EUROPEAN COMMISSION

Brussels, 8.11.2017

SWD(2017) 650 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Accompanying the document

Proposal for a Regulation of the European Parliament and of the Council setting emission performance standards for new passenger cars and for new light commercial vehicles as part of the Union's integrated approach to reduce CO2 emissions from light-duty vehicles and amending Regulation (EC) No 715/2007 (recast)

{COM(2017) 676 final}
{SWD(2017) 651 final}

Table of Contents

Glossary - acronyms and definitions Glossary - acronyms and definitions

1 1 introduction

1.1 1.1 Policy context

1.2 1.2 Legal context

1.3 1.3 Evaluation of the implementation

2 2 WHAT IS THE PROBLEM AND WHY IS IT A PROBLEM?

2.1 2.1 What is the nature of the problem? What is the size of the problem?

2.1.1 2.1.1 Problem 1: Insufficient uptake of the most efficient vehicles, including low and zero emission vehicles, to meet Paris Agreement commitments and to impr ove air quality, notably in urban areas

2.1.2 2.1.2 Problem 2: Consumers miss out on possible fuel savings

2.1.3 2.1.3 Problem 3: Risk of losing the EU's competitive advantage due to insufficient innovation in low- emission automotive tech nologies over the long term

2.2 2.2 What are the main drivers?

2.2.1 2.2.1 Driver 1: Consumers value upfront costs over lifetime costs

2.2.2 2.2.2 Driver 2: Consumers' concerns regarding zero emission vehicles (ZEV)

2.2.3 2.2.3 Driver 3 : EU standards do not provide enough incentive for further efficiency improvements and for the deployment of low and zero emission vehicles for the period beyond 2021, leading to uncertainty over future policy

2.2.4 2.2.4 Driver 4: Effectiveness of standards is reduced by growing 'emissions gap'

2.2.5 2.2.5 Driver 5: Road transport activity is increasing

2.3 2.3 Who is affected and how?

3 3 WHY SHOULD THE EU ACT?

3.1 3.1 The EU's right to act

3.2 3.2 What would happen without EU action?

3.3 3.3 Analysis of subsidiarity and added value of EU action

4 4 OBJECTIVES

5 5 WHAT ARE THE VARIOUS OPTIONS TO ACHIEVE THE OBJECTIVES?

5.1 5.1 Emission targets (level, timing and metric)

5.1.1 5.1.1 CO 2 emission target level (TL)

5.1.1.1 5.1.1.1 CO 2 target level for passenger cars (TLC)

5.1.1.2 5.1.1.2 CO 2 target level for vans (TLV)

5.1.2 5.1.2 Timing of the CO 2 targets (TT)

5.1.3 5.1.3 Metric for expressing the targets 37

5.2 5.2 Distribution of effort (DOE) 38

5.3 5.3 ZEV/ LEV incentives 41

5.3.1 5.3. 1 Context 41

5.3.2 5.3.2 Policy options 47

5.3.2.1 5.3.2.1 LEV definition (LEVD) 47

5.3.2.2 5.3.2.2 Type and level of incentive (LEVT) 48

5.4 5.4 Elements for cost-effective implementation 50

5.4.1 5.4.1 Eco-innovations (ECO) 50

5.4.2 5.4.2 Pooling (POOL) 54

5.4.3 5.4.3 Trading (TRADE) 54

5.4.4 5.4.4 Banking and borrowing (BB) 55

5.4.5 5.4.5 Exemptions and derogations 56

5.4.5.1 5.4.5. 1 'De minimis' exemptions' and 'small volume' derogations 57

5.4.5.2 5.4.5.2 Niche derogations for car manufacturers (NIC) 58

5.5 5.5 Governance 59

5.5.1 5.5.1 Real-world emissions (RWG) 60

5.5.2 5.5.2 Market surveillance (conformity of production, in service conformity) (MSU) 63

6 6 WHAT ARE THE ECONOMIC/EMPLOYMENT, ENVIRONMENTAL AND SOCIAL IMPACTS OF THE DIFFERENT POLICY OPTIONS AND WHO WILL BE AFFECTED? 66

6.1 6.1 General methodological considerations 66

6.2 6.2 Consistency with previous analytical work 68

6.3 6.3 Emission targets: me tric, level and timing 69

6.3.1 6.3.1 Metric for expressing the targets 69

6.3.1.1 6.3.1.1 Option TM_WTW: metric for setting the targets based on Well-to-Wheel approach 69

6.3.1.2 6.3.1.2 Option TM_EMB: metric for setting the targets based on embedded emissions 70

6.3.1.3 6.3.1.3 Option TM_MIL: metric for setting the targets based on mileage weighting 72

6.3.2 6.3.2 Target levels for cars (TL C) and vans (TLV) 73

6.3.2.1 6.3.2.1 Introduction 73

6.3.2.2 6.3.2.2 Economic impacts (including employment) 75

6.3.2.3 6.3.2.3 Social Impacts 99

6.3.2.4 6.3.2.4 Environmental impacts 101

6.3.3 6.3.3 Timing of the targets (TT) 109

6.3.3.1 6.3.3.1 Option TT 1: The new fleet-wide targets start to apply in 2030. 109

6.3.3.2 6.3.3.2 Option TT 2: New fleet-wide targets start to apply in 2025, and stricter fleet-wide targets start to apply in 2030. 110

6.3.3.3 6.3.3.3 Option TT 3: New fleet-wide targets are defined for each of the years until 2030. 110

6.4 6.4 Distribution of effort (DOE) 111

6.4.1.1 6.4.1.1 Methodology and introduction 111

6.5 6.5 ZEV/ LEV incentives 117

6.5.1 6.5. 1 Introduction and methodological considerations 117

6.5.2 6.5.2 Passenger cars: assessment of options with additional incentives for l ow-emission vehicles 118

6.5.2.1 6.5.2.1 Economic impacts 119

6.5.2.2 6.5.2.2 Social Impacts 128

6.5.2.3 6.5.2.3 Environmental impacts 129

6.5.3 6.5.3 Vans: assessment of options with additional incentives for low-emission vehicles 131

6.5.3.1 6.5.3.1 Economic impacts 132

6.5.3.2 6.5.3.2 Social Impacts 134

6.5.3.3 6.5.3.3 Environmental impacts 135

6.5.4 6.5.4 Macroeconomic impacts, including employment, of setting LEV incentives for cars and vans 137

6.5.4.1 6.5.4.1 Introduction and met hodological considerations 137

6.6 6.6 Elements supporting cost-effective implementation 142

6.6.1 6.6.1 Eco-innovations (ECO) 142

6.6.1.1 6.6.1.1 Future review and possible adjustment of the cap on the eco-innovation savings (Option ECO 1) 142

6.6.1.2 6.6.1.2 Extend the scope of the eco-innovation regime to include mobile air-conditioning (MAC) systems including a future review and possible adjustment of the cap on the eco-innovat ion savings (Option ECO 2) 143

6.6.2 6.6.2 Pooling (POOL) 143

6.6.2.1 6.6.2.1 Change nothing (Option POOL 0) 143

6.6.2.2 6.6.2.2 An empowerment for the Commission to specify the conditions for open pool arrangements (Option POOL 1) 144

6.6.3 6.6.3 Trading (TRADE) 144

6.6.3.1 6.6.3.1 Change nothi ng (Option TRADE 0) 144

6.6.3.2 6.6.3.2 Introduce trading as an additional modality for reaching the CO 2 targets and/or LEV mandates (Opti on TRADE 1) 144

6.6.4 6.6.4 Banking and borrowing (BB) 145

6.6.4.1 6.6.4.1 Change nothing (Option BB 0) 145

6.6.4.2 6.6.4.2 Banking only (Option BB1) 146

6.6.4.3 6.6.4.3 Banking and borrowing (Option BB 2) 147

6.6.5 6.6.5 Niche derogations for car manufacturers (NIC) 147

6.6.5.1 6.6.5.1 Change nothing (Option NIC 0) 147

6.6.5.2 6.6.5.2 Set new derogation targets for niche manufacturers (Option NIC 1) 148

6.6.5.3 6.6.5.3 Remove the niche derogation (Optio n NIC 2) 149

6.7 6.7 Governance 150

6.7.1 6.7.1 Real-world emissions (RWG) 150

6.7.1.1 6.7.1.1 Change nothing (Option RWG 0) 150

6.7.1.2 6.7.1.2 Option RWG 1: Collection, publication, and monitoring of real world fuel consumption data 150

6.7.2 6.7.2 Market surveillance (conformity of production, in service conformity) (MSU) 151

6.7.2.1 6.7.2.1 Option MSU 0 – no change 151

6.7.2.2 6.7.2.2 Option MSU 1: Obligation to report deviations and the introduction of a correction mechanism 152

7 7 Comparison of options 153

7.1 7.1 Emission targets 156

7.1.1 7.1.1 Emission target - Metric 156

7.1.2 7.1.2 Emission targets – timing 156

7.1.3 7.1.3 Emission targets – level for cars 157

7.1.4 7.1.4 Emission targets – level for vans 159

7.2 7.2 Distribution of efforts (cars and vans) 159

7.3 7.3 ZEV / LEV incentives 160

7.3.1 7. 3.1 ZEV / LEV incentives for cars 160

7.3.2 7.3.2 ZEV/LEV incentives for vans 163

7.4 7.4 Elements supporting cost-effective implementation 164

7.4.1 7.4.1 Eco-innovations 164

7.4.2 7.4.2 Pooling 164

7.4.3 7.4.3 Trading 164

7.4.4 7.4.4 Banking and borrowing 164

7.4.5 7.4.5 Niche derogations for car manufacturers 165

7.4.6 7.4.6 Simplification (REFIT aspects) 165

7.5 7.5 Governance: real-world emissions – market surveillance 165

8 8 How would impacts be monitored and evaluated? 167

8.1 8.1 Indicators 167

8.2 8.2 Operational objectives 168

List of figures

Figure 1: Overview of interlinkages between this initiative and other climate, energy and transport related initiatives at EU level Figure 1: Overview of interlinkages between this initiative and other climate, energy and transport related initiatives at EU level

Figure 2: Drivers, problems and objectives Figure 2: Drivers, problems and objectives

Figure 3: GHG emissions from cars and vans (1990-2015) Figure 3: GHG emission s from cars and vans (1990-2015)

Figure 4: Historical fleet CO2 emissions performance and current standards (gCO2/km normalized to NEDC) for passenger cars (ICCT, 2017) Figure 4: Historical fleet CO 2 emissions performance and current standards (gCO 2 /km normalized to NEDC) for passenger cars (ICCT, 2017)

Figure 5: EU-wide fleet target level trajectories for new cars under the different TLC options Figure 5: EU-wide fleet target level trajectories for new cars under the different TLC options

Figure 6: EU-wide fleet target level trajectories for new vans under the different TLV options Figure 6: EU-wide fleet target level trajectories for new vans under the different TLV options

Figure 7: Battery costs (USD/kWh) and battery energy density (Wh/L) Figure 7: Battery costs (USD/kWh) and battery energy density (Wh/L)

Figure 8: Evolution of Li-ion battery costs (USD/kWh) Figure 8: Evolution of Li-ion battery costs (USD/kWh)

Figure 9: Projected trend of greenhouse gas emissions from road transport between 2005 and 2030 under the baseline Figure 9: Projected trend of greenhouse gas emissions from road transport between 2005 and 2030 under the baseline

Figure 10: Passenger car fleet powertrain composition (new cars) in 2025 and 2030 under different TLC options Figure 10: Passenger car fleet powertrain composition (new cars) in 2025 and 2030 under different TLC options

Figure 11: Van fleet powertrain composition (new vans) in 2025 and 2030 under different TLV options Figure 11: Van fleet powertrain composition (new vans) in 2025 and 2030 under different TLV options

Figure 12: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/car) Figure 12: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/car)

Figure 13: TCO-15 years (vehicle lifetime) (net savings in EUR/car in 2025 and 2030) Figure 13: TCO-15 years (vehicle lifetime) (net savings in EUR/car in 2025 and 2030) 79

Figure 14: TCO-first user (5 years) (net savings in EUR/car in 2025 and 2030) Figure 14: TCO-first user (5 years) (net savings in EUR/car in 2025 and 2030) 80

Figure 15: Final energy demand (ktoe) for passenger cars over the period 2020-2040 under different TLC options Figure 15: Final energy demand (ktoe) for passenger cars over the period 2020-2040 under different TLC options 81

Figure 16: Share (%) of different fuel types in the final energy demand for cars (entire fleet) under different TLC options - 2025 and 2030 Figure 16: Share (%) of different fuel types in the final energy demand for cars (entire fleet) under different TLC options - 2025 and 2030 82

Figure 17: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/van) Figure 17: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/van) 86

Figure 18: TCO-15 years (vehicle lifetime) in 2025 and 2030 (net savings in EUR/van) Figure 18: TCO-15 years (vehicle lifetime) in 2025 and 2030 (net savings in EUR/van) 87

Figure 19: TCO-first user (5 years) in 2025 and 2030 (net savings in EUR/van) Figure 19: TCO-first user (5 years) in 2025 and 2030 (net savings in EUR/van) 88

Figure 20: Final energy demand (ktoe) for vans over the period 2020-2040 under different TLV options Figure 20: Final energy demand (ktoe) for vans over the period 2020-2040 under different TLV options 89

Figure 21: Share (%) of different fuel types in the final energy demand for vans (entire fleet) under different TLV options –2025 and 2030 Figure 21: Share (%) of differen t fuel types in the final energy demand for vans (entire fleet) under different TLV options –2025 and 2030 90

Figure 22: TCO-second user (years 6-10) (EUR/car) – 2025 and 2030 Figure 22: TCO-second user (years 6-10) (EUR/car) – 2025 and 2030 100

Figure 23: TCO-second user (years 6-10) (EUR/van) – 2025 and 2030 Figure 23: TCO-second user (years 6-10) (EUR/van) – 2025 and 2030 101

Figure 24: (Tailpipe) CO2 emissions of passenger cars in EU-28 - % reduction compared to 2005 Figure 24: (Tailpipe ) CO 2 emissions of passenger cars in EU-28 - % reduction compared to 2005 102

Figure 25: Cumulative (tailpipe) 2020-2040 CO2 emissions of cars for EU-28 – emission reduction from the baseline (kt) Figure 25: Cumulative (tailpipe) 2020-2040 CO 2 emiss ions of cars for EU-28 – emission reduction from the baseline (kt) 103

Figure 26: (Tailpipe) CO2 emissions of vans in EU-28 - % reduction compared to 2005 Figure 26: (Tailpipe) CO 2 emissions of vans in EU-28 - % r eduction compared to 2005 105

Figure 27: Cumulative (tailpipe) 2020-2040 CO2 emissions of vans for EU-28 – emission reduction from the baseline (kt) Figure 27: Cumulative (tailpipe) 2020-2040 CO 2 emissions of vans for EU-28 – emission reduction from the baseline (kt) 105

Figure 28: Evolution of GHG emissions between 2005 (100%) and 2030 under the EUCO30 scenario and under the baseline and different policy options for the CO2 target levels for new cars and vans considered in this impact assessment Figure 28: Evolution of GHG emissions between 2005 ( 100%) and 2030 under the EUCO30 scenario and under the baseline and different policy options for the CO 2 target levels for new cars and vans considered in this impact assessment 107

Figure 29: Additional manufacturing costs (EUR/car) for categories of passenger car manufacturers under different options DOE and with the EU-wide fleet CO2 target levels as in option TLC30 Figure 29: Additional manufacturing costs (EUR/car) for categories of passenger car manufacturers under different options DOE and with the EU-wide fleet CO 2 target levels as in option TLC30 114

Figure 30: Additional manufacturing costs relative to vehicle price (% of car price) for categories of passenger car manufacturers under different options DOE and with the EU-wide fleet CO2 target levels as in option TLC30 Figure 30: Additional manufacturing costs relative to vehicle price (% of car price) for categories of passenger car manufacturers under dif ferent options DOE and with the EU-wide fleet CO 2 target levels as in option TLC30 114

Figure 31: Additional manufacturing costs (EUR/van) for categories of van manufacturers under different options DOE and with the EU-wide fleet CO2 target levels as in option TLV40 Figure 31: Additional manufacturing costs (EUR/van) for categories of van manufacturers under different options DOE and with the EU-wide fleet CO 2 target levels as in option TLV40 115

Figure 32: Additional manufacturing costs relative to vehicle price (% of van price) for categories of van manufacturers under different options DOE and with the EU-wide fleet CO2 target levels as in option TLV40 Figure 32: Additional manufacturing costs relative to vehicle price (% of van price) for categories of van manufacturers under different options DOE and with the EU-wide fleet CO 2 target levels as in option TLV40 115

Figure 33: TCO-15 years (vehicle lifetime) (net savings in EUR/car for 2025 and 2030) for different LEV incentive options Figure 33: TCO-15 years (vehicle lifetime) (net savings in EUR/car for 2025 and 2030) for different LEV incentive options 121

Figure 34: Illustration of the impacts of option LEVT_CRED2 (net savings, TCO-15 years) in case the LEV benchmark is not exactly met (with the CO2 target of option TLC30 and the benchmark of option LEV%_A) Figure 34: Illustration of the impacts of option LEVT_CRED2 (net savings, TCO-15 years) in case the LEV benchmark is not exactly met (with the CO 2 target of option TLC30 and the benchmark of option LEV%_A) 124

Figure 35: Illustration of the impacts of option LEVT_CRED1 (net savings, TCO-15 years) in case the LEV benchmark is not exactly met (with the CO2 target of option TLC30 and the benchmark of option LEV%_A) Figure 35: Illustration of the impacts of option LEVT_CRED1 (net savings, TCO-15 years) in case the LEV benchmark is not exactly met (with the CO 2 target of option TLC30 and the benchmark of option LEV%_A) 125

Figure 36: TCO-second user (years 6-10) (EUR/car) in 2025 and 2030 for different LEVD/LEVT options Figure 36: TCO-second user (years 6-10) (EUR/car) in 2025 and 2030 for different LEVD/LEVT options 129

Figure 37: TCO- 15 years (EUR/van) in 2025 and 2030 for different LEVD/LEVT options Figure 37: TCO- 15 years (EUR/van) in 2025 and 2030 for different LEVD/LEVT options 133

Figure 38: TCO-second user (years 6-10) (EUR/van) in 2025 and 2030 for different LEVD/LEVT options Figure 38: TCO-second user (years 6-10) (EUR/van) in 2025 and 2030 for different LEVD/LEVT options 135

Glossary - acronyms and definitions

ACEA    Federation of European Car Manufacturers

BEV    Battery Electric Vehicle

CNG    Compressed Natural Gas

CO2    Carbon dioxide

EMIS    Emission Measurements In the automotive Sector (Committee of the European Parliament)

ESR    Effort Sharing Regulation

ETS    EU Emission Trading System

EV    Electric Vehicle: covers BEV, FCEV and PHEV

FCEV    Fuel Cell Electric Vehicle

FCM    Fuel Consumption Measurement

GDP    Gross Domestic Product

GHG    Greenhouse gas(es)

HDV    Heavy-Duty Vehicles, i.e. lorries, buses and coaches (vehicles of more than 3.5 tons)

HEV    Hybrid Electric Vehicle (not including PHEV)

ICEV    Internal Combustion Engine Vehicle

IEA    International Energy Agency

LCA    Life-Cycle Assessment

LCV    Light Commercial Vehicle(s): van(s)

LDV    Light-Duty Vehicle(s), i.e. passenger cars and vans

LPG    Liquified Petroleum Gas

LNG    Liquefied Natural Gas

MAC    Mobile Air Conditioning

NEDC    New European Driving Cycle

NGO    Non-Governmental Organisation

NOx    Nitrogen oxides (nitric oxide (NO) and nitrogen dioxide (NO2))

O&M    Operation and Maintenance

OECD    Organisation for Economic Co-operation and Development

OBD    On-Board Diagnostics

PHEV    Plug-in Hybrid Electric Vehicle

PM    Particulate matter

REEV    Range Extended Electric Vehicle (sub-group of PHEV)

SAM    Scientific Advice Mechanism

TLC    CO2 Target Level for passenger Cars (policy option)

TLV    CO2 Target Level for Vans (policy option)

TTW emissions    "Tank-to-wheel" emissions: emissions from the vehicle tailpipe that occur during the drive cycle of vehicles.

