Technical support to the correlation of CO 2 emissions measured under NEDC and WLTP Ref: CLIMA.C.2/FRA/2012/0006 Further details regarding the target translation 18 th December 2013 John Norris Project manager (Ricardo-AEA) Stefan Hausberger Task 1 & 2 leader (TUG) www.ricardo-aea.com
2 Contents 1. Project objective 2. Overview of target translation options 3. Option 1 advantages and disadvantages 4. Option 1 high level view on methodology 5. Option 2 advantages and disadvantages 6. Option 2 high level view on methodology 7. Conclusions and discussion
3 Project objective To develop a methodology for translating between the current NEDC-based fleet average CO 2 targets, and the associated utility function formulae, into targets of comparable stringency for CO 2 emissions measured using the WLTP Quantitative evaluation of the options from whole fleet and manufacturers and experts perspectives. For some options this will involve translating the current short and long term car and van targets and their formulae for into equivalent target levels and formulae for CO 2 emissions measured using the WLTP
4 Overview of the target translation options Option 1: Maintain NECD based targets Option 2: Introduce WLTP based targets Translating the fleet target Translating the specific emission target formulae Option 3: Mixed approach
Option 1: Maintain NEDC based targets - Advantages and disadvantages Advantages: For short/medium term, i.e. during the transition period, confirms the existing regulations on CO 2 targets. Legislation linked to the CO 2 targets at different levels remains valid. Examples: European reporting and monitoring requirements, national (e.g. UK vehicle tax bands) and local (e.g. CO 2 based charge exemptions) levels. Disadvantages: The WLTP will be introduced and the conversion from NEDC to WLTP will be required at some date in the future. Technical data monitored and reported will need to be expressed in comparable NEDC values. If NEDC targets remain after the introduction of WLTP then robust WLTP to NEDC translation methodology will be required. 5
6 Option 1: Maintain NEDC based targets - Possible methodology Outputs required: For data for which WLTP data is measured, i.e. new models for the first year from WLTP introduction date, outputs required are: equivalent NEDC CO 2 emissions, and vehicle reference mass (for manufacturers specific emissions targets) Methodology for providing outputs: Either the output data required will be included in the vehicle s Certificate of Conformity, and included in monitoring & reporting return to EEA; or meta-models will be used to translate WLTP CO 2 emissions to equivalent NEDC CO 2 emissions value.
Option 1: Maintain NEDC based targets - Possible methodology Calculations to be performed: For whole fleet average CO 2 value: The NEDC CO 2 emissions data (generated from the translated WLTP measurement) to be summed, weighted with the registration data, and averaged along with other NEDC CO 2 emissions data to give the whole fleet average For manufacturer specific emissions targets: The NEDC CO 2 emissions data (generated from the translated WLTP measurement) and the NEDC vehicle reference mass to be summed, weighted with the registration data as at present, to give the manufacturer average specific emissions. Key issues: Provision of acceptable accuracy with a reasonable administrative burden 7
8 Option 2: Introduce WLTP targets generally Advantages and disadvantages Advantages: At some time in the future all light duty vehicles will be type approved using the WLTP (drive cycle and test procedure). Translation of CO 2 emissions targets provides clarity regarding the future target. Disadvantages: All options involving the translation of NEDC CO 2 emissions targets into equivalent WLTP targets, based principally on the meta-models, will need to address the challenges of: Stakeholder acceptance of the translation methodology that the models provide acceptable accuracy, and Ensuring comparable stringency of targets.
9 Option 2: Introduce WLTP targets Sub-options Translating the NEDC fleet targets (Options 2(a)): Option 2(a)(i) - Direct translation of short term based on NEDC CO 2 emissions targets for one or more reference years; Option 2(a)(ii) - Translation of long term targets Option 2(a)(ii)(1) Using direct translation, as for Option 2(a)(i) Option 2(a)(ii)(2) Using direct translation, as for Option 2(a)(i), but including consideration of the evolution of the power trains (fuels) and technologies Translating the slope of the NEDC CO2 emission target curve, affecting manufacturer specific emissions targets (Option 2(b)): Option 2(b)(i,ii)(1) - Direct translation of the NEDC emission target formulae; Option 2(b)(ii)(2) - Adjustment of the NEDC emission target formulae so as to minimise change for manufacturers.
