmetrical annual adjustments i.e. same thresholds for increased and decreased production. This option is used in the "Limited changes" and "Targeted" option packages.

In both options, the increased allocation for increased production comes from the new entrants' reserve.

These options provide for a closer alignment to production levels building on lessons learnt from the current system and responding to requests from the European Council to this effect. A careful balance needs to be struck to avoid creating red tape and undermining incentives for further emissions reductions. For further details see Annex 5.3 .

7.3.3.Reserve for new entrants

A topic related to the previous aspect on production level and adjustments is the so-called new entrants' reserve. This is a reserve that in the current phase of the ETS is used to provide allowances to installations that are new or increase their capacity. As a result of the European Council's strategic guidance to continue free allocation, such a reserve is also needed for phase 4.

It is important to ensure that the size of the new entrants' reserve is sufficient to accommodate new demands during phase 4, but new installations are typically more carbon-efficient and therefore facilitate the gradual transition to a low-carbon economy.The analysis therefore assesses different options – guided by building on current practice while making the rules more flexible – for how to create this reserve and ensure that it will not be depleted.

The baseline option is to have a new entrants' reserve with a fixed amount of allowances (about 374 million allowances in phase 4). This baseline option is also used in the "Baseline Bbis".

Out of the screened options (see Annex 5.4 ), the one used in the option packages is the option (Option 1) with new entrants reserve set up with phase 4 or unallocated phase 3 allowances and replenished throughout 2021-2030 with unused phase 4 allowances (see Annex 14). This option is used for the "Simple", "Limited changes" and the "Targeted" packages.

7.3.4.Carbon leakage groups

The third element in the allocation formula is the carbon leakage factor. The ETS Directive currently categorises industrial sectors in two categories depending on the level of exposure to the risk of carbon leakage.

Installations in sectors and subsectors which are deemed to be exposed to a significant risk of carbon leakage are inscribed in the so-called 'carbon leakage list' (see Box 3 ) and are given free allocation at 100% of the quantity determined based on the benchmark values and production levels, i.e. the carbon leakage factor in the allocation formula is 1.

Installations in other industrial sectors not covered by the carbon leakage list receive a lower (annually decreasing from 80% to 30%) level of free allocation, i.e. the carbon leakage factor in the allocation formula is 0.8 in 2013 and shrinks to 0.3 in 2020.

In practice, there is very limited differentiation among industrial sectors since sectors accounting for more than 97 % of industrial ETS emissions are on the carbon leakage list (see Figure 3 ). Hence, practically all industrial sectors receive the same treatment irrespective of the degree of exposure to carbon leakage risks, including the ability to pass on costs.

Figure 3: Sectors on and off the carbon leakage list The chart indicates the share of sectors from industrial emissions

A more targeted approach to allocation whereby differences in terms of exposure to carbon leakage risk translate into differentiated allocation levels could have made it possible to limit or even avoid the need to apply a cross-sectoral correction factor in phase 3. It is therefore pertinent to explore the possibility of enhanced differentiation between industrial sectors.

The baseline option considers two carbon leakage groups based on the current trade and emission intensity criteria. It is used in the "Baseline B" package.

Out of the screened options (see Annex 5.3 ), the ones used in the option packages are:

A slight variation to the baseline option (option 2) with somewhat modified criteria is used in the "Baseline Bbis" package.

An alternative is an approach with no groups and a uniform carbon leakage factor (option 1), i.e. no differentiation among sectors and activities. This option is used for the "Simple" option package.

Other alternative options (options 3 and 4) with four groups ('Very high', 'High', 'Medium' and 'Low' level of carbon leakage risk) are also based on the criteria of emission intensity and trade intensity. These options are used for the "Limited changes" and "Targeted" option packages 72 .

The options vary according to the differentiation they bring among sectors – from none (all sectors in one group), to two groups (exposed / not exposed) and four groups (more detailed differentiation of the level of risk exposure).

The 'Baseline approaches' and Option 1 fail to sufficiently differentiate between the different levels of carbon leakage and lead to a high correction factor that cuts free allowances for all. The alternative approaches should lead to a more targeted distribution of allowances across sectors and provide a better outcome for the most exposed sectors.

See Annex 5.2  for further details.

7.3.5.Cross-sectoral correction factor

A fourth, potential, element in the allocation formula is the cross-sectoral correction factor which is a stopgap to ensure that the available budget for free allocation is not over-subscribed. No options are developed for this factor as the likelihood of this factor being necessary and its magnitude depends on the policy design of the first three factors. The avoidance of having to apply this factor is an important consideration in assessing different options and option packages.

7.3.6.Compensation for indirect carbon costs

Indirect carbon costs occur in many sectors across the economy and are a natural consequence of pricing carbon emissions. In the political debate, the term is used to describe the costs of carbon emissions related to producing electricity, which are passed through to industrial consumers of electricity. As electricity producers covered by the ETS do not receive free allowances to cover their emissions (and therefore have to buy needed allowances on the market) and given the situation of the electricity market (i.e. almost no imports from outside the EU and still a high proportion of fossil fuel used), electricity producers typically pass through these carbon costs in their prices to industrial customers and households alike.

As a safeguard against potential carbon leakage, the ETS Directive gives Member States the possibility to compensate certain electro-intensive industries for these costs, subject to state aid control 73 .

Some industries with high electricity intensity have criticised the fact that compensation of indirect costs is at the discretion of individual Member States 74 . They notably highlight the risk of distortion of competition due to a lack of harmonisation. As a consequence, electricity intensive industry stakeholders have asked that compensation of indirect costs is done through an EU-wide harmonised system. 75

The impact assessment therefore analyses different options of providing such compensation, whereby different aspects have to be taken into account such as whether compensation is optional or mandatory and how it is resourced, which could be national state budgets, auction revenues or free allowances.

Based on the screening, the following two options are used in the option packages:

The baseline option is to continue with the optional compensation by Member States with the key features for compensation determined at EU level, but the decision to grant compensation is discretional and depends on the respective Member State and is subject to State aid control. This option is used in the option packages "Baseline B", "Baseline Bbis", "Simple" and "Limited changes".

The other option is a mandatory Union-wide compensation scheme, financed by using national auctioning revenues. The compensation would be triggered when the carbon price exceeds a certain value, and when this situation occurs, a minimum amount of compensation is to be given by all Member States. The system would be financed by each Member State using its national auctioning revenues. Compensation would, like in baseline, be limited to certain sectors and limited to a share of the estimated carbon costs. This option is used in the "Targeted" option package.

See Annex 5.5  for further details. The options for indirect cost compensations differ in terms of the nature of compensation: optional or mandatory for Member States to provide it.The analysis indicates that more harmonised arrangements for indirect cost compensation has benefits but care is needed to avoid red tape and lock-in in emission intensive production methods.

7.4.Option packages

The options for all relevant elements presented above are combined into "Option packages" in order to facilitate the assessment of impacts. The option packages contain the elements needed to determine the amount of free allocation (see Box 4 : The basics of the free allocation system ) and some other relevant aspects, such as indirect cost compensation and new entrants' reserve.

As already mentioned, some important provisions that are currently in place will expire in 2020 (e.g. increased free allocation for sectors deemed to be exposed to a significant risk of carbon leakage). As such, the "legal baseline" option (i.e. the provisions of the ETS Directive with no change whatsoever) - 'Baseline A' does not correspond to the key requirements of the European Council, therefore this option is not further pursued.

The more relevant baseline ('Baseline B') assumes the current rules are prolonged to the next ETS phase (i.e. amendments to the ETS Directive required). The option packages will therefore be assessed against the Baseline B in order to allow for easier comparisons between policy options.

Furthermore, an additional 'Baseline Bbis' package is developed. It is similar to the 'Baseline B', but with modified carbon leakage criteria, so the sectors with low emissions are excluded from the carbon leakage list. This will shorten the carbon leakage list by some 100 sectors, but will otherwise not have a more distinctive impact, as these sectors only emit about 2% of the industrial emissions in the ETS.

Based on the initial screening performed in previous sections, and aiming to ensure a manageable level of complexity, three option packages are proposed in addition to the Baselines. These option packages are developed based on the following guiding principles:

The "Simple" package is structured around the concept of minimum level of administrative burden and complexity. It should be noted that some of the operational objectives nevertheless require administrative efforts (e.g. some data collection is still needed, but this should be less extensive).

The "Targeted" package is based on policy options aiming to ensure that the sectors most exposed to the risk of carbon leakage do not face undue costs, while, at the same time, avoiding windfall profits. This would be achieved by providing an optimal level of free allocation (i.e. regularly updated benchmarks, 2 allocation decisions to reflect more recent production data, 4 carbon leakage groups) and mandatory compensation for indirect costs to installations.

The 'Simple' and 'Targeted' packages represent different ways to address the trade-offs between the guiding principles defined by the European Council, as most of the policy options used for 'Targeted' package require intensive data collection, thereby going contrary to the minimum administrative burden principle.

The third package, 'Limited changes', uses a more conservative approach in trying to achieve the operational objectives, while building upon the current rules. This package therefore entails only moderate changes compared to 'Baseline B', instead of more ambitious exercise presented in 'Simple' and 'Targeted' packages. The 'Limited changes' mainly concerns the number of allocation decisions (two instead of one per phase) and the number of carbon leakage groups (4 instead of 2).

The table below summarises the main features of the option packages.

 Table 1: Option packages for addressing the risk of carbon leakage

7.5.Analysis of impacts

In this section, the impacts of the option packages presented in section 7.4  are compared to 'Baseline B' 77  and assessed in terms of achieving operational objectives.

7.5.1.Direct and indirect effects

Free allocation addressing the risk of carbon leakage has a direct impact on industrial installations covered by the ETS and on the budgets of the Member States. The amount of free allowances to companies has an effect on their cash-flows and profit margins (as it decreases total carbon costs). It could be argued that free allocation also indirectly affects a company's clients and final consumers, depending on the ability of each actor in the supply chain to pass the carbon costs downstream. Due to multiple variables and uncertainties, it is not possible to quantitatively assess these indirect effects.

The analysis will therefore focus on the analysis of the economic impacts (competitiveness considerations for industry), environmental impacts, social impacts and assessment of administrative complexity.

7.5.2.Environmental impacts

The environmental outcome of the ETS is determined by its overall cap and the EU climate ambition. The European Council has decided to reduce GHG emissions in sectors covered by the ETS by 43% until 2030. Consequently, the environmental outcome of the ETS is determined by the EU ambition in general terms and by the overall cap in particular. This means that a limit is set on emissions allowed, corresponding to allowances, to ensure the reduction foreseen is achieved.

When it comes to the international context, EU policies may impact on third country policies on climate and other connected areas. In view of the upcoming international negotiations and meetings (Conference of the Parties to the UNFCCC set to take place at the end of 2015 in Paris) the EU has come forward with an Intended Nationally Determined Contribution (INDC)  78  in the form of a binding, economic-wide emission reduction target.

The target includes the sectors benefitting from free allocation. However, the specific rules for the distribution of free allowances will not impact the policy choices of third countries. The general setup of the EU ETS and the benchmarking system in particular, has served as an input for the design of other emissions trading systems worldwide, but the positive impact cannot be quantified or considered in comparing option packages.

