Joint Research Centre - European Commission · Joint Research Centre ... Résultats des contrôles des émissions de polluants atmosphériques et de CO2 menés sur les 52 premiers

Joint Research Centre the European Commission's in-house science service

Serving society

Stimulating innovation

Supporting legislation

Bridging the CO2 Gap Options for policy action

and way forward

Georgios Fontaras

Lisbon – June 08th, 2016

Meeting with the SAM-HLG

Disclaimer: The views expressed are purely those of the presenter and may not in any circumstance be regarded as stating an official position of the European Commission

• Key figures & Present situation

• Building a trustworthy framework

• Current Tools: WLTP, RDE, Labeling

• US a case study

• JRC’s work on “An integrated framework”

• Summary

Outline

Key figures & Current Situation

3

• Road Transport:17.5% of EU CO2

• 800 M tonnes

• 70% of which from pass cars

• On the road today about:

• 208 M pass. cars

• 1.3 M buses/coaches

• 28.6 M Trucks

Key figures – CO2 and road transport

Key figures – Market and Industry

• About 14M new passenger cars/year sold in Europe

• Slow growth rate but increasing mobility needs

Difficult to stabilize and curb CO2

• Biggest (75%) consumer of petroleum products

• Automotive sector:

• 4% of EU GDP (~sum of the lowest 12 GDPs of EU MSs)

• 5.6% of employment

• 107€ Billion worth of exports

• About 380 € Billion in tax revenue

• Target 60% GHG reduction by 2050 (Base 1990)

• 40% by 2030 (Base 1990)

• 30% reduction for non-ETS sector (Base 2005)

• Extensive Series of Regulations, Directives and Technical

documents regulating emissions and CO2 savings measures (A

Really Long List could appear here!)

• More info available:

http://ec.europa.eu/clima/policies/transport/vehicles/cars/documentation_en.h

tm

Key figures – EU policy context

• Mass dependent emission limits for average sales weighted emissions for

each OEM

• Target in 2015, CO2 shall not average more than 130 gCO2/km (roughly

~5.2l/100km) – the target is set to 95 gCO2/km for year 2020

Target

achieved

Policy: Present situation

• Progresses appear to be in line with the requirements of 2020

targets

Source EEA

Present situation

“Emissions from official testing reported by national authorities show that new cars sold in the European Union (EU) are increasingly more fuel-efficient. Last year, new passenger cars emitted on average 119.6 grammes (g) of carbon dioxide (CO2) per kilometre, 8% below the official EU target set for 2015, according to provisional data published today (14/04/2016) by the European Environment Agency (EEA).”

Source: ICCT

...systematically different and (lately) increasingly higher But reality can be different…

But reality can tell a different story

Building a trustworthy framework

10

Vehicle type-approval system in Europe

System founded on mutual trust!

• Type approval is the European approach (other markets follow self

certification eg US)

• For entering the market an OEM has to “demonstrate” that a vehicle fulfils

the relevant EU standards (eg safety and emissions)

• Done by performing a number of tests on a vehicle (representative of a

certain family of vehicles from a vehicle manufacturer) under the

supervision of a third party (type-approval authority or technical service)

who certify the quality of the process (Directive EC 2007/46)

• Type I test concerns the following:

• Fuel/Energy consumption

• CO2 emissions

• Emissions of gaseous compunds (NOx, CO, PM/PN)

Source: ICCT 2014

Introduction of binding targets Each g of CO2 in excess can potentially cost up to 95E per vehicle sold

Binding targets & strict penalties alone, not an effective solution – need for multidimensional initiatives

Reducing the GAP: The Way forward Step 0

• The CO2 monitoring / certification / consumer information system has to be

trustworthy first and then accurate

• Most discussion revolves around accuracy or precision

• Trust is what should be built first

• Processes and results have to be:

• Comprehensible

• Traceable

• Reproducible

• Quality controlled

• The same across EU

• Validated

• Validation, feedback collection, regular methodology revision, quality control and

harmonization should be founding blocks of any framework to be proposed

The man who moves a mountain begins by carrying away small

stones. Confucius

Small steps can make a big difference

Example A: Different TA authorities and technical services may sometimes interpret existing legislation differently. Better harmonization across Europe could be a big step

Small steps can make a big difference

Example B: How an ex – post validation of the certification values could raise consumer awareness, ring bells (and trigger further actions if necessary) Wouldn’t you as consumer would like to know what are models A and B?

http://www.developpement-durable.gouv.fr/IMG/pdf/2016-04-07_-_DP_Resultats_Commission_UTAC-2.pdf

Résultats des contrôles des émissions de polluants

atmosphériques et de CO2 menés sur les 52 premiers véhicules

B) Unfair competitive advantage A) Unfair competitive disadvantage

Available Policy Instruments

17

Available policy instruments

RDE WLTP Label

• Adding completely new tests procedures at certification level may not be the most

effective option

• Focus on available tools / instruments, improve, strengthen and enhance

• Technology offers multitude of additional tools for quality and reality check!

• Ex – Post verification scheme of great importance!

WLTP - strengths

• The most important initiative in improving vehicle CO2 certification and

monitoring to date!

• A very important step ahead addresses several of the know shortcomings

• New protocol taking into account the evolution in vehicle complexity

• More realistic driving cycle

• Stricter & more realistic definitions of boundary conditions

• Less margins for exploitation (at least based on current understanding)

• World standard

• High repeatability & reproducibility

WLTP challenges (1/2)

• Becoming an extremely complex procedure. Don’t want a scheme understandable and

controllable only by professional WLTP connoisseurs

Keep it as simple as possible, don’t introduce further complexity if not really necessary

• Transparency & reproducibility issues: selection of families and boundaries, limited public

information on vehicles (what road loads?, what are WLTP-H & L vehicles, how can a third

party reproduce/check results, software optimization for hybrids?)

Keep it open and transparent, make more info data public, give the opportunity of test

reproduction to independent parties. Provide real time updated & consolidated versions of

WLTP, explanatory material on the structure of the procedure descriptions etc

• No regular updates foreseen (any protocol should be revisited once every 10 years and not

once every 40 years as with NEDC)

Foresee regular assessment and revisiting of the WLTP, every 10 years it should be

“recalibrated”

WLTP challenges (2/2)

• A lab cycle can never cover all conditions, all drivers, circumstances. Specific

mix of traffic conditions with pronounced motorway driving maybe not so

relevant for specific car categories (eg. Very small city cars)

ok for certification but for closing the gap & targeted policy interventions additional

possibilities should be used (eg ex-post verifications, market surveillance, simulation tools

user feedback)

• Strict but still not fully realistic boundary conditions (eg. No slopes, limited

temperature range, no A/C use yet, relatively low RPM)

ok for certification purposes, but for closing the gap and making targeted policy

interventions additional supporting tools should be foreseen (ex-post verifications,

market surveillance, simulation tools, user feedback)

RDE for CO2 – strengths

• RDE on its own is an important step ahead

• The most important initiative together with Euro standards in tackling

regulated pollutant emissions

• Potentially valuable tool if appropriately expanded to CO2

• Basis for “Not-to-Exceed” factors or screening mechanism for

triggering additional lab validations

• RDE tests of up to 2h generate a large data pool

• CO2 data recorded from actual on-road driving can be used in various

ways (eg. Calibrate/validate vehicle & compare with WLTP)

• Potential coverage of a wide range of conditions

• Higher Robustness against defeat strategies

RDE CO2 data – are not The silver bullet

• Lower repeatability & reproducibility of CO2 measurements

Inherent limitation prevents CO2 certification to be based mainly on RDE

Possible use in combination with other instruments (WLTP, simulations, data mining, label ,

CO2 limits etc)

• Absence of data collection and assessment mechanism for RDE CO2 (yet) – where do the results go and

to whom?