WLTP    Worldwide Harmonised Light Vehicles Test Procedure

WTT emissions    "Well-to-tank" emissions: emission occurring during fuel (incl. electricity, hydrogen) production and transport

WTW emissions    "Well-to-wheel" emissions: sum of TTW and WTT emissions

1introduction

1.1Policy context

In his State of the Union Address 2017 1 President Juncker put it very clearly: while the car industry is a key sector for Europe making world-class products, EU manufacturers will need to invest in the clean cars of the future in order to maintain their strong position. In addition, President Juncker stated "I want Europe to be the leader when it comes to the fight against climate change" and announced that "the Commission will shortly present proposals to reduce the carbon emissions of our transport sector".

The automotive industry is crucial for Europe's prosperity, providing jobs for 12 million people in manufacturing, sales, maintenance and transport and accounting for 4% of the EU's GDP 2 , including in sectors such as steel, aluminium, plastics, chemicals, textiles and ICT. The EU is among the world's biggest producers of motor vehicles and demonstrates technological leadership in this sector.

EU industry, in general, and the automotive sector, in particular, are currently facing major transformations. Digitalization and automation are transforming traditional manufacturing proceses. Innovation in electrified power trains, autonomous driving and connected vehicles constitute major challenges which may fundamentally transform the sector.

Furthermore, following the Paris Agreement 3 , the world has committed to move towards a low-carbon economy. Many countries are now implementing policies for low-carbon transport, including vehicle standards, often in combination with measures to improve air quality. These developments represent an opportunity for the EU automotive sector to continue to innovate and adapt in order to ensure it remains a technological leader.

The EU 2030 framework for climate and energy includes a target of an at least 40% cut in domestic EU greenhouse gas (GHG) emissions compared to 1990 levels. The emission reductions in the Emissions Trading System (ETS) and non-ETS sectors amount to at least 43% and 30% by 2030 compared to 2005, respectively. The Commission has recently proposed 2030 GHG emission reduction targets for Member States under the Effort Sharing Regulation 4 (covering the non-ETS sectors, including road transport) as well as a revised Energy Efficiency Directive 5 . CO2 standards for light-duty vehicles will help to meet the overall goals set out therein.

In addition to that, daily experience on traffic jams, the crisis over diesel cars emissions and the adoption of policy measures at local level to discorage car use in urban areas, have contributed to making EU consumers more aware of the impact of road transport on health and air quality.

These developments take place globally since nowadays automotive industries are increasingly integrated in global value chains. Global automotive markets are expanding faster than ever before, notably in emerging markets such as India and China. The latter, in particular, is taking full advantage of the changing automotive landscape and according to a recent report by the International International Energy Agency , in 2016 it became the country with the highest share of electric vehicles.

In addition, EU sales of passenger cars relative to global sales have decreased from 34% before the crisis (2008/2009) to 20% today. This means that EU industry will have to consider not only increasing exporting volumes but also adapting to changing demands which will require more focus on innovation to retain competitiveness.

Until now, the ambitious emission reduction standards in place in Europe have represented a fundamental tool to push for innovation and investments in low carbon technologies. But today, the EU is no longer the clear leader in this race, with the US, Japan, South Korea and China moving ahead very quickly.

As highlighted in the recently adopted Renewed Industrial Policy Strategy 6 , a modern and competitive automotive industry is key for the EU economy. However, for the sector to maintain its technological leadership and thrive in global markets, it will have to accelerate the transition towards more sustainable technologies and new business models. Only this will ensure that Europe will have the most competitive, innovative and sustainable industry of the 2030 and beyond.

The Commission's Communication 'Europe on the Move: An agenda for a socially fair transition towards clean, competitive and connected mobility for all' 7 makes clear that we want to make sure that the best low-emission, connected and automated mobility solutions, equipment and vehicles will be developed, offered and manufactured in Europe and that we have in place the most modern infrastructure to support them. The Communication identifies that profound changes in how we enjoy mobility are underway and that the EU must be a leader in shaping this change at a global level, building on the key progress already made.

This Communication builds on the earlier Commission's European Strategy for Low-Emission mobility 8 , published in July 2016, which set out an overall vision built on three pillars: (i) moving towards zero-emission vehicles; (ii) low emission alternative energy for transport; (iii) efficiency of the transport system.

The figure below presents an overview of the interlinkages between the various initiatives of the mobility package proposed by the Commission as well as other related EU climate, energy and transport related initiatives.

By pursuing an integrated approach looking both at the demand and supply side and by establishing an enabling environment and a clear vision and robust regulatory framework, the EU can create an environment that provides EU industry with the certainty and clarity needed to innovate and remain competitive for the future.

Figure 1: Overview of interlinkages between this initiative and other climate, energy and transport related initiatives at EU level

This builds on policies proposed or already implemented at national, regional and city level in the EU. Many Member States have set objectives to increase the share of zero and low emission vehicles, including both battery electric vehicles and plug-in hybrids, by 2020 9 .

However, while some Member States have made good progress in achieving their objectives, the majority of Member States has made rather slow progress 10 . Even if the objectives were to be reached, the share of electric vehicles would remain low in the EU in relation to total vehicle registrations. Furthermore, three Member States, representing 35% of total new car registrations in the EU in 2016, have announced plans to phase out CO2 emitting cars (see Table 1 ).

At the same time, many cities in the EU have implemented regulations which limit the access of certain vehicles to urban areas. Most restrictions are within the scope of so-called Low Emission Zones which either limit the city entry of the most polluting vehicles or, in some cases, impose higher fees for such vehicles if they enter the zone. Recently some cities have even announced plans to ban diesel and/or petrol cars (see Table 1 ).

Table 1: Overview of announcements at national and city level to encourage the use of zero- and low-emission vehicles

A policy framework that further stimulates the accelerated uptake of zero- and low-emission vehicles would complement the on-going efforts to address air quality problems and would be well aligned with on-going action at city, regional, and national level. Zero-emission vehicles do not only reduce CO2 emissions from road transport but deliver also in terms of air pollutant and noise emission free transport.

1.2Legal context

The EU has in place two Regulations setting CO2 targets for new passenger cars and vans, respectively, which are based upon Article 192 of the TFEU (Environment chapter):

·Regulation (EC) No 443/2009 setting a fleet-wide average target for new passenger cars of 130 g CO2/km from 2015 and 95g CO2/km from 2021, and

·Regulation (EU) No 510/2011 setting a fleet-wide average target for new light commercial vehicles of 175 g CO2/km from 2017 and 147 gCO2/km from 2020.

These regulations have been amended in 2014 through Regulation (EU) No 333/2014 and Regulation (EU) No 254/2014 in order to define the modalities for implementing the 2020/2021 targets.

Both Regulations request the Commission to carry out a review by the end of 2015, and to report on it to the Council and the European Parliament, accompanied, if appropriate, by a proposal to amend the Regulations for the period beyond 2020.

The abovementioned emission targets have been set on the basis of the New European Driving Cycle (NEDC) test cycle. From 1 September 2017 on, a new regulatory test procedure, the World Harmonised Light Vehicles Test Procedure (WLTP) 17 , developed in the context of the UNECE, has been introduced under the type approval legislation for determining the emissions of CO2 and the new targets will need to take this into account. Furthermore, consumer information on the fuel consumption and CO2 emission of new passenger cars under Directive 1999/94/EC should be based on WLTP as of 1 January 2019 18 .

1.3Evaluation of the implementation

An extensive evaluation of the existing Regulations was carried out as part of REFIT. This was completed in April 2015 and the final report of the consultants has been published 19 .

The evaluation report assessed the Regulations against the objectives set in the original legislation, which included providing for a high level of environmental protection in the EU and contributing to reaching the EU's climate change targets, reducing oil consumption and thus improving the EU’s energy security of supply, fostering innovation and the competitiveness of the European automotive industry and encouraging research into fuel efficiency technologies.

It concluded that the Regulations were still relevant, broadly coherent, and had generated significant emissions savings, while being more cost effective than originally anticipated for meeting the targets set. They also generated significant EU added value that could not have been achieved to the same extent through national measures. As regards impacts on competitiveness and innovation, the impacts of the Regulations were found to be generally positive.

Box 1 summarises the key outcomes in relation to the main evaluation criteria.

The evaluation report included some recommendations that would ensure the Regulations remain relevant, coherent, effective and efficient, including:

·With respect to relevance, a potential additional need to be considered for the post 2020 legislation is that road transport needs to use less energy. Hence, energy efficiency would become a more important metric as the LDV fleet moves to a more diverse mix of powertrains

·Concerning effectiveness, the most significant weakness identified was the current (NEDC) test cycle causing an increasing discrepancy between real-world and test cycle emissions, which has eroded a significant portion of the originally expected benefits of the Regulations. This will be largely addressed by the development of WLTP. In addition, sufficient checks are recommended to ensure that the new test does not in future years become subject to the same problems experienced with the NEDC.

·While the lack of consideration of the lifecycle and embedded emissions of vehicles was seen as a relatively minor issue, it was expected to become more significant as the proportion of electric vehicles increases.

·As regards additional incentives to develop low CO2 emission vehicles, it should be considered whether such mechanism is needed and, if so, to choose one that does not potentially weaken the target.

·A need to look at how to improve the ex-ante assessment of costs to manufacturers as the costs assumed prior to the introduction of the current Regulations were much higher than has been the case in reality.  

These recommendations are addressed when presenting the policy options in Section 5.

2WHAT IS THE PROBLEM AND WHY IS IT A PROBLEM?

Figure 1 sets out the drivers, problems and objectives that are relevant for the revision of CO2 standards for cars and vans.

While the revision will clearly contribute to all three policy objectives, it should also be clear that it does not aim to address all of the problems and drivers mentioned to the same extent. For this, complementary proposals and flanking measures will be taken, some of are scheduled to be part of the same package of mobility related initiatives. This concerns in particular the EU Action Plan on the Alternative Fuels Infrastructure Directive (limited infrastructure), the proposal for a revised Clean Vehicles Directive 2009/33/EC, as well as the proposal for a revised Directive on road charging ("Eurovignette").

The Commission is also preparing a proposal for setting CO2 standards for heavy-duty vehicles, which would further help to tackle CO2 emissions in the road transport sector.

Beside this, there are a number of areas where complementary Member State or local action would help to tackle the drivers and problems, e.g. through tax measures (in order to help lowering upfront costs, especially for zero- and low-emission vehicles), and measures promoting modal shift (i.e. lowering road transport activity.

A key driver to be addressed by this impact assessment is the lack of stringency of the existing CO2 standards for the period beyond 2021 and the related uncertainty over future standards. Other drivers are addressed to a different degree in the policy options set out in Section 5. Clarifying the policy framework beyond 2021 will help reducing manufacturers' uncertainty over costs and future investment decisions as well as tackling certain market failures. Creating a market demand for more efficient vehicles will also help to reduce upfront costs. In addition, the 'emissions gap' will be addressed.

By contrast, limited infrastructure and increasing transport activity are not directly tackled by the options considered in this impact assessment.

2.1What is the nature of the problem? What is the size of the problem?

An overview of the problems and drivers is presented in Figure 2 .

Figure 2: Drivers, problems and objectives

2.1.1Problem 1: Insufficient uptake of the most efficient vehicles, including low and zero emission vehicles, to meet Paris Agreement commitments and to improve air quality, notably in urban areas

The evaluation of the CO2 Regulations showed that the CO2 standards have stimulated the uptake of more efficient vehicle technologies, but it also highlighted that the CO2 performance of new vehicles needs to improve at a faster rate in order to achieve the Union's climate goals of at least 40% emissions reduction, as committed under the Paris Agreement, in a cost-effective way. As confirmed in the European Strategy for Low-Emission Mobility, greenhouse gas emissions from transport will need to be at least 60% lower than in 1990 and be firmly on the path towards zero. 20 With current trends in new vehicles' CO2 emissions, this cannot be achieved. More specifically, the uptake of LEV and ZEV is still very slow. In 2016, battery electric vehicles (BEV) and plug-in hybrid electric vehicles (PHEV) represented only 1.1% of the new EU car fleet (for BEV the share was only 0.41%). 21

Road transport was responsible for 22% 22 of EU GHG emissions in 2015 with a steady increase since 1990 when the share was 13%. GHG emissions from cars and vans accounted for 73% of road transport emissions in 2015; this share has remained more or less constant since 1990.

Figure 3 shows that CO2 emissions from cars and light commercial vehicles in 2015 were still 19% higher than in 1990, despite the decrease observed between 2007 and 2013. While the increase in the share of transport emissions of EU GHG emission may be due to the emissions reduction in other sectors, the evolution of GHG emissions from cars and vans shows a steady increase since 1990 with the exception of the period between 2007 and 2013 when emissions were reduced. 

Figure 3: GHG emissions from cars and vans (1990-2015) 23

While the transport sector has considerably reduced its emissions of air pollutants in the EU over the last decades, it is the largest contributor to NOx emissions (46% in total NOx emissions in the EU in 2014). Of the total emitted NOx from road transport, around 80% comes from diesel powered vehicles. In addition, the transport sector makes an important contribution to the concentration of particulate matter in the atmosphere (13% for PM10 and 15% for PM2.5). 24  

EU air quality legislation 25 sets limit and target values for the concentration of a range of harmful air pollutants in ambient air in order to limit the exposure of citizens. Today, the limit values for NO2 are being exceeded in over 130 cities across 23 Member States and the Commission has initiated legal action against 12 Member States. 

The public debate on the announcement of possible "diesel bans" in some major cities has significantly affected the share of diesel vehicles in new car registrations. For instance, in March 2017 a 5-year low in new diesel car registrations was recorded in France, Germany, Spain, and the UK. These Member States represent together almost 60% of new car registrations in the EU. 26

In the EU as a whole the share of diesel in new car registrations decreased from a peak of 53% in 2014 to 49% in 2016. At the same time the share of new petrol cars increased from 44% in 2014 to 47%. 27  

While urban access restrictions contribute to a shift from diesel to petrol with benefits in terms of lower air pollutant emissions, so far they have not triggered a significant increase in low- and zero-emission vehicles. Although new registrations of battery electric and plug-in hybrid vehicles increased by 46% by July 2017 compared to the same period in 2016, their share in total car registrations in the EU remains low at 1.2% of which 46% were battery electric vehicles 28 .

A policy framework that further stimulates the accelerated uptake of zero- and low-emission vehicles would therefore complement the on-going efforts to address air quality problems and would be well aligned with on-going action at urban, regional, and national level. Zero-emission vehicles do not only deliver benefits in terms of air pollutant and noise emission free transport but also contribute to the reduction of CO2 emissions from road transport.

2.1.2Problem 2: Consumers miss out on possible fuel savings

In understanding potential fuel savings for consumers, including initial and subsequent vehicle purchasers it is important to understand that the current average lifetime of a car is around 15 years 29 with several ownership changes. Consumers have benefitted from net savings over a vehicle's lifetime, although relatively few consumers consider fuel consumption when purchasing a new car 30 .

So far the increases in the purchase prices of more efficient vehicles, as a result of the CO2 standards, have been significantly lower than the fuel savings over the vehicle's lifetime.

According to the evaluation of the CO2 Regulations the additional purchase cost of a new car in 2013 was €183 higher compared with a 2006 vehicle due to measures to meet the CO2 standards. At the same time (discounted) fuel savings, as a result of the CO2 standards, were €1,336 for petrol cars and €981 for diesel cars over the vehicle's lifetime. Lifetime fuel expenditure savings have been lower than anticipated, primarily because of the increasing divergence between test cycle and real world emissions performance. However, even if this gap were to be reduced significantly by the introduction of the WLTP test cycle and additional governance measures (see section 5.5), there remains an important unused cost savings potential. If this potential were to be exploited through more stringent CO2 standards, consumers could benefit from even higher fuel savings. The savings are however spread differently across the vehicle's lifetime.

An analysis of second hand car and van markets and implications for the cost effectiveness and social equity of light-duty vehicles CO2 regulations 31 shows that subsequent owners of a vehicle, who on average belong to lower income groups, proportionally benefit more from fuels saving than first vehicle owners. The initial cost for the more efficient vehicle is borne by the first owner. This depends however strongly on the initial price premium for the more efficient vehicle.