Another aspect to consider is the impact of the ETS on air quality. In addition to carbon dioxide emissions, many installations covered by the EU ETS also generate a significant amount of other air pollutants (e.g. NOX, SOX and particulate matter). The reduction of carbon dioxide emissions imposed by the ETS and the cap will lead to a reduction of the other air pollutants, but the specific choices for the distribution of free allocation will not make a difference in this sense: the reduction will happen regardless with the corresponding positive effect on air quality. The overall environmental goal remains unchanged irrespective of the exact methodology to allocate allowances for free.

In terms of environmental impacts, the risk of carbon leakage has also been considered since it implies a potential increase in global emissions. Nevertheless, this is a risk that has been considered at every step of this analysis and all policy packages proposed include safeguard measures against it, making significant carbon leakage unlikely. As such, the potential impact of the packages at global level is considered to be minimal and beyond the scope of this impact assessment.

7.5.3.Economic impacts

The analysis carried out in the context of the 2030 climate and energy framework included detailed modelling of economic impacts, including sectoral impacts 79 , which concluded that free allocation of allowances would be an effective means of reducing the risk of carbon leakage and preserving the output of the concerned industries. It also notes that understanding of different levels of cost pass-through is needed in order to elaborate carbon leakage measures that provide adequate safeguards, but avoid over-compensation of industry for costs recovered through the market.

The order of magnitude of the economic impacts of free allocation modalities is estimated to be about 6.3 billion allowances, which could amount to some €160 billion 80 . These public resources need to be allocated in an optimal way, while ensuring that the polluter pays principle is not undermined. In the stakeholder consultation, NGOs emphasized that these public resources should be provided only when strictly necessary.

7.5.3.1.Quantified impacts on competitiveness

This section investigates the impact of free allocation options on the compliance costs at sector level and the possibilities to pass on these compliance costs to consumers 81 . It should be noted this analysis does not cover all aspects of competitiveness. In particular, a qualitative assessment on incentives for innovation – which will determine carbon costs in the long-term – will follow in section 8.1  below.

The purpose of the free allocation and carbon leakage rules is to address the costs which installations included in the ETS may be facing due to the ETS, potentially putting them at disadvantage vis-à-vis international competitors that do not face comparable climate policy costs. If installations in third countries would face similar costs, for instance as a result of the conclusion of an international binding agreement on climate change or the implementation of climate policies at national level, the carbon leakage risk would disappear. It should be noted that other countries may also use policy measures like free allocation. However, the more countries taking action against climate change, the less there is a need for individual countries to provide free allocation.

(a) Impact on compliance costs

The compliance costs depend on the carbon price, the level of free allocation, and the amount of emissions released during production.

The carbon price is not directly impacted by modalities for distribution of free allocation, as the total amount of allowances available (the cap) has been fixed in advance (see section 6.1  on the linear reduction factor) and is not affected by the modalities of allocation (auctioning or free allocation). The carbon price is rather influenced by other factors, such as the surplus of allowances in the system, the level of industrial production, the level of emissions, etc.

Therefore, the total carbon costs at macro level are determined by the 40 % overall GHG reduction target, and the specific 43% reduction target for the ETS. The effects of these targets have been analysed in the impact assessment for the 2030 climate and energy framework. Furthermore, it is important to recall that the European Council conclusions have also determined the total amount of available allowances for free. The difference between the option packages is therefore how this finite amount is distributed among industrial sectors. The higher the amount of free allocation, the lower the compliance costs for a sector.

The impacts of free allocation and carbon leakage options on the competitive position of ETS sectors have been analysed and compared to the outcome under 'Baseline B' 82 . Other factors affecting competitiveness, but not related to the EU ETS (e.g. global change of demand and trade patterns, labour costs, access to finance and capacity to invest etc.) are not analysed since they are independent from the ETS.

Following the call of the European Council conclusions "to provide appropriate levels of support for sectors at risk of losing international competitiveness", the carbon leakage rules should achieve a targeted allocation of free allowances across sectors such to avoid undue costs for the most efficient installations.

The 'Baseline B' package – based on the current ETS rules – fails to respond to this call because a mere continuation of the current approach – based on two carbon leakage groups – would lead to an average correction factor of some 10% to 20% over the next decade 83 . This means that free allocations across all sectors would be reduced by this percentage irrespective of their carbon leakage risk.

As shown in Table 2 , 'Baseline Bbis' would reduce by two thirds the number of sectors eligible for increased level of free allocation as a result of being on the carbon leakage list, but still lead to similar levels of free allocation as 'Baseline B'. The free allocation would only be reduced by about 2% since most of the sectors removed from the list emit very little. Consequently, this would entail some redistribution from the sectors removed from the carbon leakage list to those on the list but not significantly reduce the correction factor. The "Simple" package is not included in Table 2 since all sectors would be in one and the same group. 

Table 2: Estimated number of sectors and distribution of free allowances under 'Baseline', 'limited changes' and 'targeted 'options

The 'Simple' option package 84 offers a similar level of free allocation as 'Baseline B' and 'Baseline Bbis' for the industrial sectors that are currently on the carbon leakage list. There are some limited differences among sectors in terms of level of free allocation level compared to 'Baseline B', because updating the benchmark value by using a flat rate can have different effects for different sectors. This happens because of varying rates of emission intensity reductions linked to technological improvements and investment cycles.

At the same time, the 'Simple' option package would lead to significantly higher levels of free allocation than 'Baseline B' for those sectors that are currently not on the carbon leakage list, as they would move from the group with 30% free allocation to 90% free allocation.

Even though the coverage is reduced from 100% to 90% for most sectors, a correction factor of 5% to 10% would kick in and lead to an effective coverage of around 80% (comparable to levels under 'Baseline B').

The 'Limited changes' and the 'Targeted' options packages offer a more refined classification of sectors into four carbon leakage groups based on the same criteria – emission and carbon intensity – as in the 'Baseline' packages. Depending on the carbon leakage group (Very high, High, Medium or Low), the amount of free allowances can increase or decrease compared to 'Baseline B'. The more refined classification into four groups is estimated to minimise, if not avoid the correction factor.

Sectors in the 'Very high' group (100% free allocation) will receive more free allowances compared to the 'Baseline B' packages if the correction factor disappears.

For sectors in the 'High' carbon leakage group, the total amount of free allocation under these two option packages would be quite similar to allocation under 'Baseline B'. This would happen because the difference in carbon leakage factors 85 (80% instead of 100% under 'Baseline B') is estimated to be in the same order of magnitude as the average cross-sectoral correction factor estimated for 'Baseline B' over phase 4. So instead of a correction factor of about 20%, there will be a reduction of allocation of 20% due to the lower allocation rate (assuming the same level of benchmark updates under both options).

For sectors, which are in the group with 100% free allocation under the 'Baseline B' package but would now change to the 'Medium' and 'Low' groups, the level of free allocation would be lower ('Medium' group) or significantly lower ('Low' group) compared to 'Baseline B'.

The figure below summarises how the various carbon leakage groups impact the level of free allocation for the 15 largest sectors. All sectors are under current rules ('Baseline B') on the carbon leakage list and entitled to 100% free allocation. However, due to the correction factor of around 15-20% on average, they will effectively receive only around 80-85% of free allocation (to respect the overall constraint on the total amount of available allowances for free).

Figure 4: Estimated combined effect of carbon leakage groups and the correction factor under different option packages 86

(b) Compliance costs and possibilities for cost pass through

The compliance costs borne by sectors are ultimately dependent on their ability to pass through carbon costs to their customers. The ETS Directive already recognises this fact suggesting that the level of carbon leakage risk possibly faced by sectors depends on the extent to which it is possible for these sectors to pass through their costs without losing market share.

Although there is a general understanding that carbon-intensive sectors are able to pass through at least a part of the carbon costs, it remains to date difficult to quantify the exact rate of costs passed through per sectors or products.

It is generally recognised that some industrial sectors may face increasing competitive pressure on the international markets, potentially making it more challenging for them to sustain existing pass-through rates. At the same time, it is reasonable to assume that climate policies will become more significant in other countries over time, which could lead to increased, or at least stable, cost pass-through rates.

As the number of total allowances will further decrease over the next decade, it is possible that industries will face increased compliance costs after 2020 but this will ultimately depend on the ability to increase carbon efficiency in production, and the ability to pass on carbon costs, e.g. through more specialised products. Considering that a share of carbon costs is likely to be passed through, it can effectively limit carbon cost increases for industrial sectors. In some cases, additional carbon costs may be more than fully offset by increases in product prices. See Annex 7 for more detailed results for some of the biggest sectors in terms of emissions (steel, cement, refineries, chemicals, fertilisers, glass, paper and aluminium).

7.5.3.2.Administrative burden

Avoiding an increase of the administrative complexity is an operational objective relevant for all elements of the option packages. For the purposes of this impact assessment, the level of administrative complexity was quantified using the EU Standard Cost Model, complemented with qualitative expert judgment. In particular, the costs were estimated separately for the data collection needed for NIMs exercises 87 , benchmark update, annual adjustments for production level changes compared to baseline (capacity changes and cessations); and compensation for indirect carbon costs.

The underlying methodology and detailed calculations are presented in Annex 8.1 . In short, the option packages 'Limited changes' and 'Targeted' lead to higher costs compared to 'Baseline B' (€82 million and €110 million, respectively), reflecting a higher level of administrative complexity (due to additional data collection exercise). To note in this context, one full data collection is estimated to cost about €80 million. The annual adjustment for significant production changes is not expected to trigger substantial additional administrative costs as they replace activities currently required for 'Baseline B' (capacity changes and partial cessations). For that reason, the administrative costs of the ''Simple" package' and 'Baseline B' package are not expected to differ significantly. The 'Targeted' package shows the highest administrative complexity due to the mandatory financial support for indirect cost compensation.

7.5.4.Social impacts

7.5.4.1.Employment

The analysis of social impacts has been undertaken and is presented in detail in Annex 8.2 . It concluded that the estimated impacts across different option packages are limited in nature. In a more targeted allocation, the costs for the sectors that receive lower allocation will be higher, which a priori could lead to more employment losses for those sectors. But as explained above, since the reason for the lower allocation to some sectors would be the ability to pass on costs, this should limit the employment effects.

Total employment impact has been estimated by considering whether the additional costs are absorbed by the manufacturing sectors, or are being passed through in higher prices to the customers, resulting in decreasing sales (i.e. not absorbed by the sector).

For the option packages 'Simple', 'Limited changes' and 'Targeted', small positive impacts on employment compared to 'Baseline B' are expected in the order of magnitude of up to 5000 jobs, representing an increase of 0.1%.

An additional aspect on employment is the likelihood that an ambitious climate policy will generate a demand for low carbon technologies, and renewable energy. If industrial sectors covered by the ETS do not receive more free allocation than necessary to cover their real costs (that cannot be passed on), this should incentivize higher demand for such technology and services, leading to positive employment impacts. These effects have not ben quantified.

The employment impact of 'Baseline A' is slightly negative reflecting the additional cost for companies.