Need to establish the necessary framework

• Potentially biased trip design deviating from average driving

Trip design and definition should probably be revisited for CO2 purposes

• Being a regulated procedure , used also for other purposes ,it will always have specific limits & margins.

Test optimizations may increase the gap and result in lab-test like situations.

Definitions should be strict.

Good for not to exceed type of limits but not so good for CO2 correction factors development.

• No regular quality check of method and feedback-based corrections foreseen (how do we know it actually

works as should?)

Feedback collection and analysis mechanism necessary

As for WLTP, periodic revisit of the method necessary

RDE CO2 data – are not The silver bullet

• Lower repeatability & reproducibility of CO2 measurements

Inherent limitation prevents CO2 certification to be based mainly on RDE

Possible use in combination with other instruments (WLTP, simulations, data mining, label ,

CO2 limits etc)

• Absence of data collection and assessment mechanism for RDE CO2 (yet) – where do the results go and

to whom?

Need to establish the necessary framework

• Potentially biased trip design deviating from average driving

Trip design and definition should probably be revisited for CO2 purposes

• Being a regulated procedure , used also for other purposes ,it will always have specific limits & margins.

Test optimizations may increase the gap and result in lab-test like situations.

Definitions should be strict.

Good for not to exceed type of limits but not so good for CO2 correction factors development.

• No regular quality check of method and feedback-based corrections foreseen (how do we know it actually

works as should?)

Feedback collection and analysis mechanism necessary

As for WLTP, periodic revisit of the method necessary

RDE cannot replace the

WLTP for CO2 certification but it can become an extremely valuable source of data

WLTP & RDE: Why on their own are not enough

• Assuming the bull’s eye equals to a non-existent “single” real CO2 value

• Being official tests both have strict, known, boundary conditions

• But in reality there is no single value:

WLTP RDE × ×

× ×

Label / Information communication - strengths

• Consumer information and public awareness a key instrument

• Probably not exploited to its full potential to date

• Main communication channel to the customer

• Benchmark for apparently similar car models

• The only official source of information to your car

• Important market driver

Label / information communication - issues

• Presently relies solely to the official test result

Possibility to enhance / make more realistic based on real use fuel

consumption data collection, ex-post verifications, RDE, market surveillance

• Founded on the certification value (and any shortcomings linked to it)

Link to additional safeguards outside CO2 certification (ex-post verifications,

market surveillance, real world CO2 monit., user data, statistical corrections)

• Damaged reputation due to the CO2 GAP and its increasing trend

Improve, promote, provide info of better quality (maybe also quantity)

• Single way communication (no possibility for users to send feedback)

Foresee feedback channels eg free fuel consumption reporting, consumption

data collection during technical inspection

• Too static for 2016 (piece of paper at a dealer or a number in a webpage, in a world of

apps, mobile devices and continuous communication)

It can be made a more user friendly, more fun to use instrument

And an approach not presently available (but

technically feasible)

• Real world fuel consumption data collection is technically possible

• Real time fuel consumption should be available to the driver eg via

OBD port

• Allow for the development of codes and applications that can spur

better fuel economy

• Average life time information can be recorded by vehicles and

communicated anonymized during the compulsory technical

inspections

• Data to be collected and statistically treated to

• Monitor the real gap

• Provide feedback to drivers

• Support other tools and processes (inventories, decision making, fleet

models etc)

European CO2

Emissions

WLTP Values & Vehicle

Characteristics

RDE & Ex Post tests

& Real World Tests

Market Surveillance, Real world

Activity Data, & User

Feedback

Integrated Framework

Data Source

Data Check

User information

US’ approach (case study)

30

400g/mi = 248.5g/km

• Laboratory tests for Automaker compliance with standards

• Adjusted values for consumer information and CO2 monitoring

• http://www3.epa.gov/fueleconomy/fetrends/1975-2015/420r15016.pdf

US approach

“individual mileage will always vary and no label value can accurately predict fuel economy for all drivers under all conditions, the EPA fuel economy values are the best estimates for typical U.S. drivers and average driving conditions”