2.1.3Problem 3: Risk of losing the EU's competitive advantage due to insufficient innovation in low- emission automotive technologies over the long term

The EU automotive sector is crucial to the EU economy, including in terms of the number of direct and indirect jobs it provides. It faces global competition in terms of sales to other markets and, increasingly, from non-EU manufacturers within the EU market. The import of motor vehicles to the EU has increased from 2.5 million vehicles in 2010 to 3.4 million motor vehicles in 2016, worth € 45.7 billion. 32  

The competitiveness of industry is also related to its capacity to innovate. Looking at the relationship between the regulatory standards and industrial innovation, the Evaluation study found that EU fuel efficiency standards for new cars and vans have proven to be a strong driver for innovation and efficiency in automotive technology. 33 These targets allowed the EU manufacturers to have a first mover competitive advantage which has been especially important as the EU automotive industry exported more than 6 million vehicles in 2016, worth €135 billion. 34  

However, as shown in Figure 4 , different fuel standards have progressively been implemented around the world, in countries including China, USA, South Korea, Mexico, Brazil and India. These international targets, moving over time towards the levels set in the EU, and coupled with the commitments made on climate change targets under the 2015 Paris Agreement, demonstrate the international demand for efficient vehicles.

Figure 4: Historical fleet CO2 emissions performance and current standards (gCO2/km normalized to NEDC) for passenger cars 35 (ICCT, 2017)

Major non-EU car markets have considered or are about to introduce more ambitious policies including measures to reduce pollutant emissions. In particular, in view of increasing the deployment of zero- and low emission vehicles, ambitious policies have been developed or recently adopted in car markets that are of particular importance for the EU car industry. In the US, the Californian "ZEV" standards to support the market deployment of battery electric, plug-in hybrid, and fuel cell vehicles have also been adopted by nine other States (29% of all new cars sold in the U.S. are sold in these 10 States) (see Box 2 for more details). 36 Eight US States have signed a memorandum of understanding committing to coordinated action to ensure that by 2025 at least 3.3 million pure battery electric vehicles, plug-in hybrid electric vehicles and hydrogen fuel cell electric vehicles are on their roads. 37  

In China, new mandatory "new energy vehicle" (NEV) requirements will apply to car manufacturers as from 2019 covering battery electric, plug-in hybrid, and fuel cell vehicles (see Box 3 for more details). 38 The requirements are applicable to all manufacturers with an annual production or import volume of 30,000 or more conventional fuel passenger cars.

Over the last decade China has become the key car market with 24 million new car registrations, meaning that every third new vehicle is now being sold in China. European car manufacturers have been successful in reaching out to this new market. More than 20% of new passenger cars sold in China were from European car manufacturers/joint ventures operating in China. One third of global sales by German manufacturers, i.e. around 15 million vehicles, took place in China: 39% for the VW Group and 22% for the BMW Group and Mercedes Benz Cars 39 . Similarly, China is the most important car market for the PSA Group with more than 600,000 vehicles sold 40 .

A recent analysis of seven global automotive lead markets concludes that China is now in the "pole position" and will dominate the increasing market for electrified powertrains for the foreseeable future due to the importance of the Chinese market and a favourable regulatory framework. 41  

While Japan alone accounts for 40% of EV related patents, the EU automotive industry is the global leader in automotive patents in general. 42 At the same time patents data show that parts of the European car industry have a strong technological potential in LEV/ZEV which are however not sufficiently reflected in new products offered on the European market. 43  

This indicates that the EU industry risks losing its technological leadership and lagging behind these global trends.

2.2What are the main drivers?

2.2.1Driver 1: Consumers value upfront costs over lifetime costs

There are a number of market failures and barriers 44 which cause end-users to not necessarily purchase the most efficient new vehicles available on the market, even where this would be their optimal choice from an economic perspective, i.e. when the fuel economy benefit outweighs the additional costs for a more efficient vehicle.

When purchasing a new car, end-users tend to undervalue future fuel savings as a result of which it may not appear attractive to pay more for a more efficient vehicle. This is for instance empirically evidenced by the results of the evaluation of the CO2 Regulations, which show that fuel savings are significantly higher than the additional purchase cost of a new car (see Section 1.3). Despite existing fuel taxes, these clear financial benefits were apparently not reaped by the market, but required specific regulation to tap into such economic benefits.

Furthermore, even if the new vehicle purchasers do take account of fuel savings, it would only be rational for them to consider fuel savings for the period in which they intend to own the vehicle. As vehicles have an average lifetime of about 15 years with 4 owners, only a part of the reductions would be experienced by the initial purchaser.

In addition, a wide range of factors and elements other than fuel economy may dominate the purchase decision of a new car. Purchasers of new cars have skewed preferences away from fuel economy and towards factors such as comfort and power. 45 Another reason for the apparently economically suboptimal uptake of more efficient vehicles therefore lies on the production side. In a highly competitive automotive market, manufacturers may be hesitant to invest heavily in more efficient powertrains, knowing that competitors may have different commercial strategies (focusing on other vehicle attributes such as higher engine capacity, more comfort, etc.) that could be commercially more successful. This is in particular the case if consumers pay little attention to total cost of ownership. A regulatory framework on CO2 emissions for all new vehicles takes away the competitive risk that a manufacturer would be facing when focusing innovation efforts on fuel efficiency, while others do not.

Different purchase dynamics may apply for leased vehicles which have a share of around 30% of new registrations in the EU, with most of them being company cars. Leasing could in principle increase the attractiveness of lower CO2 vehicles, on the one hand by enabling instant payback on fuel saving ‘investments’, and on the other by helping operators optimise vehicle choice by enabling them to better take into account the costs and benefits associated with lower CO2 vehicles in the context of CO2-based national vehicle taxation schemes. However, the extent to which these factors affect the uptake of lower CO2 vehicles in practice could not be quantified due to a current lack of evidence. 46  

2.2.2Driver 2: Consumers' concerns regarding zero emission vehicles (ZEV)

Beyond the issue of undervaluing future benefits from fuel savings, the limited market uptake of ZEV is strongly influenced by additional factors. ZEV (battery EV and fuel cell EV) are still faced with much higher upfront costs 47 as compared to conventional vehicles. 48  

Consumers are also concerned about other issues regarding ZEV. As demonstrated in research 49 , a major barrier is consumer resistance to new technologies that are considered alien or unproved. As other barriers perceived by the consumers, the study mentioned battery range, charging infrastructure, reliability, safety. Furthermore, the perceived limited comfort and style were seen as limiting the attractiveness of available ZEV models.

A key barrier is 'range anxiety', i.e. the perception that the battery capacity is limited and recharging infrastructure is insufficient to ensure recharging 'on time' and at the necessary recharging speed in particular for long-distance trips. This is underlined by the fact that the electric range for the most sold battery electric vehicles in the EU is currently between 150 and 250 km.

Despite important progress and sufficient coverage in most Member States given the low uptake of ZEV so far, the infrastructure for recharging ZEV is insufficient in many Member States in particular in view of the expected uptake of ZEVs by 2020 and beyond 50 . The Commission's Communication, 'Europe on the Move: An agenda for a socially fair transition towards clean, competitive and connected mobility for all' underlines that the deployment of a network of recharging points covering evenly the whole EU road network, is a key enabling condition for zero-emission mobility. The Action Plan on the Alternative Fuels Infrastructure Directive sets out concrete measures for achieving necessary deployment rates 51 . Experience from other regions shows that with an increase in the number of electric vehicles sold investments in the necessary infrastructure increases as well. Besides, reinforced support for research and development of batteries will be provided by Horizon 2020 in the context of the new working programme 2018-2020.

Another concern among consumers is linked to the resale value of ZEV given expected further technical improvements in particular on the battery's performance (range, lifetime, costs). 52  

At the same, the market for ZEV is developing rapidly. New technologies and business models may help to overcome some of the barriers discussed above. For example, new ZEV in the compact car segment are offered in Europe with ranges of up to 380 km 53 . Some ZEV are offered with a lease contract for the battery 54 which lowers upfront costs and can address possible consumer concerns related to the battery technology.

In this context, it should be noted that consumer research in the US and Germany showed that a large share of prospective new vehicle buyers (29% in the US, 44% in Germany) would consider purchasing a battery electric vehicle (BEV) or a plug-in hybrid electric vehicle (PHEV), which indicates a substantial latent demand for such vehicles. However, it was also found that half of all consumers are not yet familiar with electric vehicles. The researchers conclude that there is an opportunity for manufacturers to quickly increase the number of potential buyers by offering more tailored EVs and deploying new business models 55 . A JRC study covering six EU Member States 56 concluded in 2012 that on average around 40% of the car drivers surveyed would consider buying an electric car when changing their current vehicle 57 .

2.2.3Driver 3: EU standards do not provide enough incentive for further efficiency improvements and for the deployment of low and zero emission vehicles for the period beyond 2021, leading to uncertainty over future policy

The current Regulations for cars and vans set targets of 95 g CO2/km for 2021 and 147 g CO2/km for 2020 respectively. In the absence of new legislation, these targets will remain at their present levels. As the current targets can be largely met by improving conventional vehicles, they do not provide sufficient incentive to invest in and in particular market alternative powertrains, in particular ZEV.

As a consequence there is insufficient uptake of LEVs and ZEVs in the EU as a result of which the necessary GHG emission reductions in the road transport sector cannot be achieved. Given persisting market failures (see Driver 1) under these conditions manufacturers are not likely to develop, produce and offer more efficient vehicles for the EU market at sufficient scale. The EU automotive industry therefore risks losing leadership in low-emission technologies for road transport.

As long as the automotive industry, including manufacturers and suppliers, does not know what will happen to targets beyond 2020/2021 and whether any additional requirements will be put in place, they do not have the regulatory certainty required to invest with confidence for the EU market. Without clarity on the long-term regulatory framework companies cannot take long-term investment decisions in order to meet future market demands and optimise compliance costs.

2.2.4Driver 4: Effectiveness of standards is reduced by growing 'emissions gap'

There is evidence of an increasing divergence between average test and real world CO2 emissions. Recent studies estimate the divergence is up to around 40% 58 . A number of factors have been identified to explain the divergence including the deployment of CO2 reducing technologies delivering more savings under test conditions than on the road, the optimisation of the test procedure as well as the increased deployment of energy using devices which are not taken into account when a vehicle is tested for its certified CO2 emissions. For example, air conditioning systems are not included when a vehicle is tested for its certified CO2 emissions but are widely installed and used, thus leading to higher real world emissions.

This increasing divergence means that the actual CO2 savings achieved are considerably less than those suggested by the test performance. Since manufacturers' compliance with their specific emissions target is assessed on the basis of the CO2 emissions as certified during the official test cycle, the 'emissions gap' undermines the effectiveness of the CO2 performance standards. In addition, the 'emissions gap' has undermined consumers' trust in the potential CO2/fuel savings of new vehicles which in turn may have affected consumers' willingness to buy the most efficient vehicles.

2.2.5Driver 5: Road transport activity is increasing

EU transport activity is expected to continue growing under current trends and adopted policies, albeit at a slower pace than in the past 59 . Despite profound shifts in mobility being underway, such as shared mobility services and easier shifts between modes, passenger traffic growth is still projected to increase 23% by 2030 (1% per year) and 42% by 2050 (0.9% per year) relative to 2010. Road transport would maintain its dominant role within the EU. Passenger cars and vans would still contribute 70% of passenger traffic by 2030 and about two thirds by 2050, despite growing at lower pace relative to other modes due to slowdown in car ownership increase.

While this increased activity is reflective of economic growth, it brings with it negative impacts in terms of GHG emissions and air quality impacts, if no additional measures are taken. It remains to be seen to what extent other developments such as autonomous driving may affect road transport activity.

2.3Who is affected and how?

The users of vehicles, both individuals and businesses, are affected because they face the cost of the energy required to propel the vehicles. Reducing the vehicle's CO2 emissions will reduce the energy required and result in a cost saving to the user. The use of technology to reduce in-use GHG emissions has a cost which is expected to be passed on to the vehicle purchaser.

Citizens, especially those living in urban areas with high concentrations of pollutants, will benefit from better air quality and less associated health problems due to reduced air pollutant emissions, in particular when the uptake of zero-emission vehicles increases.

CO2 standards require vehicle manufacturers to reduce CO2 emissions as a result of which they will have to introduce technical CO2 reduction measures. In the short-term, this is likely to result in increased production costs and could affect the structure of their product portfolios. However, demand for low- and zero-emission CO2 vehicles is expected to increase throughout the world as climate change and air quality policies develop and other countries introduce similar or even more ambitious standards, manufacturers have an opportunity to gain first mover advantage and the potential to sell advanced low CO2 vehicles in other markets.

Suppliers of components and materials from which vehicles are constructed will be affected by changing demands on them. Component suppliers have a key role in researching and developing technologies and marketing them to vehicle manufacturers. Requirements leading to the uptake of additional technologies or materials (e.g. aluminium, plastics, advanced construction materials) may create extra business activity for them. While often overlooked, EU employment in the component supply industry is as large as in the vehicle manufacturing industry.

Suppliers of fuels are affected by reduced energy demand leading to less utilisation of existing infrastructure. If demand shifts to vehicles supplied with alternative energy sources, this may potentially increase the need for other types of infrastructure and create new business opportunities and challenges for electricity supply companies and network operators.

There may also be impacts for example in the need for or type of vehicle servicing. There will also be lower maintenance requirements for battery electric vehicles.

The production and maintenance of vehicles with an electrified powertrain will pose important challenges to the workforce in the automotive sector including manufacturers and component suppliers as well as repair and maintenance businesses. The workforce will need additional and/or different skills to deal with new components and manufacturing processes.

Other users of fuel and oil-related products (e.g. chemical industry, heating) are expected to benefit from lower prices if demand from the transport sector decreases. Sectors other than transport that emit GHGs will avoid demands to further reduce emissions to compensate for increased transport emissions. In so far as these sectors are exposed to competition, this will be important for their competitiveness.

3WHY SHOULD THE EU ACT? 

3.1The EU's right to act 

The Environment chapter of the Treaty, in particular Article 191 and Article 192 of TFEU, give the EU the right to act in order to guarantee a high level of environmental protection. As mentioned in Section 1.1, based on Article 192 TFEU, the EU has already acted in the area of vehicle emissions, including adopting Regulations (EC) 443/2009 and (EU) 510/2011 which set limits for CO2 emissions from cars and vans, and with implementing legislation on monitoring and reporting of data (Commission Regulation (EU) No 1014/2010 (cars) and Commission Implementing Regulation 2012/293/EU (vans)).

3.2What would happen without EU action?

EU fuel efficiency standards for new cars and vans have proven to be a strong driver for innovation and efficiency in automotive technology. These targets allowed EU manufacturers to have a first mover advantage and to increase exports globally. Without further action in this field, it will be difficult for the EU automotive sector to retain its leading role in global markets as developing innovation and cutting-edge technologies is the only way to maintain and strengthen European competitiveness.

With all major markets with the exception of China and India projected to stall in the future, it will be important for the EU to maintain or increase the share of high-quality and high-technology vehicles on third markets, notably in those markets that are likely to grow fast. (source GEAR 2030)

Besides, without further EU action in this field it is likely there would be little additional substantial CO2 reduction from new light-duty vehicles. There may be certain expectations that in view of the current CO2 requirements and expected regulatory action in this field in third countries to which European vehicles are exported, the fuel efficiency improvement of vehicles may continue somewhat beyond this rate. However, as seen in the EU in the period between 1995 and 2006 for cars, in the absence of the mandatory CO2 standard this progress is likely to be offset at least to some degree by the increase in power, size or comfort of new cars.

Some reduction in emissions from the overall fleet of light-duty vehicles would still be expected beyond 2021 due to the continuing renewal of the existing fleet with newer cars and vans meeting the 2020/21 CO2 standards. However, transport activity would continue to increase and the overall CO2 reductions would not be sufficient to reach the targets set by the European Council in the 2030 Climate and Energy Package or contribute sufficiently to the goals of the Paris Agreement.

3.3Analysis of subsidiarity and added value of EU action

EU action is justified in view of both the cross-border impact of climate change and the need to safeguard single markets in vehicles.

Without EU level action there would be a risk of a range of national schemes to reduce light duty vehicle CO2 emissions. If this were to happen it would result in differing ambition levels and design parameters which would require a range of technology options and vehicle configurations, diminishing economies of scale.

Since manufacturers hold differing shares of the vehicle market in different Member States they would therefore be differentially impacted by various national legislations potentially causing competitive distortions. There is even a risk that national legislation might be tailored to suit local industry.

This poor coordination of requirements between countries, even if all Member States were to establish regulatory requirements for new vehicle CO2 emissions, would raise compliance costs for manufacturers as well as weaken the incentive to design fuel efficient cars and LCVs because of the fragmentation of the European market. It is unlikely that Member States acting individually would set targets in an equally consistent manner as shown by the widely differing tax treatment of new cars across the EU. This means that greater benefits will be achieved for the same cost from coordinated EU action than would be achieved from differing levels of Member State action.

With action only at Member State level we would not benefit from the lower costs which would arise as a result of the economies of scale that an EU wide policy delivers. The EU light vehicle market is currently around 16 million vehicles per year. The largest Member State market is around 3 million vehicles per year. On their own, individual Member States would represent too small a market to achieve the same level of results and therefore an EU wide approach is needed to drive industry level changes.

The additional costs which would arise from the lack of common standards and common technical solutions or vehicle configurations would be incurred by both component suppliers and vehicle manufacturers. However, they ultimately would be passed on to consumers who would face higher vehicle costs for the same level of greenhouse gas reduction without coordinated EU action.

The automotive industry requires as much regulatory certainty as possible if it is to make the large capital investments necessary to maximise the fuel economy of new vehicles, and even more so for shifting to new primary energy sources. Standards provide this certainty over a long planning horizon and they could not be implemented with the same effectiveness and certainty at Member State level.

4OBJECTIVES

General policy objective

The general policy objective is to contribute to the achievement of the EU's commitments under the Paris Agreement (based on Article 192 TFEU) and to strengthen the competitiveness of EU automotive industry.

Specific objectives

1.Contribute to the achievement of the EU's commitments under the Paris Agreement by reducing CO2 emissions from cars and vans cost-effectively;

2.Reduce fuel consumption costs for consumers;

3.Strengthen the competitiveness of EU automotive industry and stimulate employment.

These three specific objectives are on equal footing.

The first one concerns the climate objective of the Paris Agreement. Further efforts are necessary for all Member States to meet their 2030 targets under the Effort Sharing Regulation. With road transport causing one third of non-ETS emissions and emissions increasing in the last few years, reducing CO2 emissions from cars and vans is of key importance.