In the lower estimate case, where all absorbed costs are reflected in changed profits, the total employment impact will reflect the consequences of the costs passed through. In this case, the employment impacts are even more negligible.

7.5.4.2.Energy prices for households

The impacts on the energy prices for households that are supplied by district heating covered by the ETS were also assessed (Annex 8.2 ).

Overall, the estimated impacts of the 'Simple' option package compared to 'Baseline B' are reductions in heat prices of € 0.77/GJ (almost 3% of baseline price). For 'Limited changes' there will be no significant change, while for the 'Targeted' a small increase of 0.8% is projected, due to a reduction of the heat benchmark.

The 'Simple' option package results in the highest reduction due to the significantly more generous carbon leakage factor compared to 'Baseline B' which outweighs the effect of a reduced benchmark value. The limited cost reduction for 'Limited changes' results from the fact that no cross-sectoral correction factor is expected to be used for these options.

7.5.5.Comparing the options

In the following assessment step, in addition to the quantified impacts, the operational objectives as described in section 7.2 are used to compare the option packages.

7.5.5.1.Quantified environmental and social impacts

No significant differences have been identified between policy option packages in terms of environmental impacts because the required GHG emission reduction of 43% in 2030 (compared to 2005)s and air pollution are determined through the increased linear reduction factor of 2.2% (see section 6.1 ).

In terms of social impacts employment level and prices for district heating have been considered. Regarding impacts on the level of employment, no significant differences between policy option packages and 'Baseline B' have been identified. With the exception of the 'Simple' package, no significant impacts on prices for district heating have been identified. The 'Simple' package is expected to lead to lower prices in the order of magnitude of 3% of the baseline price. For the 'Targeted' package, a limited price increase of 0.8% is estimated.

In conclusion, for social impacts, overall, the differences between the options packages are rather small. The 'Simple' package could be considered as slightly better than 'Baseline B' due to the positive impact on district heating prices for households. 'Baseline Bbis' and the 'Targeted' packages could be regarded as less beneficial due to slightly higher district heating prices. The 'Limited changes' package shows no significant difference compared to 'Baseline B'.

7.5.5.2.Reflect technological progress in industry sectors

The 'Targeted' package ensures technological progress is reflected in the benchmark values more closely than other packages due to regular recalculation of all benchmarks based on collection of sector-specific data.

The 'Simple' package reflects technological progress by applying a flat-rate percentage to all benchmark values. At the same time, this simple approach based on a single flat-rate cannot account for differences between sectors in terms of their historic ability to reduce GHG emissions.

The option packages 'Limited changes', 'Baseline B' and 'Baseline Bbis' update benchmarks once prior to 2021 based on newly collected data. This allows for a reflection of technical progress made by then. However, technical developments after this update cannot be considered. Therefore, these packages can be ranked between 'Targeted' as the best package, and 'Simple'.

7.5.5.3.Fully preserve incentives for industry to innovate

More frequent benchmark updates based on full data collection (like in the 'Targeted' package) may lower innovation incentives as companies would fear that they could not earn the full return on their innovation efforts. To the contrary, a benchmark update that is based on a single or multiple flat rate(s) does not as rapidly and fully reduce allocation of the best performing sectors but leaves them a fair share of these gains. At the same time, it keeps the pressure on sectors with slow innovation rates to improve further.

Similarly, a mandatory compensation of indirect carbon costs in full, as desired by some industry stakeholders, may come at the cost that companies have less incentives to search for contracts with providers of renewable energy on the market or for Member States to support investments into renewable energy with a view to bring down the carbon intensity of electricity (and thereby the carbon costs).

7.5.5.4.Most efficient installations do not face undue carbon costs leading to carbon leakage

To ensure that the most efficient installations do not face undue carbon costs leading to carbon leakage, it is important that those installations that can operate at a carbon-efficiency level close to the applicable benchmarks receive a level of free allocation that covers their carbon costs that are not passed through to consumers.

The benchmarking approach in general leads to a low risk of undue carbon costs for the most efficient installations as those de facto set the benchmark values. However, the 'Simple' package is not sector-specific (reducing values of 2007-08 benchmarks by a flat-rate percentage) and some sectors might not be able to reduce their emissions to the same extent as others. Examples are activities with a significant share of CO2 process emissions. This specific situation could however be addressed by using multiple flat rates, and/or complementing with additional data collection. Therefore, the "Limited changes", "Targeted" and "Baseline B" packages with their sector-specific benchmark updates score better in ensuring that the most efficient installations can operate at benchmark levels.

The option packages 'Limited changes' and the 'Targeted' offer refined classification of sectors into four carbon leakage groups and are expected to avoid the application of the correction factor. The 'Baseline' packages and the 'simple' package do not achieve a free allocation that matches the carbon costs of the most efficient installations because the correction factor kicks in: the estimated average correction factor (over the period 2021 – 2030) for package 'Simple' is around 5-10% and 10-20% for 'Baseline B' and 'Baseline Bbis'. In 2030, the correction factor may even reach 35%.

It is noteworthy that 'Baseline Bbis' excludes a high number of sectors with low trade or carbon intensity from 100% free allocation, However, this is not enough to significantly reduce the correction factor because those excluded sectors would anyway be eligible only for a low number of free allowances (due to their low carbon intensity).

As part of the 'Targeted' package, mandatory financial support to compensate for indirect costs (passed through carbon costs of electricity production) would minimise the risk that the most efficient installations of electro-intensive sectors could face undue carbon costs.

To conclude, 'Targeted' package offers a range of elements to ensure the most efficient installations do not face undue carbon costs. The package 'Limited changes' addresses indirect costs to a lesser extent, but has a clearly higher capability to avoid undue costs for most efficient installations than 'Baseline B' (since in "Limited Changes" a cross-sectoral correction is not expected to be applied). 'Baseline Bbis' leads to similar results as 'Baseline B'. The 'Simple' package fails to provide a better targeted allocation as all sectors receive the same percentage of free allocation and the correction factor is also triggered.

7.5.5.5.Better alignment with production levels

'Baseline B' allows adjustments for production level decreases (partial cessation rules) and significant capacity increases. However, no additional free allowances are granted for production increases. Furthermore, the rules for the New Entrants' Reserve stay unchanged.

The 'Limited changes' and 'Targeted' packages ensure a very high level of alignment with production levels as they use two baseline periods, thus avoiding that allocation at the end of the period is based on production data more than a decade old. Furthermore, they allow annual adjustments for changing production levels in both directions (increases and decreases). This means additional allocation can be provided for significantly increased production even without a capacity increase.

For the 'Simple' package, the alignment with production levels is better than for 'Baseline B' due to the annual adjustments for changing production levels in both directions. However, only one data collection process for 10 years is part of this package.

More flexible rules for the New Entrants Reserve (with the 'Simple', 'Limited changes' and 'Targeted' packages) will provide more allowances for fast-growing installations. Furthermore, if the New Entrants Reserve is funded by unallocated allowances from phase 3 (and not from the available free allowances in phase 4), a higher amount of free allowances will be available for distribution to industries. This will help to reduce the correction factor and lead to a more targeted allocation of free allowances. It could however impair to some extent the functioning of the Market Stability Reserve.

7.5.5.6.Avoid windfall profits

Windfall profits are avoided if the allocation system provides compensation only for costs that are actually assumed by the operators, i.e. costs which cannot be passed on into product prices without a significant loss of market share. A well-targeted allocation system aims to achieve this objective.

The 'Limited changes' and 'Targeted' packages are highly targeted with their four carbon leakage groups and benchmark values reflecting updated performance data, and thus represent a lower level of risk of windfall profits than 'Baseline B'.

On the contrary, the 'Simple' package (with no carbon leakage groups) has a higher risk of windfall profits in certain sectors than in 'Baseline B' (two groups).

Furthermore, the high level of alignment with production levels of the 'Limited changes' and 'Targeted' packages leads to a lower risk of windfall profits from overestimated production levels for those two packages. The 'Simple' package allows such adjustments to a lesser extent, but is still somewhat better in that respect than the 'Baseline B' packages.

7.5.5.7.No increased administrative complexity

Administrative complexity puts a burden both on private companies, as well as public authorities.

The 'Limited changes' and 'Targeted' packages show the highest degree of administrative complexity: the frequent and detailed adjustments of all elements (benchmark values, carbon leakage groups, production levels) may lead to lengthy administrative processes and therefore reduce predictability for industry. In particular, a regular update of the benchmark values based on full data collection (from all installations in the EU) will involve a high workload for all involved parties because each benchmark value has to be re-calculated precisely taking into account lessons learnt from the first data collection exercise. To the contrary, an approach based on flat rates can build on the existing benchmarks.

With regard to the alignment to production data, the administrative costs of the 'Simple' package and 'Baseline B' are not expected to differ significantly as both packages are based on one data collection exercise. The annual adjustment for significant production level changes (as introduced in the 'Simple' package) is expected to decrease administrative complexity because these new rules would simplify the current procedures for capacity changes and partial cessations.

With regard to carbon leakage rules, a higher number of groups could require some more data from certain installations (e.g. installations exporting or using heat for several activities) and therefore slightly increase administrative costs. Finally, the 'Targeted' package foresees that all Member States shall provide compensation for indirect costs. As currently the vast majority of Member States does not provide compensation, this requirement will increase the administrative burden for those national authorities which would otherwise not provide compensation, as well as beneficiary companies in those Member States.

7.5.6.Concluding remarks

The 2030 Climate and Energy Framework – as endorsed by the European Council in October 2014 – sets important limits on the allocation of free allowances to industry:

There will only be a limited amount of allowances available for free allocation to industry as from 2021 (see section 6.2  on the auction share);

This limited amount of free allowances will decrease further in line with the linear reduction factor (see section  6.1 .).

Furthermore, the European leaders have given clear guidance that the allocation of this limited amount of free allowances should be done in such a way not to constrain the competitiveness of European industries (as long as other major economic regions have not adopted similar climate policies). The operational objectives (as listed in the table below) are therefore based on the guidance by the European leaders, and the first four operational objectives refer to competitiveness in its different dimensions.

The foregoing assessment has shown that it is possible to better target the available allowances towards those sectors that are most exposed to carbon leakage risks ("Targeted" option package):

Regular updates of benchmarks to better reflect technological progress;

More refined classification of sectors according to their carbon leakage risks;

Closer alignment of free allocation to recent production levels.

However, such a policy has its costs in terms of reduced innovation incentives and increased administrative complexity. In particular, repeated benchmark updates based on full data collection may lower innovation incentives as companies would fear that they could not retain the full return on their innovation efforts. Furthermore, each additional collection of benchmarking and production data increases administrative costs for private companies and public authorities.

Similarly, a more harmonised compensation of indirect carbon costs – out of the EU ETS auction revenues for Member States – has the benefit to offer a better protection for electro-intensive industries. But this comes at the cost that auction revenues cannot be used to support e.g. renewable electricity generation that may in the long term be the better investment to reduce the carbon costs. Furthermore, higher compensation may reduce the incentives for companies to search for contracts with providers of renewable energy on the market.