10 Option 2(a)(i) & Option 2(a)(ii)(1) - Direct translation of NEDC short and long term CO 2 emissions targets using and reference year s registrations profile Advantages: It is based on objective registrations data (Monitoring and reporting database for 2012) and meta-models for the CO 2 emissions translation Targets can be defined now, i.e. when the correlation exercise has been successfully completed, with no assumptions required regarding future technologies
11 Option 2(a)(i) & Option 2(a)(ii)(1) - Direct translation of NEDC targets Disadvantages and challenges: Takes no account of evolution of the power trains (fuels) and technologies that might be used to meet the NEDC based targets Needs to either a) be able to link each row of data in the CO 2 vehicle monitoring database with the appropriate meta-model to deduce WLTP CO 2 value from the declared NEDC CO 2 value, or b) aggregate rows of registrations data from within the CO 2 vehicle monitoring database to generate consolidated database, which is then translated. Will need to predict a WLTP CO 2 emissions value with little further data than that available in the CO 2 vehicle monitoring database. Will translate and transfer some NEDC optimisation across to the new WLTP targets
12 Option 2(a)(i) & Option 2(a)(ii)(1) - Direct translation of NEDC targets - Possible methodology Outputs required: Translated NEDC CO 2 emissions for vehicle models for which WLTP data is not measured (i.e. currently all vehicles sold). Methodology for providing outputs: Use either the meta-models or correlation functions, to translate NEDC CO 2 emissions value into equivalent WLTP value. Either: Line by line (of the monitoring and reporting database) translation, or By the aggregation of lines in the monitoring and reporting database and then translating these aggregates of data. Anticipated that the key function will be the ratio of the NEDC/WLTP CO 2 emissions, not differences, and that the ratio will be applied to the NEDC CO 2 value to generate the equivalent WLTP CO 2 value.
13 Option 2(a)(i) & Option 2(a)(ii)(1) - Direct translation of NEDC targets - Possible methodology CO 2 emissions for passenger cars Average CO 2 emissions for WLTP for reference year CO 2 (WLTP) Reference year CO 2 emissions (g/km) Average CO 2 emissions for NEDC for reference year CO 2 (NEDC) Reference year WLTP translated 95 g/km NEDC target 95 g/km NEDC target Translated target(wltp) = 95 x CO 2 (WLTP) Reference year CO 2 (NEDC) Reference year And equivalent equation for 130 g/km 2015 target
14 Option 2(a)(ii)(2) - Direct translation of NEDC CO 2 & include effect of technology evolution Advantages: Combines being based on objective registrations data and includes future technology evolution Option is based on objective registrations data (Monitoring and reporting database for 2012) and meta-models for the CO 2 emissions translation
15 Option 2(a)(ii)(2) - Direct translation of NEDC CO 2 & include effect of technology evolution Disadvantages and challenges: Starting point would be the direct translation of NEDC to WLTP CO 2 emissions for reference year using the meta-models therefore the challenges discussed for Option 2(a)(ii)(1) relating the appropriate metamodel to each row of data in the CO 2 vehicle monitoring database remain. Will require a projection quantifying how the power trains (fuels) and technologies that manufacturers are planning to use to meet the NEDC based targets will evolve Will also require meta-models to estimate impact of new technologies on NEDC & WLTP CO 2 emissions, possibly on very little physical data. Will translate some NEDC optimisation to be transferred across to the translated WLTP targets
16 Option 2(a)(ii)(2) - Direct translation of NEDC CO 2 & include effect of technology evolution Possible methodology Outputs required: Translated NEDC CO 2 emissions for vehicle models for which WLTP data is not measured (i.e. currently all vehicles sold), Agreed list of future technologies to be included, and estimate of effect of new technologies on NEDC & WLTP CO 2 emissions A fleet averaged Technology Factor multiplier to quantify the effect of technology evolution.