Implementing the Paris Agreement requires the decarbonisation of the economy including of road transport. The Low-Emission Mobility Strategy has confirmed the ambition of reducing GHG emissions from transport by at least 60% by 2050, as initially set out in the 2011 Low-Carbon Economy Roadmap and Transport White Paper.

This cannot happen without a very high deployment of zero- and low-emission vehicles. Analysis has shown that by 2050, electrically chargeable vehicles need to represent about 68-72% of all light duty vehicles on the roads. This requires a significantly increasing uptake of zero- and low-emission vehicles already in 2030 as the new vehicles of 2030 will remain on the road until the mid-2040s.

The second specific objective is related to the consumer angle of the CO2 standards, aiming to create benefits for car and van users through the sales of more efficient vehicles.

The third specific objective relates to innovation, competitiveness (including fair competition amongst EU manufacturers) and employment. While the EU automotive sector has been very successful in advanced internal combustion engine vehicles world-wide, it will need to adapt to the ongoing global transitions in the area of mobility and transport in order to maintain its technological leadership.

By providing a clear regulatory signal and predictability for industry to develop and invest in zero- and low-emission vehicles and fuel-efficient technologies, this initiative aims to foster innovation and strengthen EU industry's competitiveness in a fast changing global automotive landscape, without distorting the competition between EU manufacturers.

In addition to the three abovementioned specific objectives, the revision of the CO2 standards for cars and vans are expected to lead to two main co-benefits: improvements in air quality and increased energy security.

5WHAT ARE THE VARIOUS OPTIONS TO ACHIEVE THE OBJECTIVES? 

This Section describes the options identified to address the problems listed in Section 3 and to achieve the objectives defined in Section 4. It sets out the rationale for their selection, as well as the reasons for discarding certain options upfront, taking into account the evaluation study, the public consultation, additional stakeholder input; as well as several internal and external study reports. The options cover a number of elements, some of which are already part of the current Regulations. The options are grouped into five categories:

(I)CO2 emission targets (level, timing, metric);

(II)the distribution of effort amongst manufacturers;

(III)incentives for low- and zero-emission vehicles;

(IV)elements for cost-effective implementation;

(V)governance related issues

The following tables show how the policy options, grouped into the five key policy areas, relate to the problems and objectives

Table 2: Policy options and problems

Table 3: Policy options and objectives

5.1Emission targets (level, timing and metric)

The currently applicable Regulations (EC) No 443/2009 ("Cars Regulation") and (EU) No 510/2011 ("Vans Regulation") set a fleet-wide target of 95 g CO2/km (from 2021, with a phase-in from 2020) and 147 g CO2/km (from 2020), respectively, for the emissions of newly registered vehicles. These targets are based on the NEDC test procedure. Compared to the targets set previously, they represent an average annual reduction of 5.1% for cars (from the 2015 target of 130 g CO2/km) and of 5.6% for vans (from the 2017 target of 175 g CO2/km).

The introduction of the new test procedure WLTP, in September 2017 60 , is expected to bring the tailpipe CO2 emissions from cars and vans determined during type approval closer to the real world emissions. The WLTP will be fully applicable to all new cars and vans from September 2019 (see also Section 5.5).

WLTP is likely to result in increased CO2 emissions for most vehicles but the increase will not be evenly distributed between different manufacturers. Due to this non-linear relationship between the CO2 emission test results from the NEDC and WLTP test-procedures, it is impossible to determine one single factor to correlate NEDC into WLTP CO2 emission values. A correlation procedure 61 will therefore be performed at the level of individual manufacturer. Based on the correlation procedures and the methodology adopted for translating the individual manufacturer targets from NEDC to WLTP values, WLTP-based manufacturer targets will apply from 2021 onwards. Those targets will be confirmed by the Commission and published in October 2022. 62  

More information on the transition from NEDC to WLTP is given in Annex 5.

5.1.1CO2 emission target level (TL)

The likely increase in WLTP CO2 emission values (compared to NEDC) has been taken into account for the purposes of the analytical work underlying this impact assessment (see Annex 4.6).

Since the exact specific WLTP emission target values for 2021 can only be determined in 2022 (as described above), the new emission targets should be defined not as absolute values but in relative terms. The starting point for this are the 2021 EU-wide fleet average WLTP emission targets (i.e. the weighted average of the manufacturers' specific emissions targets for 2021). The new targets can be expressed either as a percentage reduction of those 2021 EU-wide fleet targets or as an average annual reduction rate over a given period.

The options in this section for the new EU-wide fleet average target levels ("TLC" for cars and "TLV" for vans) are defining the target trajectory over the period 2021-2030, without prejudging the target years. Options as regards the timing of the targets are set out in Section 5.1.2.

5.1.1.1CO2 target level for passenger cars (TLC)

·Option TLC0: Change nothing (baseline)

This option represents the status quo, meaning that the CO2 target level set in the current Regulation is maintained after 2021 (WLTP equivalent of 95 g CO2/km as EU-wide fleet average).

·The other options for defining the EU-wide fleet CO2 target level for passenger cars are summarised in the below table.

Option TLC_EP40 differs from option TLC40 by defining a non-linear target trajectory. This covers the strictest end of the 2025 target range referred to in the Statement by the Commission in 2014 in the context of the negotiations on the Cars Regulation 63 . This also holds true for option TLC_EP50, which defines a 2030 target that is 50% lower than the 2021 target.

Figure 5: EU-wide fleet target level trajectories for new cars under the different TLC options 64

5.1.1.2CO2 target level for vans (TLV)

·Option TLV 0: Change nothing (baseline)

This option represents the status quo, meaning that the CO2 target level set in the current Regulation is maintained after 2021 (WLTP equivalent of 147 g CO2/km NEDC as EU-wide fleet average).

·The other options for defining the EU-wide fleet CO2 target level for light commercial vehicles are summarised in the below table.

Options TLV_EP40 and TLV_EP50 are defining a non-linear target trajectory, covering the strictest end of the 2025 target range referred to in the Statement by the Commission in 2014 in the context of the negotiations on the Vans Regulation 65 . For 2025, both options cover a WLTP target equivalent to 105 g CO2/km NEDC, while in 2030 the targets are 36%, respectively 50%, lower than the 2021 targets.

Figure 6: EU-wide fleet target level trajectories for new vans under the different TLV options 66

5.1.2Timing of the CO2 targets (TT)

The following options will be considered for defining the year(s) for which new targets are set. These options apply both for passenger cars (in relation to options TLC) and for light commercial vehicles (in relation to options TLV).

·Option TT 1: The new EU-wide fleet CO2 targets start to apply in 2030.

This means that the (WLTP equivalent of the) CO2 target levels set in the Cars and Vans Regulations would continue to apply until the year 2029.

·Option TT 2: New EU-wide fleet CO2 targets start to apply in 2025 and will continue to apply until 2029, and stricter EU-wide fleet CO2 targets start to apply from 2030 on.

Under this option, the new EU-wide fleet targets for 2025 and 2030 are calculated according to the annual average reduction rates set out in Section 5.1.1.

·Option TT 3: New EU-wide fleet CO2 targets are defined for each of the years 2022-2030.

Under this option, new annual EU-wide fleet CO2 targets are calculated according to the annual average reduction rates set out in Section 5.1.1

These options include a mid-term review to assess the effectiveness of the policy.

5.1.3Metric for expressing the targets

The CO2 targets set in the Cars and Vans Regulations relate to the tailpipe emissions of newly registered vehicles, applying the so-called Tank-to-Wheel approach (TTW). The targets are expressed in g CO2 /km and apply for the sales-weighted average emissions of the EU-wide fleet. For calculating the average, each newly registered vehicle is counted equally.

Using a TTW metric allows focusing on vehicle efficiency, which has proven to be an effective way of triggering the uptake of vehicle technology and starting a shift towards alternative powertrains. However, the overall GHG emission impact of using (new) vehicles is also affected by the type of fuel/energy used to propel the vehicle, as different energy types differ in the amount of CO2 emissions generated during their production, the so-called Well-To-Tank (WTT) emissions. The sum of the TTW emissions and the WTT emissions is referred to as the Well-To-Wheels (WTW) emissions.

Furthermore, there are also CO2 emissions associated with vehicle manufacturing (including the mining, processing and manufacturing of materials and components), maintenance and disposal. These are referred to as "embedded" CO2 emissions. For determining those emissions, information is needed concerning the different phases of a vehicle's life cycle and tools such as life-cycle assessment (LCA) are often used for this purpose.

The g CO2/km metric allows comparing the emission performance of vehicles on a unit distance basis, but this does not reflect the total emissions of a vehicle over its lifetime. Vehicles with a higher lifetime mileage may contribute more to total CO2 emissions compared to vehicles that are used less intensively, even where the latter perform worse against the g CO2/km targets.

The evaluation study noted that the effectiveness of the Cars and Vans Regulations might have been reduced because some of the emission reductions achieved in terms of tailpipe CO2 emissions may have been accompanied by increased emissions elsewhere.

During the public consultation, some stakeholders also suggested to switch to other metric types to express the targets, in particular by using one of the approaches mentioned hereafter.

Well-to-Wheel (WTW) based metric

In the public consultation, stakeholders representing the fuels industry as well as some component suppliers suggested a change from the TTW metric to a WTW based metric, which takes into consideration the sum of the TTW and WTT emissions in the CO2 target levels. By contrast, consumer organisations, car manufacturers and stakeholders from the power sector did not support such a change. Public authorities had mixed views.

Metric taking into account embedded emissions

In the public consultation, most car manufacturers were against changing to this approach, whereas other stakeholder groups had diverging views.

Metric based on mileage weighting

During the public consultation, the question whether average mileage by fuel and vehicle segment should be taken into account when establishing targets received very mixed replies from stakeholders. A number of environmental and transport NGOs, some research institutions, and all respondents from the petroleum sector were in favour of doing so. By contrast, one NGO and the majority of car manufacturers were against this option. Most consumer organisations were neutral on the issue, whereas public authorities expressed split views.

In the light of the above and the views expressed during the public consultation, the following options will be considered for defining the metric of the EU-wide fleet CO2 targets. These options apply both for passenger cars and for light commercial vehicles.

·Option TM_TTW: change nothing, TTW approach

This option maintains the current metric for setting the targets, i.e. targets expressed in g CO2/km based on a TTW approach and applying for the sales-weighted average EU-wide fleet emissions.

·Option TM_WTW: WTW approach

Under this option, the target would be expressed in g CO2/km based on a WTW approach and would apply for the sales-weighted average EU-wide fleet emissions.

·Option TM_EMB: metric covering embedded emissions

Under this option, the target would be expressed in g CO2/km covering both WTW and embedded emissions and it would apply for the sales-weighted average EU-wide fleet emissions.

·Option TM_MIL: metric based on mileage weighting

Under this option, the target would be set in relation to the mileage-weighted average EU-wide fleet emissions. It could either be expressed in g CO2/km or in different units reflecting the difference in lifetime mileage between vehicle groups.

5.2Distribution of effort (DOE)

The Cars and Vans Regulations use a limit value line to define the specific emission targets for individual manufacturers, starting from the EU-wide fleet targets. This linear curve defines the relation between the CO2 emissions and a "utility parameter" (currently: vehicle mass in running order 67 ).

On this line, the EU-wide fleet target value corresponds with the average mass of the new vehicles in the fleet (M0). The slope of the line is the key factor in distributing the EU-wide fleet target as it determines to what extent vehicles (manufacturers) with a higher/lower (average) mass will be allowed/required to have higher/lower CO2 emissions than the EU-wide fleet average. The steeper the slope, the larger the difference in specific emission targets between manufacturers with "heavy" and "light" vehicles.

In order to avoid that the EU-wide fleet targets would be altered due to an autonomous change in the average mass of the fleet, the M0 values are readjusted every three years to align them with the average mass of the new fleet of the previous years.

The choice of slope of the limit value line is merely a decision on how to share efforts amongst manufacturers and does not affect the overall emission target for the EU fleet of new vehicles.

Other approaches (e.g. using another or no utility parameter, changing the slope of the line, using a non-linear curve) are possible for distributing the effort required from each manufacturer in meeting the EU-wide fleet target. The Cars and Vans Regulations explicitly request the Commission to review this modality 68 .

Most car manufacturers and consumer organisations responding to the online consultation were in favour of using a utility parameter to distribute the effort between different manufacturers. A relatively large number of stakeholders across different stakeholder groups were neutral on this question, and only a small number of stakeholders (from different groups) were against the use of a utility parameter. Views diverged on which utility parameter to use. All consumer organisations, some environmental and transport NGOs as well as stakeholders from the petroleum sector supported footprint 69 , while most car manufacturers supported mass as utility parameter. Only two stakeholders referred explicitly to another parameter (loading capacity, in the case of light commercial vehicles).

The Association of European automobile manufacturers suggested a slightly different approach for cars and vans. While maintaining a single linear curve for cars with a mass-based utility parameter (i.c. WLTP test mass 70 ), for vans they proposed to switch to a curve consisting of two linear parts with different slopes, arguing that this would better take account of the large variety in design of light commercial vehicles.

In view of this, the following options are being considered:

·Option DOE 0: Change nothing

Under this option the linear limit value curves as defined in the current Regulations are maintained. The utility parameter applied is the mass in running order and the slope of the curves is 0.0333 (cars) and 0.096 (vans). The adjustment of the M0 value takes place every three years.

·Option DOE 1: mass based limit value curve with a slope representing an equal reduction effort for all manufacturers

Under this option, the manufacturer specific emission targets would be derived from the EU-wide fleet target according to a limit value line with the mass of the vehicles as the utility parameter.

The slope of the limit value line would be determined so that it results in an equal reduction effort for all manufacturers – starting from 2021 - according to the given utility value 71 . Two variants will also be considered as part of the assessment, one using the WLTP test mass as utility parameter (instead of mass in running order) and one using a combination of two different slopes for vans (taking account of the vehicle characteristics within the lighter and heavier segments).

·Option DOE 2: footprint based limit value curve with a slope representing an equal reduction effort for all manufacturers

Under this option, the specific emission targets would be derived from the EU-wide fleet target according to a limit value line using the vehicle footprint (i.e. wheelbase multiplied by track width) as the utility parameter. The approach for defining the slope would be the same as under option DOE 1, but using footprint data instead of mass data. 

For options DOE 1 and DOE 2, other sub-options (with different slopes) had initially been considered, but were not withheld as they would either lead to unwanted effects (in case of higher slopes) or are very close to the other options explored (esp. DOE 4 in case of lower slope).

·Option DOE 3: same target for all manufacturers ("uniform target")

Under this option, the EU-wide fleet target would apply for each individual manufacturer and no utility parameter would be applied 72 . As the specific emission targets under the current Regulations vary according to the average mass of the new vehicles registered by a manufacturer, the (percentage) emission reductions required to meet the future targets would be larger for manufacturers having a higher average vehicle mass than for those having lighter vehicles.

·Option DOE 4: equal reduction percentage for all manufacturers

As in option DOE 3, no utility parameter would apply in this case. The same emission reduction percentage would be required for each manufacturer, taking its specific emissions target in 2021 as the starting point. Therefore, the future specific emission targets (in g/km) would differ amongst manufacturers, depending on their 2021 WLTP target 73 .

Under options DOE 3 and DOE 4, the future manufacturer specific emissions targets would not be affected by future changes in the average value of the utility parameter for that manufacturer's new vehicles (mass or footprint).

5.3ZEV/ LEV incentives

5.3.1Context

The transition to low- and zero-emission mobility is subject to a number of policy discussions. At the informal meeting of the Environment and Transport Ministers in Amsterdam in April 2016, Member States supported this transition and underlined the opportunities it creates 74 .

The May 2017 Communication, 'Europe on the Move: An agenda for a socially fair transition towards clean, competitive and connected mobility for all' 75 confirms that EU-wide carbon dioxide emissions standards are a strong driver for innovation and efficiency and will contribute to strengthening competitiveness and pave the way for zero and low-emission vehicles in a technology-neutral way. It also stated that options under review include specific targets for low and/or zero-emission vehicles.

The Communication builds on the earlier Commission's European Strategy for Low-Emission mobility 76 , published in July 2016, in which the Commission highlighted the important role of zero- and low-emission vehicles in delivering CO2 reductions, particularly in view of the longer-term decarbonisation objectives. Furthermore, the Commission stressed that accelerating the ongoing shift to low-emission mobility will offer major opportunities for the European automotive and other sectors to drive global standards and export their products. Fostering a domestic lead market for such vehicles is relevant from a competitive perspective, in order to create (1) economies of scale to drive down costs and (2) a competitive edge for European manufacturers and component suppliers.

The battery is a major cost component of a BEV with battery costs making up to 55% in the price of a mass manufactured BEV in 2016 77 . According to external studies, a broad range of EV support policies applied worldwide 78 is expected to contribute to a drastic reduction in cost of electric vehicles over the next decade as battery manufacturing gets cheaper 79 . Those cost reductions are however highly reliant on mass manufacturing. Analysts argue that policy is therefore critical in this respect and fuel economy regulations will play an important role in driving the scale-up in EV manufacturing over the next 5-7 years 80 .

Figure 7 summarises information available up to 2016 on the costs and volumetric energy densities of batteries currently being researched, as well as the ranges of cost reductions that can be expected from the three main families of battery technologies: conventional lithium ion; advanced lithium ion, using an intermetallic anode (i.e. silicon alloy-composite); and technologies going beyond lithium ion (lithium metal, including lithium sulphur and lithium air) 81 . Figure 8 illustrates the evolution of Li-ion battery costs (in USD/kWh) in the past decade (showing a decrease of around 70% since 2010) and a forecast of their further evolution towards 2030, based on expected demand 82 , 83 .

Figure 7: Battery costs (USD/kWh) and battery energy density (Wh/L)

Source: OECD/IEA (2017) Global EV Outlook 2017

Figure 8: Evolution of Li-ion battery costs (USD/kWh)

Source: Bloomberg New Energy Finance (2017)

In addition, the narrowing cost gap between electric cars and ICEV may put pressure on governments to gradually revise their support measures, phasing out incentives in cases where BEVs and PHEVs actually rival ICEV costs. According to a report by OECD/IEA, other regulatory instruments (such as including fuel economy regulations and local measures, such as differentiated access to urban areas) will remain important in supporting the electric car uptake needed to meet the targets characterising a low-emission future 84 .