Table 3: Overview of assessment of policy option packages compared to 'Baseline B'

Given the trade-offs emerging from the analysis and the range of guiding principles for the future carbon leakage and free allocation rules agreed by the European Council, policy makers will have choices between the following alternatives:

The 'Baseline B' and 'Baseline Bbis' packages provide predictability to industry as the current rules are continued. However, industry will be subject to a uniform cross-sectoral correction factor that may increase up to 35% in 2030 and reduce free allocation across sectors irrespective of their carbon leakage risk. Even though 'Baseline Bbis' excludes sectors with low trade or carbon intensity from 100% free allocation group, this will not significantly reduce the correction factor because those excluded sectors would anyway be eligible only for a low number of free allowances (due to their low carbon emissions).

The 'Simple' package groups together the options with low administrative complexity:

oThe benchmarks are updated based on a flat-rate – which reflects the average technological progress over the last decade – instead of going through a laborious and complex data collection exercise to re-establish new benchmark values for all sectors. This keeps administrative costs low and provides good incentives for innovation as above-average performers can keep a larger share of their profits. A further benefit is the closer alignment of free allocation to production data that comes at limited additional administrative cost.

oHowever, on the down-side, this simple benchmark and a single carbon-leakage group with 90% free allocation do not sufficiently take account of differences in technological capabilities or in carbon leakage risks across sectors. The 'Simple' package scores therefore poorly on avoidance of undue costs and reflection of technological progress.

oAn interesting improvement to better reflect technological differences could be to use three flat rates (low, medium, high) instead of one to address outliers. Even though some data collection would become necessary, the administrative complexity would still be considerably lower than with a full-fledged re-benchmarking of all individual rates.

The option packages 'Limited changes' and 'Targeted' show how to better address the differences in carbon leakage risks across sectors. Instead of two carbon leakage groups ('Low risk' with 30% free allocation and 'Very high' risk with 100% free allocation), the sectors are classified into four groups according to their risk exposure. In addition to the two existing groups, a 'High risk' group with 80% free allocation and a 'Medium risk' group with 60% free allocation are created. This more refined classification methodology provides a better targeted distribution of allowances across sectors and avoids that the most exposed sectors face undue carbon costs but also that less exposed sectors enjoy windfall profits.

Together with regular benchmark updates based on data collection (two collection exercises in 'Targeted package") and a more harmonised indirect cost compensation (only with 'Targeted' package), these packages show also the highest degree of administrative complexity. Furthermore, the frequent and detailed adjustments of all elements (benchmark values, carbon leakage group, production level) may reduce predictability and therefore investment and innovation incentives.

Given the advantages and disadvantages of the different packages, it could also be considered to choose certain measures of the 'Simple' package that perform well with regard to economic incentives and administrative simplicity – e.g. benchmark values based on (multiple) flat rate(s) and closer alignment to production data – and combine them with the carbon leakage rules (from the 'Limited changes' and 'Targeted' packages) to better target free allowances towards those sectors that are most exposed to carbon leakage rules. However, the policy choice ultimately depends on what emphasis decision-makers put on the different objectives.

8.Low-carbon funding mechanisms

The impact assessment of the 2030 climate and energy framework indicated that in the period to 2030 significant investments will be needed in the EU for energy system modernisation and reaching the 2030 climate and energy framework objectives. According to the impact assessment, the investments for meeting the 2030 objectives would be higher compared to GDP for lower income Member States, reflecting the relatively higher carbon intensity and lower energy efficiency of their economies and more limited financial resources.  Lower income Member States may face general and financial barriers (less liquid local financial markets, high risk profile and limited creditworthiness of several key actors) to mobilise the required investments which may prevent full financing by the market and limit the potential to finance the needed investments without public support. Realising the emission reductions in the lower income Member States could thus contribute to cost-effective reductions from a European perspective. It could also contribute towards the priorities of the European Energy Union with a forward-looking climate policy: more sustainable, secure, competitive and affordable energy for both citizens and businesses.

In this context, the European Council has agreed that a reserve of 2% of the EU ETS allowances will be set aside between 2021 and 2030 88 , and the proceeds from selling the allowances in this reserve will be used to create a Modernisation Fund supporting the lower income EU Member States (with GDP per capita below 60% of the EU average) in improving energy efficiency and modernising their energy systems, while ensuring simplified arrangements for small scale projects. Furthermore, the existing option for lower income Member States to allocate allowances for free to their power sectors is to be continued up to a maximum of 40% of the allowances belonging to a given Member State before redistribution. Such investments will need to be aligned with European climate and energy policies and (where relevant) support will be subject to State aid control. In contrast to the Innovation Fund, where allowances belonging to all Member States are used and projects from across the EU are eligible for support, for the Modernisation Fund, while allowances from all Member States are used, only projects in the beneficiary Member States would be eligible. The share of each of the beneficiary Member States for the Modernisation Fund follows directly from the European Council conclusions of October 2014.

In addition, the ETS Directive supports low-carbon innovation through renewables and carbon capture and storage through the existing NER 300 mechanism. The €2.1 billion of NER 300 funding awarded so far will leverage approximately €2.7 billion of private investments and mobilise €700 million from other public sources (see Annex 4.1 ). Recognising the continuing importance of technological development, in its conclusions on the 2030 framework the European Council agreed to broaden the support by including industrial innovation. To achieve this, it was agreed to set up an Innovation Fund also including industry and with an initial endowment increased to 400 million allowances. 89  

This chapter builds on existing experience and analyses options for the implementation of these low-carbon funding mechanisms in the ETS Directive.

8.1.Innovation fund

8.1.1.Problem definition

To reach its long-term decarbonisation goals 90 the EU needs to step up its efforts to rapidly introduce new low-carbon technologies to the market. An innovation-driven transition to a low-carbon economy contributes to the EU climate objectives and offers opportunities for growth and jobs. This has been recognized in the EU’s framework strategy for a resilient Energy Union with a forward looking climate change policy 91 , which highlighted as key priorities for research and development renewable energies and CCS for the power sector and industry. To reach the EU objectives, further support is needed for innovative low-carbon technologies and processes in the demonstration phase, which is a crucial step towards commercialisation and deployment.

At the end of 2014, the Special Task Force (Member States, Commission, EIB) on Investment 92 report indicated that while the EU remains a world leader in a number of medium- and high-technology sectors (including clean energy technologies), the EU position is increasingly being challenged by our global competitors, such as USA and China. As indicated in the report, an increasing number of Member States have started to cut back direct public R&D spending in their fiscal consolidation efforts. Complementary action at EU level is an effective way to maximise the development of highly innovative, low-carbon demonstration projects through EU-wide competition. Therefore, the Commission has identified the expansion of renewable energy and resource efficiency as well R&D as priority areas of the European Fund for Strategic Investment (EFSI) that will generate €315 billion of additional investment in the EU in the next three years. These efforts could be complemented by using revenues generated through the sale of EU ETS allowances to promote cost-effective emission reductions in line with the ETS.

This impact assessment analyses how to design the Innovation Fund to enable highly innovative, low-carbon first-of-a-kind (FOAK) projects in the European energy and industry sectors to support innovative low-carbon technologies and processes, especially in the demonstration phase.

8.1.1.1.Underlying drivers of the problem

Together with the price of EU ETS allowances, the need for companies to remain competitive and to create new products incentivises the development of innovative low-carbon technologies. Irrespective of the level of the carbon price, the ETS by itself may be insufficient to drive investment in R&D and trigger pre-commercial demonstration phase of new low-carbon technologies, in particular where other factors such as the high inherent level of technological risk, are present. The prevailing low carbon price has further underlined the need for public support for highly innovative technologies to achieve the necessary emission reductions. Still, the long-term development of the carbon price will be a key driver for the projects to be effectively supported by the Innovation Fund.

Innovative technologies, notably those involving FOAK projects, face considerable risks and often require public financial support to make the transition from R&D to commercialisation and to overcome the "valley of death", i.e. the transition between R&D and market uptake, when new products have to be produced and tested at commercial scale for the first time. The risk profile of demonstration investments in low-carbon innovations is often too high to attract conventional commercial finance and there is a considerable technological risk in the construction and implementation of new technologies for industrial users. Once the technologies are proven and performance is validated, the market can provide private finance to scale them up and to commercialise them.

Most EU energy-intensive industries and low-carbon energy sectors have developed sectoral low-carbon roadmaps for 2050, identifying promising future technologies to decrease CO2 emissions 93 . However, in an uncertain environment many companies might be reluctant to commit to innovation cycles, which put notable strain on their human and financial resources, and are a liability on their balance sheets. This uncertainty, in particular during the construction phase, forms a financial barrier for both small companies, who often suffer from a lack of access to capital, and large ones, who may lack sufficient financial incentives in the decision-making process to justify innovative and therefore risky investments. A stable carbon price signal is one of the elements that can improve the investment climate for low-carbon investments.

The need to provide additional incentives to trigger private investments in new low-carbon technologies has been confirmed by Commission analysis which highlighted the importance of speeding up investments in breakthrough technologies and addressing access to finance as a key challenge 94 . There is a need to channel public funds to support investments that contribute to achieving the EU policy objectives related both to climate and economic growth. Existing support has not always provided the financing required for deployment of all new technologies due to the high level of risk and large size of the projects.

Financial support for innovation could be provided through traditional grants and/or through financial instruments. The Commission 95 has highlighted that financial instruments as a form of policy intervention can have several benefits, including increased effectiveness and multiplier effects for limited public resources.

The Innovation Fund could help to bridge this financing gap by providing grants and/or financial instruments specifically targeted at the risk profile of low-carbon demonstration projects for renewable energy sources (RES), CCS and industry.

For these reasons, this analysis focuses on how to address the financial barriers preventing investments in low-carbon innovation, with the existing approach of the NER 300 as starting point. For detailed analysis of lessons learnt on NER300 see Annex 4. Section 8.2.1.3 indicates which lessons can be drawn from other existing EU instruments.

8.1.2.Operational policy objectives

Based on the underlying problem drivers and the lessons learnt from the existing NER 300 mechanism, the operational objectives for the design of the Innovation Fund are the following:

Achieve breakthrough innovation in the energy and industry sectors in Europe, while targeting support to ensure best use of limited funds;

Address financial barriers that the project developer needs to overcome when starting the project and provide incentives to commercial-scale low-carbon FOAK projects;

Avoid distortion of competition and minimise the impact on the general functioning of the European carbon market;

Set up an efficient, simple management structure.

The operational objectives will be used to derive criteria for comparing policy options and assessing their impacts.

8.1.3.Policy options

A key challenge in developing the policy options is to appropriately cover the different breakthrough technologies and proposals within the RES, CCS and industry sectors to ensure the selection of sufficiently innovative proposals. The Innovation Fund also has to ensure that appropriate types and levels of support can be offered. The question of addressing risk – both in terms of the funding rate and of reducing the operational risk – needs to be tackled. To this end, potential options are considered both for the screening and selection of projects and for how financial support is given.

8.1.3.1.Screening of projects

As under the NER 300, to be eligible projects should fall under one of the pre-defined innovative technology categories. The proposals should then be assessed to ensure they are sufficiently innovative. Eligible projects should then be ranked based on the cost-effectiveness of their performance.