US approach: Emphasis on Info and the User

US approach: Sales weighted unadjusted vs adjusted

Unadjusted ~31.3 mpg Adjusted ~25.1 mpg

Potential gap: 1.83l/100km or 20% (9.37l - 7.54) l/100 US: About 42g CO2 /km EU: About 37g CO2/km (assuming average of 125g CO2/km and 30% gap)

JRC’s approach: Bring together data of different tests and

conditions and provide customized, Vehicle – Region Specific fuel consumption estimates

34

Policy Reality

JRC’s approach to the problem

Policy Reality

Regulatory complexity

Costs

Diversity of conditions

Increasing Vehicle complexity

Flexible, effective procedures

Simple, meaningful & accurate metrics

Monitor & Achieve targets

Lab and real world tests & Simulation

JRC’s approach to the problem

CO2MPAS tool

• A vehicle micro – simulator for the introduction of WLTP

• Developed by the JRC

• Model calibration is based on WLTP provided enough vehicle data.

Potentially RDE data can be used for WLTP back calculation

• The tool is currently in test phase of beta release

• JRC combined CO2MPAS with fleet data for global CO2 calculation

(average CO2 over WLTP)

Green-driving tool

Green-driving tool

• On the basis of the main vehicle (shape, powertrain, fuel, etc.) and

usage characteristics (number of passengers, aerodynamic

features, electric devices, initial temperature etc.) a user can

calculate fuel consumption on a selected trip

• The fuel consumption calculation is based on CO2MPAS and the

WLTP. Potentially RDE data can be used.

• The tool is currently under development. First BETA version

presented during JRC’s Open Day (28/May ‘16)

• In the future, the tool could be used on any specific vehicle if

needed to address a big part of the “gap”

Project: REAL CO2

• Possibility to obtain fuel consumption directly from vehicle

• Investigating:

• Possibilities for storing data at vehicle level

• Possibilities for communicating real fuel consumption data during

vehicle technical inspection

• Similar data already available data

• Companies big databases of real-life fuel consumption

• Making use of such kind of data could be a first option to

supplement official values or perform market surveillance

• Confidentiality seems a possible barrier

• A preliminary investigation is being carried out at the JRC

Summary

Summary 1/2

• There is no silver bullet in CO2 . There is no test that can capture all

possible operating conditions. By definition CO2 is not one value but a

distribution of values: who drives what, where and under which conditions.

• Need to re-build and maintain confidence in the system – need for ex-post

verifications & some form of surveillance mechanism

• WLTP is a very important step forward and has to be supported. Challenges

should be carefully considered and the procedure should be strengthened.

• RDE test or its elements could be further studied as a possible

complementary procedure to WLTP. RDE based certification not suitable for

CO2 emissions. RDE can be a valuable source of data

• Labeling and consumer information channels still have great potential

• Technology should be harnessed to improve our understanding of the

problem and provide information. Real activity data can be recorded,

collected and analyzed. Drivers must be involved to the extent possible.

Summary 2/2

• Real savings are achieved in real-life. What can the EC do?

Exploit existing tools to the best extent

Help “upgrade” fuel consumption to one of the top purchase criteria market

driven reductions in CO2

Stress the environmental importance of low CO2 emitting vehicles

Open to the public: drivers are not dumb. They should have access to more info

regarding their cars (through official channels)

Provide on-line tools & apps that can educate drivers and promote behaviors

and CO2 saving technologies

Draft best practices guidelines and information packages for the drivers

Educate the public make data sources available!

JRC Science Hub: www.ec.europa.eu/jrc

Twitter: @EU_ScienceHub

LinkedIn: european-commission-joint-research-centre

YouTube: JRC Audiovisuals

Vimeo: Science@EC

Stay in touch