Regulatory incentives might thus be needed to help overcome the barriers to the market uptake of ZEVs and LEVs.

The vehicles incentivised should have a significant potential contribution to reducing the CO2 emissions of the new car and van fleet. The types of vehicle most relevant in this respect are the following:

·Battery electric vehicles (BEV) and fuel cell electric vehicles (FCEV), both having zero tailpipe CO2 emissions and a limited market uptake so far.

·Plug-in hybrid electric vehicles (PHEV) with sufficiently low tailpipe CO2 emissions.

In their replies to the public consultation, a majority of stakeholders across all stakeholder groups was in favour of some mechanism to encourage the deployment of LEV/ZEV, except for consumer organisations which were mostly neutral on whether and how LEVs/ZEVs should be incentivised. Environmental and transport NGOs were mostly in favour of a flexible mandate, differentiating between LEV and ZEV and allowing trading among manufacturers. European car manufacturers argued for considering broader policy issues such as grid management, infrastructure and taxation policy.

The European Automobile Manufacturers Association (ACEA) is opposed to sales mandates for LEV/ZEV as it considers the market uptake to be mainly driven by public incentives, in particular fiscal measures, which would give car manufacturers limited control to meet such mandates. 85 They also refer to experience in markets with existing mandates where customers are not willing to buy LEV/ZEV. Car manufacturers also point to the need to increase the number of publically available charging points which does not fall under their responsibility.

At the same time, over the past few years, several major car manufacturers have been announcing their global ambitions for the sales of electric cars, which would result in a strongly increasing deployment of those vehicles in the following years. Table 4 summarises a number of those announcements.

Table 4: List of manufacturer's announcements on electric car ambition (adapted from 'Global EV outlook 2017' (OECD/IEA, 2017) 86 )

*Note: Chinese manufacturers include BYD, BJEV-BAIC Changzhou factory, BJEV-BAIC Qingdao factory, JAC Motors, SAIC Motor, Great Wall Motor, GEELY Auto Yiwu factory, GEELY Auto Hangzhou factory, GEELY Auto Nanchong factory, Chery New Energy, Changan Automobile, GAC Group, Jiangling Motors, Lifan Auto, MIN AN Auto, Wanxiang Group, YUDO Auto, Chongqing Sokon Industrial Group, ZTE, National Electric Vehicle, LeSEE, NextEV, Chehejia, SINGULATO Motors, Ai Chi Yi Wei and WM Motor.

Despite this willingness by manufacturers to strongly expand their offer of EVs, the IEA 92 argues that at this stage of the electric car market deployment, policy support remains "indispensable for lowering barriers to adoption". In this context the IEA notes that mandates in combination with targets provide a clear signal to manufacturers and customers.

As a follow-up to the EMIS Inquiry Committee, the European Parliament 93 in April 2017 called on the Commission to fully engage in and implement a low-emission mobility strategy and "to come forward with a draft regulation on CO2 standards for the car fleets coming onto the market from 2025 onwards, with the inclusion of Zero-Emission Vehicles (ZEV) and Ultra-Low Emission Vehicles (ULEV) mandates that impose a stepwise increasing share of zero- and ultra-low-emission vehicles in the total fleet with the aim of phasing out new CO2-emitting cars by 2035".

A regulatory instrument to enhance the uptake of LEV has been established since the early 1990s in California with the "ZEV Regulation", which requires manufacturers to market a certain percent of vehicles with (near-)zero tailpipe emissions (see Box 2) 94 . Similar mandates also apply in nine other States of the US 95 . In September 2017, China adopted new energy vehicle (NEV) mandates (see Box 3) for the sales of electric cars, which, combined with government and local incentives for customers, manufacturers and the development of infrastructure, have seen a very strong growth in the past few years. Most recently, Quebec has adopted a ZEV mandate 96 . In the light of this policy context, the Impact Assessment is considering several options described below.

5.3.2Policy options 

·Option LEV 0: Change nothing

This option assumes that, apart from the fleet-wide CO2 emission targets, the legislation will not include provisions, which would specifically aim to increase the number of ZEV or LEV registered. The assessment of this option will therefore be based on the assessment of the TLC and TLV options

For the other policy options for incentivising ZEV/LEV, three key elements are considered: (i) the definition of a low-emission vehicle, as this determines the scope of the incentive, and (ii) the type of incentive and (iii) the level of the LEV incentive.

In addition, elements related to the implementation of the incentive need to be considered, such as compliance assessment (incl. the link with the CO2 target), differentiation between OEMs and between different types of LEV.

5.3.2.1LEV definition (LEVD)

In order to identify which vehicles would qualify for the LEV incentive, it is necessary to define what constitutes a LEV. This requires consideration of the metric and threshold to be used.

An option initially considered was to use the zero emission range of a vehicle (in km) for defining a LEV. However, this approach was not considered further, as the link of this metric with CO2 emissions is less outspoken, and only limited data is available to decide on an appropriate WLTP value. This view is also supported by the majority of stakeholders from different stakeholder groups which clearly preferred the use of CO2 emission performance as the criterion for defining LEV, with proposed thresholds ranging from 15g CO2/km to 50g CO2/km.

Therefore, as regards the CO2 emission threshold, only the options for defining a LEV according to its tailpipe CO2 emissions will be further considered, as summarised in Table 5 .

Table 5: Options considered for the LEV definition (LEVD)

The higher threshold for vans under option LEVD_40 compared to cars (LEVD_25), is explained by their larger average mass compared to cars and by the uncertainty over the feasibility of bringing a sufficient number of PHEV vans with emissions below 25 g CO2/km to the market.

For option LEVD_50, the 50 g CO2/km threshold is the same one as set in Article 5 of the Cars and Vans Regulations for vehicles to be eligible for generating super-credits. With the change from NEDC to WLTP, type approval emissions from PHEV with emissions around 50 g CO2/km are not expected to change significantly (see Annex 4.6).

However, covering such a broad range of vehicles without any further distinction would not take account of the expected improvement in battery efficiency and the corresponding decrease of CO2 emissions from PHEV. Furthermore, the actual performance of PHEV on the road is strongly influenced by the type and duration of trips undertaken, external conditions (temperature) and consumer behaviour (charging, use of electric equipment).

Therefore, a distinction is proposed under this option between ZEV and other LEV, by counting each LEV in relation to its CO2 emissions. While each ZEV would thus count as one vehicle, all other LEV would count as less than one vehicle, according to the following formula: .

In this way, the incentive is targeted towards vehicles having near-zero emissions, which avoids over-incentivising PHEVs with a short electric range.

5.3.2.2 Type and level of incentive (LEVT)

Additional regulatory tools for incentivising the uptake of ZEV/LEV currently used are mostly based on a ZEV/LEV sales mandate (e.g. California) and/or a crediting system, through increasing the weighting of a ZEV/LEV in the calculation of average emissions or providing emission credits based on the sales share of qualified vehicles.

Under the current Cars and Vans Regulations, a "super-credit" modality has been established to incentivise manufacturers to produce vehicles emitting less than 50 g CO2/km. During a limited number of years, such vehicles may be counted as more than one vehicle for the purpose of calculating the average specific emissions of a manufacturer.

For cars, super-credits applied between 2012 and 2015 in relation to the 130 g CO2/km target and will again apply (with lower multipliers) between 2020 and 2022 in relation to the 95 g CO2/km target (with a cap of 7.5 g CO2/km per manufacturer over the three years). For vans, super-credits only apply between 2014 and 2017 in relation to the 175 g/km target (for a maximum of 25,000 vans over that period).

However, as already highlighted in the impact assessment underlying the 2012 proposals for amending the Cars and Vans Regulations 103 , a super-credit system has significant drawbacks as it reduces the stringency of the CO2 target and thus the effectiveness of the Regulations in reducing CO2 emissions. The increase of CO2 emissions depends inter alia on the multiplier used and the number of eligible vehicles. For example, with a multiplier of 3.5 (which was applicable in the Cars Regulation in 2012-2013 and in the Vans Regulation in 2014-2015), CO2 emissions could increase by 3% to 15% depending on the proportion of vehicles qualifying for super-credits.

The evaluation study confirmed that super-credits could potentially weaken the targets, but noted that this had not yet materialised (in 2015) in view of the very low uptake of vehicles emitting less than 50 g CO2/km and as all major manufacturers were meeting their targets at that time even without taking super-credits into account.

However, as the share of vehicles with low emissions is expected to increase over time, maintaining the super-credit modality, as included in the Cars and Vans Regulations, would bear a high risk of weakening the CO2 target.

This analysis is confirmed in recent studies 104 , 105 which highlight the substantial environmental cost of electric vehicle multipliers or super-credits, in particular as the share of low-emission vehicles in the fleet starts to increase. Super-credits are seen as a counterproductive long-term vehicle policy. As an example, it is calculated that with an electric vehicle penetration at 28% of new vehicle sales in Europe, the regulation would lose 41% of its intended CO2 benefits when allowing super credits. Furthermore, as CO2 targets get stricter, super-credits could even discourage the further deployment of LEVs after 2020 due to the multiple counting. Maintaining even a small multiplier of 1.33 (the lowest value used in the current Cars Regulation) could cause the market uptake of LEV to be reduced by 6-7% by 2030.

Finally, by applying a multiplier from the first LEV registered on, the current super-credits system fails to send a clear signal to manufacturers and authorities about the expected share of LEV in the fleet.

The main drawbacks of the super-credit system could be mitigated or overcome by redesigning it into a crediting system, which would incentivise the uptake of LEV beyond a given level and would avoid undermining the CO2 target levels.

In view of the above, the following three options are considered:

·Option LEVT_MAND: LEV mandate

Under this option, each manufacturer's new vehicle fleet would have to include at least a given share of LEV.

·Option LEVT_CRED1: LEV crediting system with one-way adjustment of the CO2 target

This option builds on and improves the current super-credits system. The LEV incentive would take the form of a crediting system in connection with a manufacturer's specific CO2 target. A benchmark would be defined for the share of LEV in the new fleet in a given year. The specific CO2 target of a manufacturer exceeding this LEV benchmark would be adjusted as follows: each LEV registration above the benchmark would be rewarded on a 1%/1% ratio, meaning that a manufacturer registering 1% more LEV than the benchmark would get a 1% less stringent CO2 target. The CO2 target adjustment would be limited to 5% in order to avoid it to be weakened too much. Assessing compliance would be done only against the CO2 target. Not meeting the LEV benchmark would have no consequences for this compliance assessment.

·Option LEVT_CRED2: LEV crediting system with two-way adjustment of the CO2 target

This option only differs from option LEVT_CRED1 in that a manufacturer not meeting the LEV benchmark level would have to comply with a stricter specific CO2 target. Again, each LEV registration below the benchmark would be counted at a 1%/1% ratio, meaning that a manufacturer registering 1% less LEV than the benchmark would get a 1% more stringent CO2 target. The CO2 target adjustment would also be limited to 5%. Not meeting the LEV benchmark would therefore be reflected in the compliance assessment through a more stringent CO2 target.

As regards the percentage of the new vehicle fleet serving as the LEV mandate (LEVT_MAND) or benchmark (options LEVT_CRED), three options are considered for cars, labelled LEV%_A, LEV%_B and LEV%_C, and two options for vans, labelled LEV%_A and LEV%_B. The values chosen for the LEV mandate/benchmark are incremental compared to the LEV shares in the new vehicle fleet under option LEV0, while taking account of recent announcements by vehicle manufacturers as regards their expected LEV share. This is further explained in Section 0.

The assessment will be based on applying the same LEV mandate/benchmark for all manufacturers. The option of differentiating between OEMs has been not been withheld.

5.4Elements for cost-effective implementation

5.4.1Eco-innovations (ECO)

Article 12 of the Cars and Vans Regulations provides manufacturers with the possibility to take into account CO2 reductions achieved by innovative technologies whose CO2 reducing effect cannot be demonstrated through the official test procedure. Vehicle manufacturers and component suppliers may apply for the Commission's approval of a technology as an eco-innovation, if it fulfils the following basic conditions:

·The supplier or manufacturer must be accountable for the CO2 savings achieved;

·The technologies must make a verified contribution to CO2 reduction;

·The technologies must not be covered by the standard test cycle CO2 measurement or by mandatory provisions covered by the so-called Union's integrated approach to reach 10 g CO2/km (Article 1 and Article 12(2)(c) of the Cars Regulation 106 , see below for more information).

Where an approved eco-innovation technology is fitted to a manufacturer's vehicles, the average specific emissions of that manufacturer may be reduced by the CO2 savings from applying that technology, up to a maximum of 7 g CO2/km per year.

The Commission is empowered to adopt detailed provisions on the application procedure, including on the implementation of the criteria listed above. So far, the Commission has adopted more than 20 decisions approving eco-innovations for use in cars, for instance LED lighting systems and more efficient alternators. No applications have yet been submitted with regard to vans.

Both the previous impact assessment 107 and the evaluation study concerning the Cars and Vans Regulations concluded that eco-innovations are effective and efficient as they help to reduce CO2 emissions at a lower cost than alternative options. While it could be argued that the stringency of the targets as measured on the official test procedure would be reduced by this modality, this effect is balanced by the delivery of ‘off cycle’ emission reductions which cannot be measured on the test procedure and by setting a cap on the contribution of those reductions to the target achievement.

During the public consultation, a very large majority of stakeholders across all stakeholder groups was in favour of taking account of CO2 emission reductions arising from eco-innovations. Moreover, the evaluation study concluded that there is evidence supporting that the introduction of the Regulations has had a positive impact on innovation through encouraging higher R&D, and the development and deployment of fuel efficient technologies in the market. A phase-out of the eco-innovation modality will therefore not be considered as an option.

The evaluation study as well as stakeholders have however raised the issue of the administrative burden linked to the application and certification of savings as an issue and have suggested that the eco-innovation regime could be simplified in order to ensure a wider up-take of eco-innovations in the EU fleet.

Under the Cars and Vans Regulations (Article 12), the Commission is empowered to adopt detailed provisions on the application process, through which it may address any issues related to the administrative burden for industry and/or authorities. The Implementing Regulations 108 set out the requirements for applications as well as for the certification by type approval authorities of the CO2 savings from the approved technologies.

A revision of the Implementing Regulations is currently underway, with a view of adapting it to the new test procedure WLTP, but also to introduce a number of simplifications without changing the robustness of the assessment of the applications or the certification of the savings. The revision includes consideration of the US approach of determining off-cycle technologies with pre-defined CO2 savings as well as the possibility for amending existing approval decision upon request by stakeholders or at the Commission's initiative.

In view of this, it can be concluded that the current concept of eco-innovations is both efficient in that approved innovations will reduce CO2 emissions and cost-effective in that their cost should be lower than alternative options, while not causing any significant adverse effects with regard to the stringency of the targets.

Moreover, the current design of the provisions provides the Commission with the necessary powers to address effectively the concerns raised by stakeholders and identified in the relevant studies with regard to the administrative burden.

Against that background, it is considered that the current design of the eco-innovation modality is fit for purpose and can be maintained for the period 2022 to 2030. However, two issues require further consideration: the cap for the CO2 savings and the current exclusion of mobile air-conditioning systems from being eligible as eco-innovations. Manufacturers have in the context of the introduction of the WLTP requested an increase in the 7 g CO2/km cap. Manufacturers as well as component suppliers have also called for including mobile air-conditioning systems in the eco-innovation regime, pending any further regulation of such systems under the type approval legislation.

Cap for the CO2 savings

The current eco-innovation regime includes a cap of 7 g CO2/km for the CO2 savings that may be taken into account for compliance purposes. The cap applies regardless of the target level and vehicle category concerned. Until now, the up-take of eco-innovations has been limited (less than 1 g CO2/km in average savings for the manufacturer with the highest number of eco-innovations). It is however expected that the amount of eco-innovation credits used by manufacturers will increase significantly towards the target years 2020-2021.

The 7 g CO2/km cap has been set by reference to the emissions tested on the NEDC, while the EU-wide fleet CO2 targets for the period 2022 to 2030 are to be based on the emissions measured on the new WLTP type approval test. By setting a cap on the eco-innovation savings, a balance is ensured between incentives given to efficiency improvements demonstrated on the official test procedure and those given for the development of more efficient and new technologies that are not covered by that test. That balance also takes into account the fact that the target level is set on the basis of the test procedure emissions only.

The majority of technologies that have already been approved as eco-innovations will continue to fall outside also the WLTP test and will thus still be eligible as eco-innovations. There is however still uncertainty with regard to the level of the savings that can be expected from those technologies within the new testing framework as well as for the potential for other off-cycle technologies.

Against that background, and in order to ensure a smooth transition from the NEDC to the WLTP testing conditions, it is proposed to maintain the cap at the level of 7 g CO2/km pending the availability of more information with regard to the level of eco-innovation savings under the new WLTP test procedure.

In order to be able to take into account the experience that will be gained from the implementation of that procedure in the next couple of years, it is appropriate to consider an option providing the Commission with an empowerment to review the level of the cap so as to ensure that incentives given to eco-innovations remain balanced and effective over time.

Mobile air-conditioning systems (MAC systems)

Under the Cars and Vans Regulations, measures that are covered by the so-called "integrated approach" as defined in the 2007 Commission Communication on A Competitive Automotive Regulatory Framework for the 21st Century 109 are not eligible as eco-innovations 110 . This includes, inter alia, MAC systems.

All measures related to this "integrated approach", with the exception of MAC systems, are subject to mandatory measures. This concerns tyre pressure monitoring systems, tyre rolling resistance limits, gear shift indicators, fuel efficiency standards for vans and the use of biofuels. Mandatory measures addressing the efficiency of MAC systems have not yet been introduced and the WLTP test procedure, developed in the context of the UNECE, will not cover such systems in a foreseeable future.

Different studies 111 have pointed to the absence of measures addressing the efficiency of MAC systems as a draw-back, considering that MAC systems are one of the most important energy consumers on board vehicles, representing an average increase in fuel consumption in the order of 9% 112 . Furthermore, these systems are becoming standard equipment in new vehicles. The share of new cars equipped with MAC has risen from around 10% in 1993 to 85 % in 2011 113 .

Against that background, it is appropriate to consider the option of incentivising more energy efficient MAC systems within the context of eco-innovations. More efficient MAC could reduce the overall fuel consumption by at least 1 or 2% 114 .