Industry projects might merit a different approach, since putting projects with different characteristics and needs in direct competition might not deliver a good technological spread of proposals. Ranking industry projects only on the basis of their performance would likely reward only the large ones, which may benefit from economies of scale and, thus, have better overall cost performance. The selection would then risk primarily reflecting the differences between sectors and in physical boundaries, rather than level of innovation.

Certain sectoral low-carbon roadmaps for 2050 developed by industry have identified possible future technologies with a high level of detail, for instance in the forest and fibre sector 96 . However, more work is still needed to identify innovative low-carbon breakthroughs in many other industrial sectors.

In industrial sectors, capturing the relative merits of proposals in terms of innovation based on a directly comparable criterion could be done through a qualitative assessment, but this approach would make a direct comparison more difficult than a quantitative criterion such as cost-per-unit performance. Special attention must be paid to maintaining an objective way of comparing diverse industry projects, to ensure that funding will deliver innovation breakthroughs with EU value added. For these reasons, comparing improvements to a pre-defined benchmark 97 or the replicability potential of a technology 98 could be considered as eligibility criteria when making the support scheme operational.

Alternatively, innovation could be used as a ranking criterion. Industry projects' innovative potential could for example be rated from 1 to 4 99 by comparing projects' technologies to the state-of-the-art and measuring their availability amongst multiple vendors, degree of development and potential for scale-up. Where the innovative quality is judged to be equal and the funds are insufficient to fund all proposals within the same rating, the cost-effectiveness of performance could be used as a second criterion.

It should be noted that innovation should be used either as an eligibility criterion or as a ranking one, to avoid confusion in the selection process.

8.1.3.2.Conditionality of Awards

NER 300 awards are linked to projects achieving operational performance. In order to cover some of the risks, the project can receive a certain degree of upfront funding upon a Member State's guarantee, followed by 100% of funding when demonstrating at least 75% of the targeted performance.

An alternative approach in order to widen the risk coverage in the Innovation Fund would be to award part of the funding on the basis of achieving milestones in the construction phase. A fixed percentage (e.g. 30%) of the additional costs of innovation could be granted conditional on finalising steps in design, permitting and construction. This is the approach followed by the US ARPA-E programme 100 , advancing high-potential, high-impact energy technologies. ARPA-E selectees may request a Budget Plan Payment (BPP), to obtain reimbursements on a prospective basis in order to purchase the necessary equipment or services required to achieve specific milestones.

Linking funding to the achievement of specific milestones has proved beneficial also in the EU based on one of the lessons from the 2009 European Energy Programme for Recovery (EEPR). Under this programme, grants were awarded to highly strategic projects in gas and electricity infrastructure, offshore wind and carbon capture and storage 101 . This could apply to both energy and industry projects, since there is no material difference in the construction process.

As an example, the hypothetical steps in the construction of an innovative offshore wind farm could be considered to identify some steps which could be considered milestones: reaching final investment decision (FID), construction of all foundations, construction of all turbines or installation of all turbines and the start of the delivery of power to the grid. All milestones should be relevant in terms of justifying the partial granting of investment support.

8.1.3.3.Type of instrument

The NER 300 mechanism provides cash grants. Such support is attractive both for financial institutions, for which innovative projects are often not bankable, and for project sponsors, which find in grants an effective way of improving their financial standing (see also section 8.1.5.2 ). The NER 300 programme has linked such grants to operational performance to deliver funding to highly innovative FOAK projects characterised by long planning and construction times, a high risk profile and hard-to-predict budgets.

An alternative could be to switch to a financial instrument. Financial instruments are designed to address market failures or sub-optimal investment situations which have proven to be economically viable but do not give rise to sufficient funding from market sources due to their particular risk structure. As an illustration, financial instruments can include guarantees on first loss that could allow banks to provide loans to riskier projects than they would otherwise support. An alternative could be to invest public equity, in which case projects would be selected through a permanent financing facility on an open-call basis. Due to the high risk involved in FOAK demonstration, the Innovation Fund should target projects with higher risk levels when compared to the investments supported through existing financial instruments.

8.1.3.4.Maximum rate of funding

NER 300 covers up to 50% of the projects' additional costs of the innovation (for example, the CCS part of a power plant or additional costs for a RES plant compared to a fossil fuel reference plant). RES projects on average requested NER 300 funding equivalent to 39% of their additional costs. Due to 'funding cap' 102 , no more than €300 million could be awarded to the only CCS project, which covers 34% of its additional costs.

In the consultation on carbon leakage provisions and innovation support 103 , more than half of the industry respondents indicated that in their view, there is a particular need to strengthen the EU's innovation support for the implementation of large-scale pilot projects. In the general consultation on the ETS revision, different industrial sectors, Member States and civil society members (including trade associations and NGOs) have highlighted the importance of appropriate risk sharing in the innovation process, for example through an increase in the current co-financing rate. At the same time, for industry innovation, the projects are not as large as for CCS. They are thus not likely to be limited by the 'funding cap' of 15% share of the total amount available. However, a low limit on the funding rate per project could have a negative impact on the financial viability of highly innovative and therefore risky industrial projects.  

Access to finance is a major barrier for implementation of innovative projects and in certain cases the current 50% co-financing requirement has been too high for project promoters to secure implementation and may also have resulted in some projects not participating.

For these reasons and to reduce the significant financial barriers (e.g. lacking or not ensured national funding, additional operating support, feed-in-tariffs or similar national support schemes, equity, long-term debt financing or revenues 104 ) as well as the level of technological risk, the maximum funding rate in the Innovation Fund could be increased to up to 75%.

For CCS, 10 out of 11 submitted proposals were not confirmed, mainly due to a lack of sufficient additional funding and as such, increasing the rate will improve their financial viability. As regards other technologies, existing experience suggests that for most subcategories of RES projects, the actual requested funding for such projects from the NER 300 will on average be lower than the maximum allowed. At the same time, many awarded RES projects have not yet achieved a Final Investment Decision, which could be a reason to also consider allowing an increased maximum rate for RES projects, if further experience confirms this is needed.

As the lessons learnt from the NER 300 are limited to a few years of implementation and the NER 300 has a more limited scope than the Innovation Fund, the exact impact of using higher rates is challenging to evaluate. Specific needs should, therefore, be assessed separately for RES, CCS and industry through more extensive market testing in the context of the implementing legislation. Comparing options which include the current maximum funding rate of 50% to an increased maximum rate 75% gives a first indication of the potential range of effects. To provide additional information on what these thresholds mean for project promoters, a short sensitivity analysis in (Annex 14 ) shows the level of possible funding for a range of RES, CCS and industry projects, across a range of funding rates.

If support was provided through financial instruments rather than grants, this would imply a different approach. While the detailed parameters for such support would need to be further elaborated, financial instruments could for example take the form of equity participation or of a guarantee on the first loss covering a certain percentage of the additional cost of the innovation. With such an approach, there would be no directly comparable level of funding, but rather a maximum coverage of risk related to the investment with the aim to increase the financial viability of the project.

The State aid guidelines for Energy and Environment, which apply until 2020, note that centrally managed funding, including also NER 300 funding 105  does not constitute State Aid. When such funding is combined with state aid, only the latter is considered for determining notification thresholds and maximum intensities. The existing maximum aid intensities vary between technologies and other factors, such as size of the targeted recipients or innovative nature of the project. For the Innovation Fund, the maximum funding rates and further design of the operational modalities would need to be consistent with future State Aid rules. Moreover, the use of different financial instruments (for instance equity vs. guarantees) may have different effects and the further design of the operational modalities for such instruments should be consistent with future State Aid rules. 

8.1.3.5.Parameters not varied in the options

There are a number of design features, which have proven appropriate under the NER 300 programme and it is proposed not to vary these in the options, except for an adjustment to the higher volume (400 instead of 300 million allowances) and longer duration (10 instead of 5 years) of the Innovation Fund:

The NER 300 management structure, consisting of cooperation between the European Commission, Member States and the EIB, was effective for developing and selecting projects, for monetising allowances and for managing revenues. While the simplification of the interaction between the three bodies and a reduction of administrative burden for project sponsors should be considered, this will mainly be subject to a future implementing measure and will be assessed in this context.

The 'funding cap' (limit of maximum 15% allowances per project) is proposed to be maintained 106 . A higher limit would increase the potential for large projects to participate, but would result in a lower total number of projects funded.

The limit of 3 projects per Member State introduced in the NER 300 Decision was appropriate to ensure adequate geographical coverage under the programme (see Annex 4.1 ). It could be maintained or adjusted to 4 projects per Member State, dependent on other design features such as the maximum funding rate and the resulting total number of projects. This element will be subject to a future implementing measure and will be assessed in this context.

Two calls for proposals within 5 years were adequate under the NER 300 programme, to ensure stability in funding and monetisation of allowances. This could be maintained or adjusted to 4 calls, depending on the appropriate timing of monetising 400 million allowances, as described in Annex 4.1  

The current knowledge-sharing provisions could be streamlined to ensure smoother implementation. An element of knowledge sharing should be kept as there is a clear public interest in replicability of the projects. Industrial sectors that have an intra-sectoral cooperation already incorporated, i.e. that have elaborated their low-carbon roadmaps, could have better possibilities of developing breakthrough technologies than individual companies acting alone. Previous examples, such as e.g the ULCOS 107 process of the steel sector, illustrate that such cooperative approach is possible and accepted by industry. In case several companies of a particular sector join resources, the benefits of knowledge-sharing – and of their possible licensing - could be included as one of the design features, taking into account respect of competition rules. These knowledge-sharing provisions are set in implementing legislation and will be assessed in this context.

Indicative shares for CCS and RES projects with a smooth spill over possibility between the groups were crucial under the NER 300 programme to ensure the allocation of all available funds. A similar indicative share could be set for industry projects.

8.1.4.Option packages

Four option packages are taken into consideration, by combining the options for the elements set out in chapter 8.1.3 . The legal baseline scenario would imply no changes to the ETS Directive and hence no continuation of support to innovation. The alternative baseline scenario represents a continuation of the current practice by extending the current NER 300 rules to industry, increasing its endowment to 400 million allowances, and its duration to 10 years.

Option 1 envisages tailored rules and selection criteria applying for industry, and the current rules for RES and CCS. The risk approach would be changed for all categories (industry, RES and CCS) by higher co-funding rates and early disbursement of part of the funds following the achievement of construction milestones. The essence of this option is to provide alternatives to share investors' risks and create a more conducive climate for innovative investments.

Option 2 foresees the creation of a permanent financing facility selecting projects continuously on a first-come-first-served basis considering the innovation merits of the proposal. This option assumes financial instruments instead of grants.

Table 4: Option packages for Innovation Fund

8.1.5.Analysis of impacts

8.1.5.1.General impacts

Complexity and administrative burden

In the alternative baseline and Option 1 the innovation support would be awarded via calls for proposals, primarily carried out under the responsibility of the European Commission. Building on the NER 300 experience, well-organised and managed calls would not increase the complexity or administrative burden compared to current practice. However, specific rules for the inclusion of industrial projects might increase the complexity of the selection assessment, since elements such as replicability and ranking based on innovation and cost-effectiveness of performance would need to be considered.