It should also be noted that the US has introduced an off-cycle regime, according to which manufacturers that provide efficiency improvements in MAC systems can generate CO2-efficiency credits. The credits generated by the use of efficient MAC systems represented an equivalent of around 1.9 g CO2 /km in 2014 and in 2015.

It is therefore proposed to consider the option of extending the scope of the eco-innovation regime to include MAC systems.

In view of the above, the following options are considered:

·Option ECO 0: Change nothing

·Option ECO 1: Future review and possible adjustment of the cap on the eco-innovation savings

This option would maintain the current provisions of Article 12 of the Cars and Vans Regulations but would introduce an empowerment for the Commission to review and, where found appropriate following an assessment, adjust the 7 g CO2/km cap set on the eco-innovation savings.

·Option ECO 2: Extend the scope of the eco-innovation regime to include MAC systems

This option would also maintain the provisions of Article 12 of the Cars and Vans Regulations including the empowerment to adjust the cap as described in ECO1 but would remove the exclusion of MAC systems from being eligible as eco-innovations. The design of the methodology for determining the efficiency of MAC systems would result from an application by a manufacturer or supplier which would have to be assessed and approved by the Commission.

5.4.2Pooling (POOL)

The current Regulations (Article 7) offer individual manufacturers the possibility to form a "pool" for the purposes of meeting their emission targets. Such agreement enables a group of manufacturers to be counted as one entity for the purpose of compliance with the joint target. This allows manufacturers to decide on the most efficient way of complying with the targets. All manufacturers covered by the scope of the Regulations, which have not been granted a derogation (see section 5.4.5), could be part of a pool.

Pooling has been extensively used under the current Regulations. In 2015, pooling was used by 49 car manufacturers, responsible for 81% of all new car registrations in that year and by 25 van manufacturers, responsible for 70% of all new van registrations in that year. Forming a pool has prevented several manufacturers from exceeding their individual specific emissions target (in 2015 this was the case for 23 car manufacturers and 4 van manufacturers, which were member of a pool) 115 .

The vast majority of pools have been formed by manufacturers belonging to the same group of connected undertakings. Independent manufacturers may also form pools, however, until now this possibility has been rarely used. A pool formed by independent manufacturers would, in accordance with competition rules, have to be open to the participation of any other manufacturer requesting to participate. This reduces somewhat the utility of such, so called "open", pools with regard to compliance planning.

In order to enhance pooling as an instrument for all manufacturers to reduce compliance costs, the conditions under which open pools may be formed by independent manufacturers and under which conditions another manufacturer may request to join an existing open pool could to be clarified. An option is therefore introduced whereby the Commission is empowered to complement the existing provision by developing specific criteria for the open pool arrangements, in particular with a view to address any relevant competition aspects.

In view of the above, the following options should be considered:

·Option POOL 0 – change nothing – current pooling regime

·Option POOL 1 – an empowerment for the Commission to specify the conditions for open pools arrangements

5.4.3Trading (TRADE)

Trading has been suggested as a complement to pooling in order to provide additional flexibility for manufacturers in meeting the targets. Trading would allow individual manufacturers (or pools) to trade credits depending on their performance. This means that when a manufacturer (pool) overachieves its specific CO2 emissions and/or LEV mandate, this would result in credits that could be sold to another manufacturer (pool), which would otherwise not meet its target.

The main distinction compared to pooling is that trading would not require an upfront decision by manufacturers on how to ensure compliance with the target. The decision to trade could take place only at the time the provisional performance of the manufacturer (pool) is known.

Just as for pooling, trading would support the meeting of the CO2 targets or a LEV mandate or a combination of both.

In the case of a LEV mandate (LEVT_MAND), different design options are possible, mainly in relation to how any LEV generating credits are being accounted for in relation to the CO2 targets 116 .

Under a LEV crediting system (options LEVT_CRED), a separation between LEV credits and CO2 emission credits would not be necessary. For example, a manufacturer that does not achieve the LEV benchmark would have to meet a more stringent CO2 target. If that leads to non-compliance with the CO2 target, the manufacturer would have to buy credits from a manufacturer overachieving on its CO2 target.

In light of the above the following options are considered:

·Option TRADE 0: Change nothing – no trading

·Option TRADE 1: Introduce trading as an additional modality for reaching the CO2 targets and/or LEV mandates

Under this option, individual manufacturers (or pools) (which do not benefit from a derogation) would be allowed to exchange CO2 and/or LEV credits on an 'ad hoc' basis. This would require the establishment of a register to ensure full transparency and accountability of all transactions among manufacturers.

Trading would be allowed for cars and for vans separately (not amongst them).

5.4.4Banking and borrowing (BB)

Banking and borrowing are mechanisms used in different regulatory environments setting policy targets for individual actors with the aim of increasing flexibility and therefore lowering the cost of compliance. The rationale is that the overall desired outcome should be achieved by a certain time, while acknowledging that the optimal route to that point may differ between actors.

For the LDV CO2 legislation, banking would mean that when in a given year the average specific CO2 emissions of a manufacturer (pool) are below its specific emissions target, the manufacturer (pool) can carry over the difference between its emissions and its target as CO2 credits for future compliance purposes. In case its average specific CO2 emissions exceed the specific emissions target in one of the following years, the manufacturer (pool) can offset these excess emissions with the ‘banked’ CO2 credits from preceding year(s).

Borrowing would mean that, in a given year, a manufacturer (pool) could comply with its CO2 target by 'borrowing' CO2 credits, which have to ‘paid back’ in subsequent years.

In order to ensure that the EU-wide fleet CO2 target set for a certain date is actually met, banking and borrowing needs to be limited. For the definition of such a limit, the timeline for the new CO2 emissions target(s) (options TT, see Section 5.1.2) is critical.

If new targets are only set in discrete years (e.g. 2025 and/or 2030), it would be necessary to define a target trajectory against which emissions in the intermediate years would be compared for the purpose of granting credits. This would avoid that too many credits are accumulated before 2025, respectively 2030, which otherwise would allow a manufacturer (pool) to significantly exceed the target and hence undermine the intended CO2 emission reductions for that time period. In case of annual CO2 targets, these would, by definition, provide for such a trajectory.

However, even if a trajectory is set, there may still be a risk of too many credits being accumulated over time. In order to prevent this, banking could be limited to certain time periods (e.g. 2025-2030) or even to one year (e.g. 2025, when overachieving the applicable target or the trajectory). In the latter case, credits could only be used for compliance with the 2030 target. Finally, the use of banked credits could be limited to the year 2030 and no credits could be used after that year (assuming that a target will remain in place in subsequent years).

Links with the LEV incentives

In case of option LEVT_MAND, the above considerations would equally be valid in relation to the LEV mandate.

However, the situation is different in case of a LEV crediting system (options LEVT_CRED) where compliance assessment is based on the CO2 target only and therefore already makes a link between the LEV benchmark and the CO2 target. Hence, under that option, banking would only be necessary in relation to the CO2 target.

In light of the above-mentioned considerations, the following options are considered:

·Option BB 0: Change nothing

Under this option, no banking or borrowing would be allowed.

·Option BB 1: Banking only

Under this option, banking of CO2 and/or LEV credits would be allowed, but no borrowing.

·Option BB 2: Banking and borrowing:

Under this option, both the banking and borrowing of CO2 and/or LEV credits would be allowed.

5.4.5Exemptions and derogations

The Cars and Vans Regulations acknowledge that CO2 targets should be determined differently for smaller manufacturers as compared to larger ones, taking account of their capability to meet such standards. The Regulations therefore contain the following derogations:

·A de minimis exemption (cars and vans), which was introduced in the legislation in 2014 for manufacturers responsible for less than 1,000 newly registered vehicles per year. This exempts small manufacturers, in many cases SMEs, from meeting a specific CO2 emissions target and hence from applying for a derogation, thus reducing administrative burden;

·Small volume derogations (cars and vans): manufacturers (or a group of connected undertakings) responsible for between 1,000 and 10,000 cars registered per year or between 1,000 and 22,000 vans registered per year can apply to the Commission for an individual target consistent with their reduction potential;

·Niche derogations (cars only): manufacturers (or a group of connected undertakings) responsible for between 10,000 and 300,000 cars registered per year can apply for an individual target in 2021, corresponding with a 45% reduction from their 2007 average emissions.

5.4.5.1    'De minimis' exemptions' and 'small volume' derogations

De minimis exemptions reduce compliance and administrative costs for small manufacturers which are in many cases SMEs. Since they are exempt from meeting a specific CO2 target they have no compliance costs for adapting their vehicles to meet CO2 standards. The evaluation study estimated that the exemption reduces the administrative burden for the eligible manufacturers by around € 25,000 per manufacturer. It also facilitates the market entry of new manufacturers whilst having no significant impacts on the CO2 reductions of the overall EU vehicles fleet. During the public consultation, small car manufacturers underlined the importance of this exemption, with no other stakeholders questioning it.

The evaluation study also identified the small volume derogations as a potential weakness, but also confirmed that its impacts in this respect had been relatively small. Most stakeholders also supported this derogation regime, although some environmental NGOs and public authorities were opposed.

In 2015/2016, 23 car manufacturers benefitted from this derogation, 18 of which had less than 1,000 registrations and could thus have benefitted from the de minimis exemption (many small manufacturers continue to apply for derogations since EU derogations are required to avoid penalties when selling vehicles on the Swiss market). Without a derogation (or exemption) all of these car manufacturers would have exceeded their specific emissions target.

Six van manufacturers (or pools) applied for this derogation in 2015/2016, three of which had less than 1,000 vans registered in these years and were thus eligible for the de minimis exemption  117 . Four other manufacturers, which were eligible for the derogation, did not apply for it as they met their 'default' (Annex I) target.

In considering possible options, it does not appear appropriate to completely exempt this group of manufacturers from meeting any CO2 targets in view of the emission reduction potential in this segment, including the introduction of alternative powertrains. On the other hand, applying the same targets as for large volume manufacturers, based on the limit value curve, would mean that the reduction effort imposed on the small volume manufacturers would be significantly higher compared to large volume manufacturers taking account of their capability to meet emission standards (e.g. smaller fleet, fewer models).

The options of complete exemption or applying the same targets as for large volume manufacturers are therefore not considered further.

While some manufacturers applying for the derogation have pointed to the administrative burden of the application procedure as an issue, it should be noted that the Commission is empowered to define the detailed provisions on the application procedure and assessment criteria. These concerns can effectively be addressed through a simplification of the current applicable rules which are defined under comitology 118 .

In view of the above, the impact assessment does not consider specific options to change the existing regime of de minimis exemptions and small volume manufacturers.

5.4.5.2    Niche derogations for car manufacturers (NIC)

The Cars Regulation allows a 'niche' car manufacturer to meet a fixed emission reduction percentage set in relation to its emissions in 2007 (25% reduction by 2015 and 45% by 2021) instead of the 'default' emission target according to the limit value curve (Annex I to the Regulation). It should be noted that the percentage emission reduction between the 2015 and 2021 'niche' derogation targets is the same as the one between the fleet-wide targets set in the Regulation for those years (130 g/km and 95 g/km, respectively).

In 2015/2016, eight manufacturers or pools were eligible for a niche derogation but only five have applied to the Commission. Four out of the eight 119 were below their 'default' (Annex I) specific emissions target in one or both years and so strictly speaking did not need a derogation to comply with the Regulation.

It results from the evaluation study that this derogation potentially weakens the delivery of CO2 emissions reductions. If all of the eligible manufacturers would apply for the derogation, the number of cars covered could then increase by up to five times 120 .

During the public consultation, car manufacturers supported the continuation of this derogation regime but a majority of environmental and transport NGOs as well as all consumer organisations were against it.

Taking into account those considerations, the following options will be considered:

·Option NIC 0: Change nothing

This would mean maintaining the current provisions of the Cars Regulation. As a result, the 'niche' manufacturers would have to continue to comply after 2021 with the current derogation target, i.e. 45% reduction from their 2007 average emissions.

·Option NIC 1: Set new derogation targets for 'niche' manufacturers

Under this option, new "niche" targets would be defined for the period post-2021 on the basis of the overall CO2 reduction targets defined for the EU-wide fleet (TLC, see Section 5.1.1). The starting point for the 'niche' manufacturers would be their specific emission target for 2021. This approach would be in line with the reduction pathway set in the current Regulations between 2015 and 2021.

·Option NIC 2: Remove the 'niche' derogation

Under this option, no 'niche' derogations would be foreseen. This would mean that the 'niche' manufacturers would be covered by the same rules as the larger manufacturers as regards the target levels (see Section 5.1.1.1), distribution of effort (see Section 5.2) and LEV/ZEV incentives (see Section 5.3.2).

5.5Governance 

The CO2 emission targets for LDVs are set and enforced using as reference a standardised type approval test, taking place in a laboratory. This approach is used worldwide and allows for comparability, reproducibility, verifiability and planning certainty. The effectiveness of the targets in reducing CO2 emissions in reality depends on the one hand on the representativeness of the test procedure with respect to average real-world driving, and on the other hand on the extent to which the vehicles placed on the market conform to the reference vehicles tested at type approval.

As highlighted in the evaluation report and in the opinion of the Scientific Advice Mechanism 121 , it is widely accepted that the currently used NEDC laboratory test is no longer representative of today's driving conditions and vehicle technologies. Evidence taken from a number of sources indicates a growing divergence over the past years, up to around 40%, between the certified emissions and the emissions of vehicles driven on European roads 122 .

Factors which have contributed to this divergence include: the deployment of CO2 reducing technologies delivering more savings under test conditions than on the road; exploitation of flexibilities in the test procedure; growing deployment of untested energy consuming devices; driver independent circumstances like weather, road conditions or trip types; driving style and driving modes 123 .

During the public consultation there was very strong support across all stakeholder groups for the Commission to explore the potential to further reduce the divergence between the test cycle and real world emissions. Only representatives of car manufacturers and one component supplier were against this. All stakeholder groups, except for car manufacturers, supported establishing additional driving tests to give values closer to real driving emissions.

Application of the WLTP, which is mandatory in the EU for all new car types from September 2017 and for all new cars and vans from September 2019, will result in more realistic CO2 values.

However, the longer-term effectiveness of the shift to WLTP in closing the gap will depend on the extent to which it will remain representative of real-world driving circumstances and on the degree to which it is enforced, including via market surveillance instruments.

The following sections set out the options considered in relation to these two governance aspects.

5.5.1Real-world emissions (RWG)

The effectiveness of technologies applied to reduce the CO2 emissions of vehicles is affected by the actual driving conditions. This effectiveness can therefore not be fully captured by a laboratory emissions test procedure, in particular given the rapid evolution of these technologies.

Therefore, it is generally accepted that the emissions determined through a test procedure differ from the actual emissions achieved in the real world 124 . As such, this is not problematic for designing CO2 targets relying on type approval values, as long as the expected divergence between the test procedure and the real world emissions can be estimated correctly (see Annex 6). However, for the CO2 targets to fulfil their objective, it is important that any remaining divergence under the test procedure does not increase over time.

This consideration also applies with regard to consumer information. Type approval values of CO2 and fuel consumption are used by consumers to compare different vehicles' performance in terms of fuel efficiency. In order avoid that consumers are misled with regard to the performance of vehicles, information on how type approval values compare to real world values should be readily available. Easy access to real world fuel and energy consumption data should contribute to achieve that and would also be an important step towards increased transparency and rebuilding consumer trust in the automotive industry as well as in the type approval system.

The following two options are considered with a view to address both the need to verify and ensure the representativeness of the new test procedure and to provide consumers with robust real world data on vehicle CO2 emissions and fuel consumption:

·Option RWG 0:    Change nothing

This option assumes that the new test procedure WLTP, its periodic revision and the (proposed) revision of the type approval testing 125 would be sufficient to ensure the representativeness of the test procedure over time, with a limited and stable divergence with respect to average real-world emissions. It also assumes that the CO2 and fuel consumption data resulting from the WLTP test would be sufficient in terms of consumer information.

·Option RWG 1:    Collection, publication, and monitoring of real world fuel consumption data

This option considers two main complementary sources: firstly, the collection by manufacturers of real world fuel and energy consumption data from new vehicles and their publication on-line or by other easily accessible means. Secondly, the monitoring and assessment by the Commission of the manufacturer data and, if appropriate, from national sources, such as from periodic technical inspections, with a view to continuously evaluate the representativeness of the WLTP.

The implementation of these measures would require an empowerment for the Commission to determine the conditions for the collection and publication of the data, inter alia taking into account relevant data protection requirements. This empowerment would enable the development of a methodology to access, monitor and evaluate on a regular basis the average real world CO2 emissions of the new vehicle fleet (and/or sub-fleets thereof) and determine how that evolves in comparison to the corresponding type approval values. The findings based on that evaluation would be an essential element to be considered in a review of the WLTP test procedure and, where necessary, of the CO2 emission standards.

These measures require the availability of relevant data on real world fuel and energy consumption which are described below.

Standardized 'fuel consumption measurement device'

The Commission is currently preparing an amendment, in the context of the type approval legislation, of the WLTP Regulation 2017/1151 to lay down an obligation for manufacturers to fit a standardized 'fuel consumption measurement device' in the new vehicles.

This measure is not covered as an option in this impact assessment as it concerns an obligation under type approval legislation through a comitology act. It should however be noted that the cost for cars to be equipped with standardised, accurate and accessible on-board fuel-consumption measurement devices is estimated to be very low - in the order of 1 euro per vehicle 126 . - and they already exist in today's vehicles 127 , 128 ,, but the information is not accessible Moreover, this enabling technology has already been mandated in California 129 as of 2019.

The data resulting from such fuel consumption measurements would provide a robust basis to verify the representativeness of the WLTP type approval emission values and monitor the gap. It would also provide consumers with reference real world data on the basis of which they can assess how their own fuel economy compares to the average real world fuel consumption. In addition, it would enable simplified on-road fuel consumption measurements on a large number of vehicles.

An empowerment would be required for the Commission to develop the necessary provisions for the collection of the data as well as for the conditions for access and publication. This approach is in line with the recommendation of the European Parliament (following the work of its EMIS Committee) 130 , the opinion of the Scientific Advice Mechanism 131 , as well as the technical assessment by the Commission's Joint Research Centre (see Annex 6).

Other data sources

In the absence of standardised on-board fuel sensors, real-world fuel consumption data can be gathered via self-reporting platforms or fleet operators 132 , 133 even though such data are subject to inherent bias. The gap can also be estimated using a simulation software like the Green Driving Tool developed by the Commission's Joint Research Centre 134 , 135 .