In Option 2, the permanent financing facility would allocate support to projects selected on a first-come-first-served basis. This continuous selection would provide applicants with more flexibility as their proposals would not be tied to the timing of calls for proposals. In terms of project evaluation, for a comparable number of projects to be reviewed under Options 1 and 2, no substantial differences are expected in the overall administrative burden. Structuring financial instruments to address the specific needs of different technologies would require considerable know-how to assess and allocate the risks properly and this could increase complexity. At the same time, under Option 2 this role would need to be assumed by financial institutions such as the EIB that already has extensive experience in structuring complex deals. An expanded role for such a financial institution may be reflected in an increase of the fees charged for the activities carried out when compared to the baseline scenarios. The transparency of the level of support should also be ensured. The experience in setting up and operationalising the new European Fund for Strategic Investments under the EU Investment Plan would also be relevant to ensure a simple and transparent governance structure.

When compared to the alternative baseline, the complexity and administrative burden of Option 1 and Option 2 would be higher, due to the additional factors to be taken into account during the selection process.

Complementarity to other EU instruments

A range of instruments has been developed at EU level to support the development of innovation activities, and research and development more broadly 108 , and these are expected to support total R&D investments of approximately €48 billion through 2020 109 . Additionally, at the end of 2014 the Commission unveiled the new European Fund for Strategic Investments under the EU Investment Plan 110 . It is expected to make an important contribution in the short term to the climate and energy investments, highlighting the importance of infrastructure, energy efficiency and renewables 111 . 

While it cannot be anticipated what EU level instruments will exist in 2021, the complementarity of the Innovation Fund to existing policy instruments would come from its specific focus on support for low-carbon innovation at the pre-commercial demonstration phase. Managing the Innovation Fund at EU level ensures that it complements and reinforces other existing instruments.

Providing support in the form of a performance-based grant (as in the alternative baseline or option 1) or through a financial instrument which explicitly targets projects with a higher level of risk when compared to other existing EU-wide instruments, such as the existing InnovFin 112   programme (as in option 2) can be complementary to other EU instruments.

When compared to the alternative baseline, the complementarity of both Option 1 and Option 2 is similarly positive, and explained primarily by the focus of the support on low carbon first of a kind demonstration projects and the expansion of scope to industrial projects.

8.1.5.2.Economic impacts

Several economic impacts are considered in this section. Firstly, the overall ability of the fund to attract and support a large number of innovative projects is considered to be the result of three impacts: the ability to attract innovative projects, the fund's effectiveness in addressing barriers to allow these projects to be realised and finally the expected leverage, indicating the volume of additional investment mobilised as a result of the support offered. Two other economic impacts at a higher level are also considered, the effects on competitiveness and the EU added value realised.

Effectiveness in addressing barriers for low-carbon innovation

The NER 300 lessons learnt (see Annex 4.1 ) showed that the current NER 300 rules, i.e. the alternative baseline option, were effective in encouraging the development of projects for RES innovation. However, the deadlines for reaching the final investment decision within two years and starting operations within four years proved to be too ambitious due to factors such as the long preparation, construction and permitting timelines of FOAK projects. In addition, this approach does not seem to adequately address the risks for pre-commercial demonstration projects in energy intensive industries and CCS projects. This is supported both by the NER 300 experience of only 1 CCS project awarded and by the majority of stakeholders 113 . 

Various stakeholders indicated that the current NER 300 rules would not be sufficient to trigger innovation for the energy intensive industry, since project sponsors would have to bear the high financial and technical risks of capital intensive investments. Since the return on such investments can be reaped only in the longer term, in the short run such projects do not seem to be ranked as a first priority in companies' internal decision-making processes. As a result, innovation might not occur at the speed needed to meet the EU's long-term decarbonisation objectives. It is likely that such investments would rather be realised in those Member States where higher levels of national public support are offered, while in others the emission reduction and innovation potential could remain underdeveloped.

In that regard, Option 1 could substantially lower the project sponsors' needs for co-financing, given the higher funding rate and the early disbursement of part of the funds following the achievement of milestones. The funding provided through a grant would lower the costs borne by project sponsors, improving the financial viability of the project. At the same time, linking the grant to the achievement of milestones would be more attractive when compared to the alternative baseline. Under the alternative baseline, the funding is tied to operational performance at a later stage and discounts will be applied by the company to reflect uncertainty about the current value of funds that will be received in future.

Option 2 which envisages the creation of a financial instrument (e.g. guarantees, loans or equity investments, and risk-sharing instruments) 114 , can also address barriers faced by project promoters. If providing public equity, the projects could be supported from the initial stages of development. Such public equity investment could lower the need for project promoters to raise debt and private equity. As an alternative, if the financial instrument is implemented as a guarantee that would cover for a percentage of the losses in case of a loan default, this would allow financial institutions such as the EIB, national promotional banks and/or commercial banks to lend to riskier projects than they would have done otherwise. This also addresses projects in the earlier stage of development compared to the alternative baseline, i.e. continued current rules. In certain cases, financial instruments might not address the large financing gap in the demonstration of FOAK projects (e.g. CCS projects). Financial instruments can stimulate a significantly higher level of private investments, but this will likely be in technologies that are closer to commercial scale deployment.

In conclusion, while both Option 1 and 2 have potential to address barriers for low-carbon innovation when compared to the alternative baseline, Option 1 is expected to be more effective due to the combination of a more targeted risk sharing approach and the possibility of a higher rate of funding.

Potential to attract innovative projects

The lessons learnt from the existing NER 300 mechanism (Annex 4.1 ) show that the current rules and selection mechanism have resulted in a portfolio of projects for renewables that were identified in many cases as highly innovative or even game changing (80%) in the existing framework. At the same time, while multiple proposals were initially submitted, only one CCS project has been confirmed.

The alternative baseline option would imply applying a one-size-fits-all approach, using the existing rules for RES, CCS and industry. As discussed in Section 8.1.5.2 , there is a risk of not attracting a similar number of innovative project proposals in the CCS and industry sectors. As a result, in particular CCS projects could still face similar risks as the existing NER 300 projects, such as uncertain national funding, lack of private equity or long-term debt financing.

Option 1 has a higher potential to attract innovative projects with its risk sharing elements. Raising the maximum funding rate (up to 75%) has the potential to encourage more innovative project proposals to be submitted for the Innovation Fund. In addition, directly targeting innovation for industry proposals as the basis for the ranking of projects during the selection process (compared to the current practice of innovation as eligibility criterion) is expected to make the system more attractive for innovative projects with a lower technological maturity. As discussed above, increasing the maximum funding rate would also facilitate addressing the financial barriers of such innovative projects, if they are selected.

Option 2 involves providing support to projects through financial instruments, such as loans, guarantees or equity which would target projects with a higher risk level compared to existing instruments (e.g. the current InnovFin programme, with the exception of the recently launched InnovFin Energy Demo Projects (EDP) pilot facility which does cover such higher risk levels). This could enable a pipeline of more innovative projects with a higher risk profile to access financing when compared to existing EU and EIB financial instruments, for example by covering the first loss which would be incurred or through co-investing equity (See Section 8.1.5.2 ). A decision on supporting a project would be carried out by an entrusted financial entity such as the EIB that would evaluate the projects based on economic, financial, technical and environmental criteria. There is a risk that such an approach would be more beneficial to closer to market pre-commercial technologies that have higher short-term revenue generating potential, overlooking technologies that take longer to develop but are more interesting in the long term. Therefore under Option 2 financial instruments might not be sufficient to fully address the financial barriers faced by breakthrough innovation projects without additional public interventions, such as grants. In conclusion, while financial instruments could be targeted at earlier stages in project life cycle when compared to the performance based grant in the alternative baseline, there is a risk that such an approach would be likely to attract technologies closer to commercialisation. Whether Option 2 could attract more high risk innovative projects when compared to the alternative baseline would therefore depend on how the instrument is operationalised.

In conclusion, option 1 has the highest potential to attract innovative projects when compared to the alternative baseline, while option 2 would be expected to lead to comparable outcomes.

Leverage

A similar leverage as for the NER 300 funding (see Annex 4.1 ) is expected for the alternative baseline. The leverage of Option 1 would likely be lower due to the higher co-financing rate and hence the lower need for private investments. Additionally under Option 1 part of the funds would be reimbursed prior to proof of performance. While funding partial success or failure can still deliver benefits in the form of knowledge sharing, from a financial perspective this would imply the loss of some funds compared to disbursement based on proving operational performance.

On the other hand, significantly higher overall leverage could be reached under Option 2. Preliminary evidence shows that the EU contribution to financial instruments mobilises a global investment exceeding the EU contribution by 4 to 10 times on average 115 . Since the Innovation Fund addresses inherently riskier innovation projects, the leverage realised would likely be in the lower range but still significantly higher than for traditional grants under Options 1 and the alternative baseline. While project failures may lead to overall higher losses than with grants or debt instruments, financial instruments proceeds might reflow into the fund and be used to finance additional projects.

When compared to the alternative baseline, Option 1 would result in a lower leverage due to the higher funding rate, while Option 2 would be expected to result in an increase of the leverage as financial instruments can mobilise more additional funding.

Competitiveness

Low-carbon innovation in the energy and industry sectors would improve the overall EU competitiveness by supporting low-carbon technologies in which the EU has global technological leadership. Better and more efficient technologies will benefit the entire supply value chain and ultimately consumers. Furthermore, such innovative technologies can create a substantial number of new jobs and generate new business opportunities. There is also evidence that low-carbon technologies bring larger overall economic benefits, as they generate more knowledge in the economy, which in turn can be used by other innovators to further develop new technologies, leading to a virtuous cycle of innovation. 116  

Additionally, knowledge-sharing requirements under the Innovation Fund would contribute to the dissemination of the results across EU borders, technology exchange and associated catching-up effects. It could also help to reduce the innovation gap between and within sectors and/or Member States.

The Innovation Fund could improve the prospects for EU companies to increase exports in rapidly growing low-carbon markets, thus enhancing EU competitiveness. The extent to which the different options would improve EU competitiveness would depend on how effectively the fund incentivises innovation in new low-carbon technologies and processes. The alternative baseline and Option 1 would strengthen the global competitive position of EU businesses selling low-carbon technologies. The inclusion of replicability as an eligibility criterion for industry projects in the assessment would also ensure a measure of the overall potential for improved competitiveness through a significant contribution to overall industrial emissions across the energy intensive industry sector in the EU. This could help to ensure higher benefits for competitiveness.

Option 2 is expected to result in higher leverage and may be more attractive to investments which are close to commercialisation. This could lead to quicker realisation of the projects and a more rapid diffusion of the technologies supporting competitiveness, but there is a risk that to enable breakthrough innovative FOAK demonstration projects additional support might be needed in the form of grants.

Consequently, Options 1 and 2 both have the potential to improve competitiveness compared to the alternative baseline.

EU added value and geographical distribution

The Innovation fund would target support towards projects with EU-wide significance. Through EU-wide coordination, it would be possible to reach the requisite scale for highly capital intensive demonstration projects in RES, CCS and industry across Member States in particular compared to a scenario where only national schemes exist. The EU focus could provide additional benefits in terms of the development of technological standards at EU-level. Under the alternative baseline and options 1 and 2 such benefits would also be extended to the energy-intensive industry sectors.

The Innovation Fund is to cover projects in industry, CCS and RES. Investment projects in all Member States, including small-scale projects. The experience with the NER 300 projects, selected and ranked through two calls for proposals by EU-wide competition, showed the EU value added of the programme in terms of maximising innovation and decarbonisation benefits. The geographical and technological spread of innovative projects (see Annex 4.1 ), combined with the knowledge-sharing requirement for project sponsors, is likely to result in an effective knowledge spill-over throughout Europe. As indicated in Annex  4.1 , the rules relating to the maximum number of projects per Member State did not guarantee that projects were funded in all Member States, but acted as a relatively light safeguard against a high number of project proposals being approved in only one Member State. The alternative baseline for the Innovation Fund would lead to a comparable outcome in terms of geographical distribution and contribution of EU value added. On the other hand, for both options 1 and 2, the expansion of scope to include industry is an opportunity to develop a promising project pipeline in lower income Member States which have high potential for economic growth. This could contribute to a wider geographical distribution.

Since Option 1 also includes replicability as one of the eligibility criteria for industry projects, this could help ensuring an even higher EU value added, since projects' technology could be deployed / licensed at a larger scale to other similar installations. Additionally, under the alternative baseline, project promoters should have received a guarantee from Member States to receive up-front financing. In Option 2, financial instruments would need to include a provision for maximum geographical concentration ratio to ensure geographical balance. In fact, through a large portfolio of projects with different risk profile, an optimal level of portfolio risk could be achieved. As a result, both Option 1 and Option 2 can be considered to represent a higher potential EU added value than the alternative baseline.

8.1.5.3.Environmental impacts

The Innovation Fund addresses investments in low-carbon innovation for CCS, RES and energy intensive industries. All three scenarios would facilitate the commercialization of new low-carbon technologies that would facilitate achieving the long-term decarbonisation objectives in the EU. The environmental benefits of low-carbon technologies will increase over time as the technologies are replicated and deployed on a larger scale.

In the longer term, more direct environmental benefits such as increased use of renewable energy resources, improved energy efficiency, improved local air quality and related health benefits are expected.

It should be noted that projects supported under the three options could have a different impact regarding verified avoidance of CO2 emissions. Under the alternative baseline, funding would continue to be disbursed strictly upon proof of avoided CO2 emissions. Options 1 and 2 would be less directly linked to such reductions, as the support could be paid before the project enters into operation.

With regard to environmental impacts, Option 1 and 2 are thus considered to have a similar impact when compared to the alternative baseline.

8.1.5.4.Social impacts

Although it is not possible to quantify the impacts on employment in the individual Member State, a positive impact on employment such as the creation of new high value-added jobs across the entire supply chain, could be expected in proportion to the level of investments triggered by the Innovation Fund.

As indicated in the impact assessment accompanying the Communication from the Commission "Horizon 2020 - The Framework Programme for Research and Innovation" 117 , a wealth of evidence demonstrates the crucial role that research and innovation play for the creation of more and better jobs, for productivity growth and competitiveness, and for structural economic growth. To boost future productivity and growth, it is critically important to generate breakthrough technologies and translate them into innovations that are taken up by the wider economy.

8.1.6.Comparing the options

In the previous section, seven different impacts, in particular economic impacts, were analysed as presented in the table below. These impacts contribute to the operational policy objectives that were identified (section 8.1.2 ) as follows. A positive score on the impacts "potential to attract innovative projects", "EU added value and geographical distribution", "competitiveness" and "leverage" would contribute to the objective of achieving breakthrough innovation in the energy and industry sectors in Europe, while targeting support to ensure best use of limited funds.

The impact "Effectiveness in addressing barriers for low-carbon innovation" relates directly to the objective to address financial barriers that the project developer needs to overcome when starting the project and provide incentives to commercial-scale low-carbon FOAK projects.

Thirdly, a positive score on the impact "minimise complexity and administrative burden" contributes to the objective to set up an efficient, simple management structure. Finally, while the objective of minimising the impact of the mechanism on the general functioning of the European carbon market is not directly reflected in the comparison of the options, there is a further discussion of potential impacts in relation to the timing and monetisation (Annex 13 ).

Table 5: Comparison of options for the innovation fund

When screening and comparing the options, two pluses indicate a major improvement when compared to the alternative baseline, one plus indicates an improvement, zero indicates no change and one minus indicates deterioration.

The analysis of the options shows that the options involve several trade-offs with regard to the resulting impacts. While all seven criteria presented above are relevant to comparison of the options, relative to the other criteria the ‘Effectiveness in addressing barriers for low-carbon innovation’ and ‘Potential to attract innovative projects’ are key factors contributing towards the fund objectives.

Option 1 would be very effective in addressing specific barriers for low-carbon innovation by substantially lowering the financial barriers for project sponsors. It would provide cash grants (opposed to a financial instrument such as equity or a guarantee in Option 2) combined with a higher funding rate (up to 75%, as opposed to 50% in the alternative baseline) and the option for early disbursement of part of the funds following the achievement of construction milestones (opposed to the disbursement of funds only on the basis of achieving operational performance as in the alternative baseline).

While it would be expected to result in a lower leverage and a lower number of projects being supported than either the alternative baseline or Option 2, Option 1 represents the highest potential to address the specific barriers to support the commercialization of breakthrough innovation for CCS, RES and energy intensive industry. This package would likely attract the highest number of applications of innovative FOAK projects in the energy and industry sectors and ensure effective project implementation.

Option 1 could also deliver significant EU value added by taking into account replicability of industry projects, while still rules providing safeguards to allow a wide variety in the geographical and technological spread of projects within the EU and achieving a critical mass of funding which would not be attainable by Member States alone.

Similar outcomes could also be realised by the first-come-first-served selection process provided by Option 2, but this option would offer smaller scope for a comparison between numerous project proposals and may need to be combined with additional public support in order to enable more far reaching innovative projects to be realised. Both Option 1 and Option 2 are expected to provide significant benefits with regard to competitiveness and to be complementary to existing EU instruments.

An Innovation Fund implemented through Option 1 (provision of grants), could be closely coordinated with other EU-level and national level support schemes. To increase the impact of the Innovation Fund under Option 1, an increased level of coordination between grants and financial instruments could be beneficial to address market failures. Duplication should be avoided, but a combination of such instruments could cater to a wider set of technologies and projects as grants and financial instruments normally do not fully cover the same underlying risks and could be regarded as complementary. How to ensure the coordination between the Innovation Fund and relevant existing financial instruments should be assessed in the context of the implementing legislation for the Innovation Fund.

More detailed assessment of the relevant modalities should take place in the context of the preparation of a future implementing legislation for the Innovation Fund.

8.2.Modernisation Fund

8.2.1.Problem definition

8.2.1.1.Context

The European Council in October 2014 agreed that a reserve of 2% of the EU ETS allowances will be set aside between 2021 and 2030 118 , and the proceeds from this reserve will be used to create a Modernisation Fund to support the EU Member States with lower income (with GDP per capita below 60% of the EU average) in improving energy efficiency and modernising their energy systems, while ensuring simplified arrangements for small scale projects. There are 10 beneficiary Member States: Bulgaria, the Czech Republic, Estonia, Croatia, Latvia, Lithuania, Hungary, Poland, Romania, and Slovakia 119 . The creation of the Modernisation Fund implies a net transfer of 223 million allowances, translating into potentially as much as € 5.5 billion, to the beneficiary Member States from the remaining EU Member States (See Table 34  in Annex 11 ). 

8.2.1.2.Underlying drivers of the problem

At the end of 2014, the Special Task Force (Member States, Commission, EIB) on Investment 120  as one of its tasks analysed the market barriers for investments to various sectors in the EU. It concluded that across sectors macroeconomic uncertainty, insufficient structural reforms, incomplete single market as well as regulatory constraints negatively affect the investment climate. Administrative burden has also been identified as a major bottleneck. The report highlighted that for the energy sector the barriers to investment, and hence potential solutions, differ between grids, production projects, energy efficiency and distributed renewable energy projects. See Annex 10 for details on the financial barriers for investment in the energy sector.

The Modernisation Fund could play a role in addressing specific financial barriers. While the European Council already agreed on the establishment of the fund and its purpose to modernise the energy sector and improve energy efficiency, it did not define the governance structure, including the details of the respective roles of the beneficiary Member States, EIB and other institutions.

Since the governance structure matters for the effectiveness of funding mechanisms, this impact assessment focuses on the design of the fund to make best use of the expertise and knowledge of the various institutions involved. As the Modernisation Fund is a new funding mechanism created with allowances which belong to the Member States and as such are not part of the EU budget, this Impact Assessment aims to determine, as a first step, to what extent the different institutions would be involved in setting the eligibility and selection criteria for the projects to be supported by the fund. Thus at this stage no options are developed on the eligibility criteria and the specific type of support used (grants, financial instruments or a mix of the two). The further details on the modalities of the fund, including eligibility criteria and specific type of support used (grant, financial instrument or a mix of the two) would need to be clarified and analysed in a subsequent implementing legislation following further assessment and/or decided by the board of the fund, depending on the option chosen for the governance structure.

8.2.1.3.Lessons learnt

The Modernisation Fund is a new policy mechanism, so it should draw on the experience of existing initiatives related to the ETS such as the optional free allocation to the power sector under Article 10c of the ETS Directive (see section 8.3.2 and Annex 4.2 ) and the funding of innovative investments through the NER300 initiative (see Annex 4.1 ). More general lessons on governance of investment platforms can be drawn from the Commission, EIB and Member States’ experience in implementing financial instruments, and from the work related to setting up of the European Fund for Strategic Investments under the EU Investment Plan. 121

One element that has been highlighted by private stakeholders in the recent discussions of the EU Investment Plan is the importance of robust project quality criteria and an independent selection of projects 122 . Additionally, Member States have expressed political commitment 123 to have financial instruments play a more important role in the multiannual financial framework for 2014-2020, preliminary data suggest that the uptake can be higher 124 . The experience gained during the 2007-2013 implementation period with financial instruments 125 developed by the EU and EIB shows that the effectiveness and efficiency of financial instruments can be enhanced by implementing fewer financial instruments with larger volumes to ensure critical mass 126 .

As discussed in Section  8.3 , the free allocation to the power sector in the lower income Member States indicates the need for a simple, transparent and clear approach to make effective use of available resources. Finally, the NER300 programme (see Section 8.1 and Annex 4.1 ), demonstrates the EIB’s expertise to cooperate with the Commission and Member States in due diligence to select projects and act as an agent to monetise allowances. An important element to ensure that the Modernisation Fund can start financing projects in 2021 is to time the auctioning of allowances in such a way as to provide certainty of available funds, while also avoiding a negative impact on the carbon market.