CO2 measurements are also performed at type-approval (ex-ante) as part of the Real Driving Emissions (RDE) procedure for pollutant emissions introduced gradually as of 2017 136 . Their measurement is necessary to validate the procedure itself. However, there is no evidence to date for the degree of representativeness of these data with respect to corresponding ex-post average real-world driving emissions, and there is a risk of bias and inconsistency across the tested vehicle types (see Annex 6).

Other option considered: elaboration of an ex-ante CO2 real-driving emissions procedure, including the determination of a not-to-exceed limit

In their response to the public consultation, some environmental and transport NGOs and car drivers associations suggested to develop a dedicated new RDE test protocol for CO2 emissions using Portable Measurement Equipment Systems (PEMS). In addition, binding not-to-exceed limits for CO2 emissions would be introduced. These not-to-exceed limits would be based on the difference between the emissions measured during the WLTP test cycle and the new RDE procedure for CO2 emissions. This would add another level of compliance checking, in a similar way as for air pollutant emissions.

The feasibility of such an approach is highly uncertain due in particular to the high variability in the CO2 emissions under real world conditions. RDE CO2 test results are strongly influenced by external factors, such as temperature, humidity, and driving behaviour. Consequently, the test results cannot offer the precision needed for regulatory purposes, such as target setting, compliance checking or for imposing financial penalties. In a laboratory test – such as the WLTP – such external factors can be controlled and the test values can as a consequence ensure the necessary legal certainty and precision 137 .

Custom-tailored test protocols of individual manufacturers (or groups) may provide more realistic fuel consumption and CO2 emission values than a laboratory test. They can provide useful information to consumers. However, such protocols rely on test data from a limited number of vehicle models and selected drivers, and make use of monitored real world emissions of these specific fleets. As a consequence, these test protocols are not exposed to the same variability or uncertainties as compared to a more generic protocol that would have to apply in an equivalent way and with similar accuracy to any vehicle on the EU market.

In view of the above, the elaboration of an EU-wide ex-ante CO2 real-driving emissions procedure at type-approval, including the determination of a not-to-exceed limit for the purpose of target setting and compliance checking does not appear feasible and is therefore discarded from further analysis.

5.5.2Market surveillance (conformity of production, in service conformity) (MSU)

As recommended by the European Parliament following the work of its EMIS Committee 138 and stressed by several consumer organisations and environmental NGOs 139 , it is necessary to put in place the means to detect irregularities in the CO2 and fuel consumption data.

Recent test campaigns performed by independent laboratories, have provided indications of CO2 emission values deviating significantly from the values determined at type approval 140 . Such deviations may undermine the achievement of the reduction objectives, distort competition among manufacturers and undermine consumer confidence in the type approval fuel consumption data.

Type approval tests are performed on a vehicle, which is representative of a certain vehicle family. The CO2 emissions of each vehicle produced that the manufacturer attributes to that family must conform to the emissions of the approved type. The manufacturer certifies this in a certificate of conformity which is issued as a condition for placing a vehicle on the market.

Each year, Member States report the CO2 emission values recorded in the certificates of conformity of the newly registered vehicles to the Commission. On that basis, the Commission determines the annual average specific emissions of a manufacturer for the purpose of checking compliance with the specific CO2 emissions target.

For the CO2 reduction objectives to be achieved, it is essential that the CO2 emissions of the vehicles placed on the market conform to the type approved values.

Under the type approval legislation, the conformity of the CO2 emissions is currently verified only at the stage of production. Some vehicles are selected from the production line by the manufacturer and tested to verify that the CO2 emissions are in line with those of the approved type. If this is not the case, the manufacturer has to take measures to bring the vehicles to be produced into conformity or perform a new type approval test.

A procedure for verifying the CO2 emissions of vehicles on the road, i.e. a so called in-service conformity test, is not yet in place. However, a proposal for setting up such a procedure is under discussion by the co-legislators 141 . In case in-service tests would not be retained in type approval legislation following the on-going co-decision process, an empowerment for the Commission to set up an independent testing of vehicles in use could be considered as part of this proposal.

In view of this, the following options are considered to ensure that the emissions of vehicles placed on the market continue to adequately reflect the CO2 emissions determined at type approval, to minimise the risk of deviations occurring and, if they occur, to ensure that the consequences for the CO2 reduction objectives can be adequately addressed.

·Option MSU 0: Change nothing

This option would mean that the CO2 monitoring provisions set out in the Cars and Vans Regulations 142 and the associated implementing legislation continue to apply. This allows the Commission to amend the CO2 monitoring data reported by Member States where manufacturers have found and notified errors in that data 143 . The verification by a manufacturer is voluntary and there is no explicit obligation placed on either manufacturers or Member States to report to the Commission deviations found from the type approved CO2 emission values.

·Option MSU 1: Obligation to report deviations and the introduction of a correction mechanism

Under this option, an obligation would be introduced in the legislation requiring Member States and manufacturers to systematically inform the Commission of any findings resulting from conformity of production tests or, where applicable, from in-service conformity tests, and inform of deviations from the type approved CO2 emissions affecting the monitored CO2 data.

The monitoring data for a manufacturer would be corrected in those cases where serious deviations from the type approval values have been detected which cannot be technically or otherwise justified. The empowerment would allow the Commission to define the way in which deviations may be detected and how these should be reported to the Commission as well as taken into account for the compliance checking. This could build on measures defined within the framework of the type approval legislation, or as an independent testing procedure to be defined under the CO2 regulations.

6WHAT ARE THE ECONOMIC/EMPLOYMENT, ENVIRONMENTAL AND SOCIAL IMPACTS OF THE DIFFERENT POLICY OPTIONS AND WHO WILL BE AFFECTED?

6.1General methodological considerations 

The quantification of the impacts, in particular as regards the target levels, distribution of effort and LEV incentives - see Sections 6.3.2, 6.4 and 0 - relies on a suite of models and a dedicated set of cost curves covering a broad range of up-to-date technologies for reducing CO2 emissions from cars and vans.

These cost curves, which show the CO2 reduction potential and costs for over 80 technologies, were determined as part of a study 144 on which car manufacturers, suppliers and other stakeholders provided input and were extensively consulted. The technologies considered include those that are currently already utilised in vehicles in the marketplace, as well as those expected to be available in the near future, and also options that have been proposed or are under development that could feasibly be introduced to the marketplace in the 2020-2030 period. Starting from a detailed assessment of these technologies, a total of 252 cost-curves on a WLTP basis was generated for different combinations of powertrain type (conventional, PHEV, BEV, FCEV), vehicle segment (four size classes for cars and three for vans) and year (2015, 2020, 2025 and 2030).

In the preparation of the cost curves, which represent a cost-optimal combination of technologies to be fitted in the vehicles to reach specific CO2 reduction levels, the possibility (or impossibility) to combine technologies has been duly taken into account, as has their pre-existing market penetration in the vehicles fleet, and overlaps in the CO2 saving potential of technologies when combined into packages.

In addition, for the purpose of analysing the sensitivity of cost assumptions apart from the "medium" costs, a number of cost-curves were developed illustrating the impact of low and high technology cost estimates. These different cost estimates were calculated using a methodological approach developed and refined in consultation with stakeholders and a statistical model to assess the uncertainty in the future cost projections. The "medium" cost case represents the most likely scenario resulting from significant future technology deployment to meet post-2020 CO2 targets. The projected future costs of BEV, PHEV and FCEV powertrains take into account economies of scale and potential rates of learning on the cost reduction of key components (i.e. notably batteries and fuel cells) in different market deployment scenarios. These costs have also been reviewed in the light of the more rapid than expected reductions in battery costs.

The PRIMES-TREMOVE model is used to project the evolution of the road transport sector. This model was consistently used for climate, energy and transport initiatives in the past, including for the 2016 Commission initiatives concerning the Effort Sharing Regulation (ESR), the Energy Efficiency Directive (EED), the Low-Emission Mobility Strategy, the Eurovignette Directive, as well as impact assessment on the Clean Vehicles Directive which was conducted in parallel. In addition, macro-economic models (GEM-E3, E3ME) and the DIONE model developed by the JRC have been used. All analytical models used are described in detail in Annex 4.

The baseline used for this impact assessment builds on the "Reference Scenario 2016" (Ref2016) 145 , which was used as the baseline for the ESR and EED proposals and the Low-Emission Mobility Strategy. In this scenario the market uptake of advanced technologies is estimated to remain rather low, not allowing for economies of scale, i.e. costs for these advanced technologies staying high.

The baseline includes a few policy measures adopted after the cut-off date of Ref2016 (end of 2014). Furthermore, some further differentiation in the model assumptions was needed in view of new information from specific studies, in particular:

(1)Updated cost curves were used, as explained above. The new costs are lower than the costs used as assumptions in Ref2016 and other previous analytical work performed with the PRIMES-TREMOVE model.

(2)Based on a recent JRC study 146 and other publications, a higher gap between emissions measured during NEDC testing and those in real driving conditions has been applied, on average about 37% for cars and 33% for vans 147 .

(3)The transition from the NEDC to the WLTP test cycle has been factored in by converting NEDC to WLTP emission values, using conversion factors derived by the JRC for this purpose (see Annex 4.5). For conventionally fuelled vehicles, these conversion factors are 1.21 for cars and 1.30 for vans, with specific values depending on the segments and powertrains.

Finally, the latest set of data from monitoring the implementation of the Cars and Vans Regulations (2015) has been used to properly reflect the current fleet composition and the turn-over rate for cars.

The baseline assumes that the EU-wide CO2 standards for the new passenger cars and vans fleets remain at the same level as in the current Regulations after 2020/2021 (i.e. 95 g CO2 /km for cars and 147 g CO2/km for vans). This would lead to a reduction of CO2 emissions in the period between 2020 and 2030 due to the renewal of the fleet and the reduction of the technology costs over time, which triggers the uptake of more efficient vehicles. However, in absence of new targets, the CO2 emissions reductions remain limited. Figure 9 shows that the GHG emissions from road transport are expected to decrease by 17% in 2030 with respect to 2005.

Figure 9: Projected trend of greenhouse gas emissions from road transport between 2005 and 2030 under the baseline

In particular, CO2 emissions from passenger cars and vans reduce by 26% and 17% respectively between 2005 and 2030. In a context of projected growing activity, these reductions are achieved due to the penetration in the fleet of more efficient vehicles. The monitored type-approval CO2 values of new passenger cars and vans decrease respectively by 14% and 11% between 2020 and 2030.

The resulting composition of the new passenger car and van fleet in 2025 and 2030 is shown in Section 6.3.2.1 in Table 6 (TLC0) and Table 7 (TLV0). The uptake of LEV remains limited, especially when considering that by 2050 the fleet share of these vehicles should be around 68-72% in view of the longer term emission reduction objectives.

A detailed description of the baseline projections is presented in Annex 4.

6.2Consistency with previous analytical work

A comparison was performed between different options for the CO2 targets for new cars and vans considered for the period after 2020 in this impact assessment and under the "EUCO30" scenario 148 , which is underlying several Commission climate, energy and transport policy proposals adopted in 2016. This scenario achieves the EU-wide 2030 targets regarding greenhouse gas emissions in the ESR sectors (a 30% reduction compared to 2005), and regarding final energy demand (27% renewable energy and 30% energy efficiency). The results are shown in Section 6.3.2.4.3 and in Annex 4.

6.3Emission targets: metric, level and timing 

6.3.1Metric for expressing the targets

6.3.1.1Option TM_WTW: metric for setting the targets based on Well-to-Wheel approach

The two main arguments most frequently used by stakeholders calling for a change from the current tank-to-wheel (TTW) to a well-to-wheel (WTW) metric mainly relate to the following aspects:

I.the need to account for the well-to-tank (WTT) emissions of electricity generation in comparison with those from fossil fuels, in particular in a context where the power sector is not yet fully decarbonised;

II.the need to acknowledge the role of low-carbon fuels like bio-ethanol, bio-methane or synthetic fuels produced from renewable electricity when setting reduction targets for CO2 emissions from cars and vans in order to incentivise the use of those fuels.

As regards the first argument, it needs to be remembered that greenhouse gas emissions from the power and refinery sectors in the EU are already covered by the EU emissions trading system (ETS). Furthermore, the power sector is also affected by measures to attain the Renewable Energy target.

With respect to the second argument, the Commission's 2016 RED-II proposal 149 sets mandates on the fuels sector for 2030. This means that EU policy is already in place for incentivising the deployment of renewable electricity and low-carbon fuels across all sectors, including transport.

Thus, moving to a WTW metric would de facto constitute double regulation for the fuels sector as well as the power sector. In the medium term, the impact of this double regulation on the emissions from those sectors in combination with other EU ETS sectors would likely be negligible as the total emissions of the EU ETS sectors are covered by a cap that declines every year. In fact, power sector emissions are reducing at a faster rate than that of any other sector. According to the projections based on the Reference Scenario 2016, about 65% of electricity generated in the EU in 2030 will be carbon free 150 .

Projections taking into account newly proposed policies 151 show a carbon free share of more than 70%, and overall project a decrease of GHG intensity in the power sector of around 40% between 2015 and 2030. With the continuation of the linear reduction factor in the ETS beyond 2030, further reductions of the greenhouse gas intensity of the power sector will be realised. The WTW emissions of electric vehicles in particular can therefore be expected to reduce over time.

As the WTW emissions are not a property of the vehicle alone, it would be hard if not impossible to establish metrics which are accurate, fair and cost-effective. In fact, conventional powertrains are sufficiently flexible to use different fuel types within certain specifications and therefore it is not possible to determine ex-ante for a given new vehicle on which fuels it will actually run or to which extent these would be low-carbon fuels. PHEV and BEV will run on any form of electricity, no matter how it is produced, with PHEV also capable of running on liquid fuels. Hence, uncertain ex-ante assumptions would have to be used to account for the potential use of low-carbon fuels in the metric expressing the CO2 emission performance.

Alternatively, some fuel producers propose to use an ex-post crediting approach for based on actual fuel use and the respective GHG emission factors. While it is theoretically possible to establish WTT factors for the many different fuels used in vehicles 152 , there are numerous practical barriers to overcome to actually agree on such figures, which also vary geographically as well as over time. Lessons can be learned from the discussions regarding the monitoring requirements for upstream emissions in the context of the Fuel Quality Directive 98/70/EC (FQD), where stakeholder concerns about large administrative burden contributed to the political decision not to insist on detailed monitoring of emissions from well to tank and instead to discontinue regulating CO2 in the FQD after 2020. Similarly, as in the implementation of the Renewables Directive the issues of indirect land use change (ILUC) and the sustainability of imported low-emission fuels would have to be addressed. For the WTW based CO2 targets, the exact same issues would have to be faced, but in addition a discussion would be needed regarding electricity.

Even in the case of an ex-post crediting system, highly uncertain ex-ante assumptions would have to be made about the availability of such credits when setting new CO2 targets for cars in order to maintain a sufficient level of incentives for accelerating the adoption of efficient and clean technologies in cars and vans.

As the actual emission reduction potential, the market availability and the penetration rate of low-emission fuels falls outside the direct control of the automotive industry, ACEA advocates maintaining the current tank-to-wheel metric.

Additional information on WTW emissions can be found in Annex 8.1.

6.3.1.2Option TM_EMB: metric for setting the targets based on embedded emissions

Apart from the WTW emissions, which cover the use phase of the vehicle and the production of the fuels used, there are also "embedded" CO2 emissions associated with vehicle manufacturing (including the mining, processing and manufacturing of materials and components), maintenance and disposal.

It is estimated that those embedded emissions currently cause around 16% of the total lifetime CO2 emissions of EU cars 153 . Additional information on embedded emissions can be found in Annex 8.1.

The evaluation study concluded that the further uptake of technologies improving the fuel efficiency of conventional (internal combustion engine) vehicles would have a limited impact on production emissions, and that the tailpipe CO2 emission savings achieved through such measures would outweigh by far any additional production emissions.

Nevertheless, it was also noted that the relative importance of embedded emissions may increase in the long-term, in particular when the proportion of vehicles using alternative powertrains is increasing.

A number of recent studies highlighted the potential emissions associated with the production of batteries for electric vehicles. However, the emission factors calculated vary significantly depending on the type of battery in terms of materials and energy density and the source of energy used for its production 154 . Furthermore, it is anticipated that the significance of batteries in the overall carbon footprint of electric vehicles could decrease very significantly due a number of factors, including the anticipated increase in gravimetric energy density reducing the materials use per kWh, the reduced GHG intensity of the power sector (see above) and materials used in battery manufacture, improved recycling processes, and an extension of the battery lifetime. Improved overall vehicle efficiencies would also contribute to this by reducing the size of the battery needed for a given electric range. All of this would cause the GHG emissions from the lifecycle of a BEV to drop by 40% between 2020 and 2030, in particular, if combined with establishing a strong battery manufacturing base within the EU in the near future.

Another study 155 highlighted the technical complexity of the issue, and the high administrative burden of covering embedded emissions in a meaningful way. In addition, trade policy issues might be raised as in the case of the emissions from fuels produced from Canadian tar sands during the implementation of the FQD. Such highly complex and detailed emission reporting would need to rely on life-cycle assessment (LCA) reporting by manufacturers which would have to cover all relevant downstream emissions from a huge number of suppliers of materials and car parts within the EU and from third countries. Developing a meaningful and robust methodology with guidelines and tools would be lengthy and costly.

Using a pre-described LCA approach that is sufficiently meaningful and providing the right incentives for reducing the embedded emissions would not only be extremely complex in terms of methodological approach, but would also be very difficult to enforce.

If such a LCA methodology could not be established, fixed default values for including embedded emissions in the metric would have to be used. However, this would have very limited added value as it would just give incentives for reducing the amount of materials used, but not take account of the differences between the emissions related to various materials.

In response to the public consultation, most car manufacturers were against covering embedded emissions in the metric, while other stakeholder groups had diverging views. The steel industry mentioned that the eco-innovation scheme should be complemented with an LCA credit option.

For the reasons above, including embedded emissions in the metric in a meaningful way is not deemed feasible with an effort proportionate to the expected benefits due to the technical complexity of the issue and the prohibitively high cost of data collection at the level of granularity required.

In the coming years, voluntary reporting on embedded emissions of the most relevant segments along the supply chain and testing various methodological approaches could offer further insights to manufacturers on the overall carbon footprint of car manufacturing. This could be combined with regularly monitoring the progress made with reducing the embedded emissions through dedicated studies.

6.3.1.3Option TM_MIL: metric for setting the targets based on mileage weighting

Information on vehicle lifetime mileage was gathered in the context of two studies on behalf of the Commission. A first one 156 investigated differences in lifetime mileage between vehicle categories. A follow-up study 157 gathered additional data and analysed the total mileages of vehicles of different ages with the aim of describing how annual mileage varies and accumulates during the vehicle lifetime.

It was found that diesel cars on average run higher mileages than petrol cars, but no size-related differences in mileage were identified for vans.

Introducing mileage-weighting when calculating the fleet-wide average emissions, by weighting the CO2 emissions of each type of car by the distance typically travelled over its lifetime, would impose a proportionately more stringent target on larger and heavier vehicles. According to the findings of the study, this would in turn slightly reduce by 1.6-1.8% the overall fleet-wide cost of achieving the same CO2 reduction.

A main challenge encountered during these studies was to find appropriately detailed data at Member State level and important data gaps remain in this respect.

A more recent study 158 , building on the aforementioned data, concluded that accounting for different lifetime mileages would have a relatively limited impact on the effectiveness, costs and competitiveness. Furthermore, it was highlighted that establishing robust and broadly agreeable mileage numbers for different vehicle types and categories depending on the utility value or other characteristics would be very complicated.

In light of the above, there are a number of uncertainties around the feasibility of establishing a robust mileage database to implement this option.

6.3.2Target levels for cars (TLC) and vans (TLV)

6.3.2.1Introduction

As regards the CO2 emission performance of new passenger cars, due to the continuous overall improvement of car technologies some autonomous improvement is expected to occur under the baseline. On average, WLTP CO2 emissions in 2030 are estimated to be 14% lower than in 2021. For vans, a similar effect is seen, bringing down emission by 10% in 2030 compared to 2020. In fact, the improvements already captured in the baselines TLC0 and TLV0 are very similar to the results of the options TLC10 and TLV10. Therefore, there is no need to consider the latter options further.

Table 6 and Figure 10 show the impact of the remaining six target options on the composition of the EU-wide fleet for passenger cars in 2025 and 2030.

At moderate target levels up to TLC30, the change in composition of the fleet will be rather gradual compared to the baseline. For instance, with a 30 % target the share of gasoline and diesel cars in 2030 will still make up almost three quarters of the total fleet, compared to slightly more than 80% in the baseline. Only at the higher target levels, the change would be more rapid. In the most ambitious scenario, the gasoline and diesel car share would decline to a little more than 55%.

It should be noted that for option TLC20 the new fleet composition results in an over-achievement of the CO2 target constraint. This is because for all policy options, the introduction of the CO2 target constraint is assessed in the context of the broader policy on low-emission mobility, i.e. in conjunction with enhanced availability of recharging infrastructure and better user acceptance of advanced powertrains as higher mileage of EVs reduces range anxiety. These factors result in an enhanced up take of more advanced power trains. In combination with cost-beneficial improvements of ICEVs, this leads to a situation that the 20% target is somewhat overachieved. This effect is also illustrated in the results regarding final energy demand (section 6.3.2.2.1.4) and CO2 emission trends over time (section 6.3.2.4.1), where the TLC20 results are somewhat optimistic, when comparing the different policy options.

Table 6: Passenger car fleet powertrain composition (new cars) in 2025 and 2030 under different TLC options

Figure 10: Passenger car fleet powertrain composition (new cars) in 2025 and 2030 under different TLC options

Table 7 and Figure 11 show the impact of different target level options on the composition of the EU-wide fleet of new vans in 2025 and 2030. This shows that for vans the change would be a little less pronounced than for cars. Under the 30% target, in 2030 almost four fifths of the vans would still be equipped with a more efficient combustion engine. At the highest level of ambition considered, this share would fall to a little less than 55%.

Table 7: Van fleet powertrain composition (new vans) in 2025 and 2030 under different TLV options

Figure 11: Van fleet powertrain composition (new vans) in 2025 and 2030 under different TLV options

6.3.2.2Economic impacts (including employment)

In this section the following impacts are considered:

(I)Net economic savings from different perspectives

(II)Energy system impacts

(III)Macro-economic impacts, including employment

Net economic savings taking different perspectives

The direct economic impacts of the abovementioned options have been assessed by considering the changes (compared to the baseline) in capital costs, fuel costs 159 , and operating and maintenance (O&M) costs for an "average" new vehicle (car or van), registered in 2025 or in 2030.

An "average" new vehicle of a given year is defined by averaging the contributions of the different segments of small, medium, large vehicles and powertrains by weighting them according to their market penetration as estimated. The PRIMES-TREMOVE model projects the new fleet composition in a given year as a result of the need to comply with the requirements of the new policy. Therefore, the different policy options lead to different projected fleet compositions, characterised by different shares of powertrain types (diesel, gasoline, battery electric, plug-in hybrids, etc.) in the different market segments. The net savings for an "average" vehicle are calculated by averaging the costs and savings of the different powertrain types and segments, using the projected shares as weights. Since these shares change among the different scenarios, and they change for the new vehicles of 2025 and those of 2030, the cost indicators are used to represent the economic impacts for the new fleet of 2025 and 2030.

For this analysis, the following indicators have been used:

·Net economic savings over the vehicle lifetime from a societal perspective

This parameter reflects the change in costs over the lifetime of 15 years of an "average" new vehicle without considering taxes and using a discount rate of 4%.

·Net savings from an end-user perspective, using two different indicators:

oTotal Cost of Ownership (TCO) over the vehicle lifetime (TCO-15 years)

This parameter reflects the change in costs over the lifetime of 15 years of an "average" new vehicle. In this case, given the end-user perspective, taxes are included and a discount rate of 11% for cars or 9.5% for vans 160 is used.

oTCO for the first user, i.e. net savings during the first five years after registration (TCO-first user):

This parameter reflects the change in costs, during the first five years of use, i.e. the average time the first buyer is using the vehicle. Again, taxes are included and a discount rate of 11% for cars or 9.5% for vans is used. The calculation also takes account of the residual value of the vehicle and the technology added with depreciation.

Sensitivities

As explained in Section 6.1, apart from the cost curves based on the "Medium" technology cost estimates, a number of other cost-curves were developed as part of a sensitivity analysis. While the overall economic analysis of the policy options (TLC and TLV) relies on the use of the Medium costs, some sensitivities were run to investigate the effect on the net costs (savings) in case technology costs would decrease faster than anticipated under the Medium cost case. This additional assessment also allows looking into a situation where costs evolve differently for different powertrain types. This is particularly relevant for EV in view of the importance of the battery cell costs and the higher uncertainty over how these costs will evolve in the future very much depending on market penetration.

Two other sensitivities explored are related to the future oil price and to the evolution of the share of diesel cars in the fleet.

Energy system impacts

In view of the close link between the LDV CO2 standards and energy use in the transport and fuel sectors, the energy system impacts have been analysed, considering the final energy demand, the final energy demand by energy source and the impact on the electricity system.

Macro-economic impacts

The broader macro-economic impacts of the different TL options have been analysed for the LDV sector (passenger cars and vans) as a whole. Therefore, the results are presented for cars and vans together in Section 6.3.2.2.3.

While the below Sections provide an overview of the main findings of the assessment and some illustrative tables and figures, the detailed results of the calculations of the net savings and their components are given in Annex 8.

Passenger cars (TLC)

6.3.2.2.1.1Net economic savings over the vehicle lifetime from a societal perspective

Table 8 and Figure 12 show the net savings (EUR per vehicle, expressed as the difference with the baseline) over the vehicle lifetime from a societal perspective for an average new passenger car registered in 2025 and in 2030 under the different TLC options. The net savings observed are the result of differences in capital costs, fuel cost savings and O&M costs.

Capital costs – which in this case are equal to manufacturing costs - increase with stricter fleet-wide CO2 target levels as reducing CO2 emissions will require additional more expensive technologies to be implemented. For a car registered in 2025, the average additional capital cost ranges from 115 EUR (TLC20) to 1,411 EUR (TLC_EP40). In 2030, it ranges from 419 EUR (TLC20) to 2,752 EUR (TLC_EP50) per car.

At the same time, stricter targets will lower fuel costs as the fuel efficiency of the cars improves and more alternative powertrains are deployed, both measures reducing the amount of fuel consumed. Fuel cost savings per car range from 354 EUR (TLC20) to 1,394 EUR (TLC_EP40) in 2025 and from 1,159 EUR (TLC20) to 2,558 EUR (TLC_EP50) in 2030.

O&M costs show little variation between the different options, as they depend on the insurance and maintenance costs for the different segments and powertrains which compose the PRIMES-TREMOVE optimised fleet.

Both in 2025 and in 2030, net savings occur for options TLC20, TLC25, TLC30 and TLC40, ranging from 78 EUR (TLC40) to 152 EUR (TLC30) per car in 2025 and from 565 EUR (TLC40) to 902 EUR (TLC25) per car in 2030. Option TLC_EP40 results in net savings in 2030 (512 EUR per car), but not in 2025 (net costs of 42 EUR per car) while under option TLC_EP50 net savings are just below zero in both 2025 and 2030.

As can be seen from Table 8 and Figure 12 , the highest net savings can be realised with options TLC25 and TLC30 in both 2025 and 2030.

Table 8: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/car)

Figure 12: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/car)

In principle, in order to estimate the net economic savings over the vehicle lifetime from a societal perspective, one should include also the external benefits (or avoided external costs). For the options assessed here, the most important effect concerns additional benefits in terms of avoided CO2 costs over the lifetime of a vehicle as compared to a baseline vehicle.

Table 10 gives an overview of the estimated additional avoided CO2 costs for cars in 2030 for the different options assessed 161 . It shows that these external benefits increase as the CO2 target gets stricter.

Table 10: Avoided CO2 cost (EUR/car) over a car's lifetime

6.3.2.2.1.2TCO-15 years (vehicle lifetime)

Figure 13 shows the TCO over 15 years (EUR per car) of an average new passenger car registered in 2025 and in 2030 under the different TLC options (expressed as the difference with the baseline).

It shows that in both years and under all options considered there are net savings for the end-users over 15 years. The savings per car in 2025 range from 253 EUR (TLC_EP40) to 436 EUR (TLC30) and they increase in 2030, ranging from 389 EUR (TLC_EP50) to 1,374 EUR (TLC25).

The highest net savings for the total cost of ownership over 15 years can be realised with a CO2 target as in options TLC25 or TLC30.

Figure 13: TCO-15 years (vehicle lifetime) (net savings in EUR/car in 2025 and 2030)

6.3.2.2.1.3TCO-first user (5 years)

Figure 14 shows the net savings (EUR per car) from a first end-user perspective for an average new passenger car registered in 2025 and in 2030 under the different TLC options (expressed as the difference with the baseline).

The trends seen are very similar to those found for the analysis from a societal perspective (see above).

Capital costs increase as the fleet-wide CO2 target levels get stricter and range from 90 EUR (TLC20) to 1,104 EUR (TLC_EP40) for the average car registered in 2025. In 2030, they range from 328 EUR (TLC20) to 2,154 EUR (TLC_EP50) per car.

At the same time, stricter targets will lower fuel costs and fuel cost savings per car range from 348 EUR (TLC20) to 1,286 EUR (TLC_EP40) in 2025 and from 1,025 EUR (TLC20) to 2,354 EUR (TLC_EP50) in 2030.

O&M costs show little variation between the different options and are generally positive in 2025 and negative (i.e. lower than under the baseline) in 2030.

For the first user, both in 2025 and in 2030, net savings occur under all options considered, ranging from 171 EUR (TLC_EP40) to 263 EUR (TLC30) per car in 2025 and from 282 EUR (TLC_EP50) to 818 EUR (TLC30) per car in 2030.

The results of the following two sensitivities are given in Annex 8.2:

(I)sensitivity regarding the effect of varying cost assumptions;

(II)sensitivity regarding the effect of a varying international oil price.

Figure 14: TCO-first user (5 years) (net savings in EUR/car in 2025 and 2030)

6.3.2.2.1.4Energy system impacts 

Figure 15 shows the impact of the different TLC options on the final energy demand for passenger cars over the period 2020-2040.

Under the baseline (TLC 0), the final energy demand for passenger cars is 170,300 ktoe in 2020 and it decreases over time as cars being subject to the CO2 targets set in the current Cars Regulation enter the fleet. In 2030, the final energy demand for passenger cars is 13% lower than in 2020, and the effect of the current targets continues afterwards, i.e. in 2040 final energy demand is 16% lower than in 2020).

Under the different policy options regarding the CO2 target level, the final energy demand for cars reduces further, and the effects of more stringent CO2 targets become more outspoken from 2030 on as more and more cars which are subject to those targets enter the fleet.

The EU-wide fleet targets for CO2 also affect the composition of the car fleet in terms of the powertrains used and hence have an impact on the demand per type of energy source in the transport sector.

Figure 16 shows the share of different fuel types used in the entire passenger car fleet (i.e. not only the newly registered cars) in 2025 and 2030. Diesel and gasoline by far remain the main fuels used in 2025 and 2030. Even for the most ambitious target level, there is only a gradual shift away from fossil to alternative fuels, in particular electricity and hydrogen. The shift is more outspoken in 2030 and for the options with more stringent CO2 targets. For the other fuel types (biogasoline, biodiesel, gaseous fuel), there are very limited changes amongst the different options considered.

Figure 15: Final energy demand (ktoe) for passenger cars over the period 2020-2040 under different TLC options

Figure 16: Share (%) of different fuel types in the final energy demand for cars (entire fleet) under different TLC options - 2025 and 2030

Electricity consumption

Table 9 shows the share of the total EU-28 electricity consumption used by cars and vans (together) in 2025 and 2030 for different CO2 target level options. It illustrates that, even with the strictest targets considered, the share of electricity used by light-duty vehicles up to 2030 is not more than a few percent.

Table 9: Electricity consumption by cars and vans with respect to total electricity consumption (EU-28) under different options for the EU-wide CO2 target levels

Diesel and gasoline demand

Table 10 shows the cumulative savings of diesel and gasoline in the period 2020 to 2040 with respect to the baseline for different scenarios. Considering the combination of options TLC30 and TLV40, the cumulative savings between 2020 and 2040 are equivalent to around 150 billion euros at current oil prices

Table 10: Cumulative diesel and gasoline savings (Mboe) over the period 2020 to 2040 under different policy options with respect to the baseline

Sensitivity – effect of decreasing share of diesel cars

In view of recent developments following the diesel emission scandal and the persistent air quality issues in a number of cities across the EU, the share of diesel cars in the fleet of newly sold cars has declined in a number of EU Member States 162 .In order to assess the potential effects, two sensitivities were designed with lower diesel car fleet shares, as shown in Table 11 .

Table 11: Share of diesel cars (incl. diesel hybrids) in the new car fleet under the two "Low Diesel" sensitivities - % reduction compared to the baseline

The resulting fleet composition under these two sensitivities is shown in Table 12 , compared with the fleet composition in case of TLC25 and TLC30. It makes clear that diesel cars are largely substituted by gasoline cars with rather limited increases in PHEV, BEV and other (gaseous fuel) cars.

Table 12: Passenger car fleet composition in 2025 and 2030 under the "Low Diesel" sensitivities compared to options TLC25 and TLC30

Table 13 shows the resulting tailpipe CO2 emission reductions for cars between 2025 and 2040, taking 2005 as the reference year, under the "Low Diesel" scenarios in conjunction with the EU-wide fleet CO2 target of option TLC30. It shows that the impact of the declining diesel share is limited. CO2 is reduced only slightly less than under option TLC30 when using the initial fleet composition. This is due to the modelled gap between test cycle and real-world emissions, which is slightly lower for diesel cars compared to gasoline cars. Therefore, a declining share of diesel cars leads to a small overall increase in the gap between type approval and real world emissions, hence a lower emissions reduction.

Table 13: (Tailpipe) CO2 emissions of passenger cars in EU-28 - % reduction compared to 2005

In terms of economic impacts, the three tables below show that net savings from a societal perspective, vehicle lifetime perspective and first-user perspective decrease as diesel shares are declining. This is mainly due to a decrease in the fuel savings when the market shares of diesel car decrease. However, from any of the three perspectives there will still be significant net savings, especially when approaching 2030.

Table 14: Net economic savings from a societal perspective (EUR/car)

Table 15: TCO– lifetime (15 years) – net savings in EUR/car

Table 16: TCO- first user (5 years) – net savings in EUR/car

Light commercial vehicles (TLV)

6.3.2.2.1.5Net economic savings over the vehicle lifetime from a societal perspective

Table 17 and Figure 17  show the net savings over the vehicle lifetime from a societal perspective for an average new van registered in 2025 and in 2030 under the different TLV options (expressed as the difference with the baseline).

Capital costs – which in this case equal manufacturing costs - increase with stricter EU-wide fleet CO2 target levels as reducing CO2 emissions will require additional more expensive technologies to be implemented. For a new van registered in 2025, the average additional capital cost ranges from 232 EUR (TLV20) to 1,469 EUR (TLV_EP50). In 2030 (when stricter targets apply), it ranges from 426 EUR (TLV20) to 2,439 EUR (TLV_EP50) per van.

At the same time, stricter targets will lower fuel costs and fuel cost savings per van range from 1,002 EUR (TLV20) to 2,529 EUR (TLV_EP40) in 2025 and from 2,063 EUR (TLV20) to 4,261 EUR (TLV_EP50) in 2030.

O&M costs show little variation between the different options, apart from TLV_EP50, where these costs are significantly lowered in 2030.

Both in 2025 and in 2030, net savings occur under all options, ranging from 810 (TLV20) to 1,369 EUR (TLV_EP40) per van in 2025 and from 1,687 EUR (TLV20) to 2,386 EUR (TLV40) per van in 2030.

Overall, net savings for vans are significantly higher than for cars under options with similar emission target reduction percentages due to the much higher fuel cost savings achieved as vans start reducing from a significantly higher CO2 efficiency standard. Importantly, the highest benefits occur at target levels of 30% and 40% in 2030. This could help improving the competitiveness of many SMEs.

Table 17: Net economic savings over the vehicle lifetime from a societal perspective in 2025 and 2030 (EUR/van)