8.2.2.Policy Objectives

As discussed above, the impact assessment focuses on the governance structure of the Modernisation Fund. The fund should support transition to low carbon economy in beneficiary Member States and contribute to ensuring EU-value added and support the completion of the EU internal energy market, while ensuring that the specific market barriers and national priorities in the beneficiary Member States are addressed. In this respect, the impact assessment will also assess the potential role the Commission could play. This reflects also the views of stakeholders: on the one hand, many stakeholders 127 highlighted the importance of reflecting national priorities, while on the other hand, many indicated the importance of minimising distortion to the internal market and contributing to the EU long-term climate objectives 128 .

An additional concern raised by many stakeholders 129 is that the governance structure and decision-making process should minimise administrative burden, be feasible for operational implementation, and be coherent in order to catalyse additional investments. In particular, the process should be simplified for small-scale projects.

8.2.3.Development and screening of policy options

8.2.3.1.Policy options

The main options assessed relate to the governance structure of the Modernisation Fund and in particular the different roles that Member States, the European Commission and the EIB can play to advance its objectives. The approach taken involves varying their role of these actors in aspects, such as who determines the investment guidelines (including criteria for project eligibility and selection) monitors performance, and selects projects and programmes. While various combinations are possible, the key consideration is that the governance structure needs to reflect the specific strengths and expertise of the different institutions involved.

Eligibility, selection criteria, investment guidelines and monitoring

A main factor that can be varied is the composition of the Steering Board, which will further define the rules and guidelines insofar as this has not been done in the implementing legislation of the Modernisation Fund. Membership of the Steering Board could either be primarily reserved for the beneficiary Member States, or for a balanced representation of all 28 Member States and the Commission.

Variation 1: the 10 beneficiary Member States have control over defining the eligibility criteria, selection criteria and investment guidelines of the fund and monitoring its performance. The Commission provides general guidance.

Variation 2: the 28 Member States together with the Commission are involved in defining eligibility, selection criteria and investment guidelines and monitoring the operations of the fund.

Variation 3: The Commission defines the eligibility, selection criteria and investment guidelines and monitors performance. Within this framework beneficiary Member States pre-approve the project pipeline in line with these criteria, but do not directly define the priorities of the fund.

Project selection

With regards to project selection, the role of the EIB and other institutions is examined. There are two general options of how the EIB can be involved in the selection of projects and programmes:

Variation 1: the EIB is involved in due diligence and takes an advisory role 130 . The Commission and beneficiary Member States are responsible for project selection and approval, following advice by the EIB. The Steering Board would need to justify investment decisions that deviate from the EIB advice.

Variation 2: the EIB is delegated to take the role of a fund manager on behalf of the beneficiary Member States and the Commission, and therefore takes investment decisions in line with investment guidelines. The Steering Board would need to justify a refusal to support investment decisions taken by the EIB.

The EIB could also act as an agent to monetise the EU ETS allowances. Finally the EIB may be one of the institutions that implements the financial instruments selected under the Modernisation Fund, together with national and regional promotional banks.

8.2.3.2.Option packages

Three different combinations for the governance structure are examined to evaluate the specific impacts and reflect the differences in stakeholder views on the relative roles of the beneficiary Member States, other Member States, the Commission and the EIB 131 :

Baseline - a legal baseline would imply that 2% of the ETS allowances would be allocated to all 28 Member States rather than being used for modernising energy systems and improving energy efficiency in the beneficiary Member States 132 . The lower income Member States would then receive only some 87 million allowances, less than a third of the 310 million in the Modernisation Fund.

Option 1 - large discretion and responsibility for the 10 beneficiary Member States as the only representatives on the Steering board, setting eligibility criteria and investment guidelines. The EIB would have a limited advisory role, including performing due diligence, while the Commission could administer calls for proposals for grants. The implementation could be either through grants and/or financial instruments.

Option 2 - a wider representation on the Steering board of all 28 Member States and the Commission as well as a greater role in implementation for the EIB. The implementation would be through grants and/or financial instruments, for which the EIB could have an enhanced role as fund manager, but remaining accountable to the Steering Board to which it would report. The Commission could administer calls for proposals for grants, for which the EIB performs due diligence.

Option 3 - the beneficiary Member States would identify a pipeline of projects to which funds should be allocated, but projects would conform to eligibility criteria and general principles for project selection analysed and set in the fund’s implementing legislation. The Commission would administer any calls for proposals, for which the EIB would perform due diligence. For the purposes of the analysis, the implementation would be through a grant rather than financial instruments 133 . 

An option in which each beneficiary Member State implements the funds individually at national level is not analysed further, since this would have a limited added value compared to instruments that the beneficiary Member states already have at their disposal 134 . Such an option also has the potential to have a higher administrative burden, a higher distortive effect on the common energy market and to be less suited to promote investments with beneficial cross-border effects.  

The options are summarised below:

Table 6: Retained options for examining for the Modernisation Fund

8.2.4.Analysis of Impacts

The environmental, social and economic implications of the creation of the Modernisation Fund are driven by the European Council strategic guidance that determined the number of allowances available for the fund, the criteria based on which Member States are determined as eligible beneficiaries, as well as the method for allocation among Member States and the type of project supported 135 . Annex 11  assesses the impacts of having a Modernisation Fund against the baseline scenario of not having a Modernisation fund at all. In summary, the effects of creating the Modernisation Fund when compared to the baseline are:

Economic impacts: Compared to the baseline scenario, the Modernisation Fund would help trigger important investments in the beneficiary Member States that could contribute towards the key aspects of creating a strong European Energy Union with a forward looking climate policy. It is expected to contribute not only towards cost-effectively reaching the 2030 climate and energy objectives, but also towards diversifying the energy mix, improving security of supply and lowering import dependency.

Environmental impacts: The creation of the Modernisation Fund has environmental and health benefits for the beneficiary Member States compared to the baseline scenario. Through facilitating investments in energy efficiency and modernising the power sector, the fund would contribute to lowering greenhouse gas emissions in the beneficiary Member States under both the ETS and the non-ETS sectors. Additionally, by lowering fossil fuel consumption, local air quality may improve leading to health benefits in the beneficiary Member States and contributing to lower costs related to controlling emissions of air pollution.

Social impacts: Since the Modernisation Fund would contribute to realizing investments to facilitate the energy sector modernisation and improving energy efficiency in the beneficiary Member States, when compared to the baseline, it is expected to have a positive impact on employment for the beneficiary Member States, for example, through the creation of jobs through small scale projects. The final impact will depend on the amount of realized investments and in which specific sectors these are implemented.

After being compared to the baseline option of not creating a Modernisation Fund, the three options identified and presented in Section 8.2.3  will be compared with one another in their ability to achieve the operational objectives as outlined in Section 8.2.2 . the following criteria will be used in order to assess the relevant differences in achieving the specific policy objectives, categorized as follows:

Effectiveness: a governance structure should be able to catalyse additional investments, address specific barriers that limit investments in the modernisation of the energy sector and in increased energy efficiency in lower income Member States, including for small scale projects. Since a different set of instruments might be needed to address the market barriers in the different sectors, the implications of the options on the possibility to use grants, financial instruments or a mix of both would be considered.

Coherence: a coherent governance structure should appropriately align the interests of the institutions involved, while achieving transparency in the use of the funds. The coherence evaluates the extent to which the options present a governance structure that would give confidence to private investors. In this context, the transparency of the fund is evaluated against the possibility for the European citizens and the private sector to be informed about the setting up and operations of the Modernisation Fund.

Market distortion: Evaluate to what extent there is a risk of distorting the internal energy market.

Administrative burden: Minimize the complexity and the administrative burden of setting up and operating the Modernisation Fund. This includes consideration of the administrative burden for project promoters.

8.2.4.1.Effectiveness

The three options are expected to be more effective in modernizing the energy systems and improving energy efficiency in the lower income Member States compared to the baseline scenario as they channel specific funds towards these objectives.

Depending on the barriers faced, the different types of projects in the lower income Member States may require public intervention in the form of grants, financial instruments, or a mix of the two (see Annex 10 for discussion on the barriers in different sectors, including networks, energy efficiency, and power generation). The specific type of support provided through the Modernisation Fund would depend on the project eligibility that would be determined at a later stage, depending on the chosen governance option. Based on this, a suitable form of public intervention would need to be determined based on the specific market needs. It may be needed to contribute part of the funding towards providing technical assistance facilities to speed up project preparation, and improve the quality of projects.

Grants are typically used to support non-revenue generating projects that at the same time advance the EU policy objectives 136 . Grants require a certain rate of co-financing by project promoters, so assuming a co-financing rate of 50%, grants would lead to 2 fold leverage on investment. However, this rate could vary so that the minimum needed support is provided. Financial instruments, on the other hand, can realise higher leverage and scale-effects and enable cost-effective use of public resources. Evidence from the use of financial instruments implemented by the Commission and the EIB in the energy sector shows that financial instruments, depending on how they are structured, can leverage public money in the range of 6 to 15 137  times. It is possible to also blend financing options.

The options on governance have some implications for the design of any future grants and/or financial instruments. Options 1 and 2 allow flexibility to balance between addressing sub-optimal investment situations that need grant support, and maximizing leverage by using the funds through financial instruments. On the other hand, under Option 3, the proceeds would be disbursed through grants so focus will likely be given to specific sub-sectors where providing grant support would not distort the internal energy market.

Under Option 1 the use of financial instruments would potentially be more limited than under Option 2. As discussed in Section 8.2.1.3 , the experience with the European Structural and Investment Funds indicates that the use of financial instruments in Member States may be limited due to bottlenecks in administrative capacity and in some cases market maturity. Therefore, under Option 1, the beneficiary Member States might opt for more traditional means of supporting investments such as grants. Under Option 2, the EIB can use its extensive experience with financial instruments and the expertise to ensure the effectiveness and competitiveness of such instruments. Therefore, once project eligibility is determined under this option it is more likely that the use of grants would be limited only to situations where financial instruments cannot help adequately.

Regarding the possible use of financial instruments, under Option 1, it is more likely that national and regional promotional banks would be more involved in structuring the financial instruments, which, in turn, could lead to a more fragmented approach resulting in different structures and financial terms offered in the different Member States for similar projects. This may result in a suboptimal level of risk sharing and diversification. On the other hand, it would allow promotional banks in the beneficiary Member States to strictly tailor these instruments to the specificity of the domestic market, in particular to support small scale projects.

On the other hand, in Option 2 the management role of the EIB could allow applying a harmonised approach across the beneficiary Member States. Standardised financial instruments have the advantage to offer consistent financial terms to project promoters and intermediary financial institutions. They could also contribute to a more optimal level of risk-sharing, which can be achieved with a larger portfolio of projects with different risk profile across the beneficiary Member States. The EIB could play a pivotal role in providing more assurances to project promoters and ensuring that evaluation and selection are performed ensuring value for money. Regarding small scale projects, under Option 2 it will be necessary to work with financial intermediaries 138 . It allows to aggregate together similar projects of smaller size across the beneficiary Member States to build a critical mass, diversify the risks and make such projects more attractive to private investors. Specific targeted calls for proposals could be organised for small scale projects (e.g. less than €5 million) under Option 3, as well as for awarding grants under Options 1 and 2.