Option 2(a)(ii)(2) - Direct translation of NEDC CO 2 & include effect of technology evolution Possible methodology Methodology for providing outputs: Assume only relevant for longer term targets (not short term targets); Use directly translated target from Option 2(a)(ii)(1) as starting point; Obtain view from manufacturers and experts regarding what additional technology penetration and evolution will occur by 2020 relative to the baseline year, to what extent and to which vehicle categories it will be applied; Using meta-models/correlation functions establish effect of emerging technologies on both NEDC and WLTP CO 2 emissions For those emerging technologies where there is a difference between the NEDC and WLTP CO 2 emissions: 17 Segment reference year database into vehicle categories, and gauge to what extent the emerging vehicle technologies are already being used; Using the proportion of all registrations of that vehicle category as proportion of all registrations, and the difference between the NEDC and WLTP CO 2 emissions calculate a technology factor (TechFac)
18 Option 2(a)(ii)(2) - Direct translation of NEDC CO 2 & include effect of technology evolution Possible methodology CO 2 emissions for reference year for passenger cars CO 2 emissions (g/km) WLTP translated 95 g/km NEDC target Average CO 2 emissions for WLTP for reference year CO 2 (WLTP) Reference year Average CO 2 emissions for WLTP for reference year after making allowance for technology evolution Average CO 2 emissions for NEDC for reference year CO 2 (NEDC) Reference year Average CO 2 emissions for NEDC for reference year after making allowance for technology evolution Technology adjusted WLTP translated 95 g/km NEDC target 95 g/km NEDC target Key equation: WLTP Target (Option 2(a)(ii)(1)) = TechFac x WLTP Target (Option 2(a)(i)(1)) In illustration technology adjusted WLTP target increases because effect of emerging technologies give larger CO 2 reduction for NEDC than for WLTP
19 Options 2(b) Translating the specific emission target formulae
20 Option 2(b)(i) - Direct translation of NEDC emission target formulae Advantages: Is simple, being directly based on the overall target factors Can be used for Option 2(a)(i) & Option 2(a)(ii)(1) simple directly translated targets Option 2(a)(ii)(2) Directly translated targets including some consideration of the evolution of technologies Disadvantages: Challenge of translating the NEDC mass to WLTP test mass Might not minimise change for manufacturers when taken together
21 Option 2(b)(i) - Direct translation of NEDC emissions target formulae Outputs required: Values of the two key parameters in the specific emissions target formulae (e.g. as given in Annex 1 of Regulation EC/443/2009) Specific emissions of CO 2 = 130 + a x (M M 0 ) i.e. translated value for the gradient (a) And for the vehicle reference mass, which in Regulation EC/443/2009 is the mass in running order (as defined in Directive 2007/47/EC)
22 Option 2(b)(i) - Direct translation of NEDC emissions target formulae Methodology for providing outputs: For gradient of emissions target formulae (a) For passenger cars the relationships between the NEDC emissions target formulae (Reg EC/443/2009 and EC COM/2012/393) are: Gradient (parameter a ) changes by factor of 95/130, i.e. ratio of fleet average targets, and Reference mass (parameter M 0 ) is currently unaltered. Methodology for this option is to apply the same changes to the gradient that mirror those calculated for the changes in fleet targets I.e. if overall long term target for passenger cars goes from 95 g/km to Y g/km the revised gradient would go from 0.0333 g/km/kg to 0.0333 x Y g/km/kg.
23 Option 2(b)(i) - Direct translation of NEDC emissions target formulae Methodology for providing outputs: For vehicle reference mass of emissions target formulae (M 0 ) Possible options are: Retain the NEDC Vehicle mass definition given in Directive 2007/47/EC and develop translation algorithm between the WLTP test mass (which will be a value between TM L and TM H for each vehicle sold in for a model depending on the range of equipment fitted) and Vehicle mass as defined in Directive 2007/47/EC, (For this option M 0 would not need to change), Translate the NEDC Vehicle mass into an equivalent WLTP value through analysing the correlation between the WLTP vehicle test mass and the value of the NEDC Vehicle mass that applies to the vehicle (For this option the revised M 0 would be translated from its current value by the amount indicated from the correlation WLTP vehicle test mass and the value of the NEDC Vehicle mass )
24 Option 2(b)(ii) Adjust the NEDC emission target formulae to minimise change Advantages: Is designed to minimise systematic change For example, if the technologies used by the manufacturers of heavier vehicles were less effective at CO 2 emissions reduction over WLTP than NEDC, relative to the technologies used by the manufacturers of heavier vehicles, then no change in gradient would lead to the manufacturers of heavier vehicles having to work disproportionately harder to meet the targets. An appropriate increase in the slope would minimise this, to some extent. Could also minimise systematic change caused by the change in definition of vehicle mass between NEDC and WLTP I.e. caused by change in scale of mass axis. However, adjustment of slope to improve manufacturer to manufacturer comparable stringency will not impact directly on overall emissions target.
25 Option 2(b)(ii) Adjust the NEDC emission target formulae to minimise change Disadvantages: Is more complex than a simple change of gradient Challenge of translating the NEDC mass to WLTP test mass
26 Option 2(b)(ii) - Adjust the NEDC emission target formulae to minimise change Outputs required: Values of the two key parameters in the specific emissions target formulae (e.g. as given in Annex 1 of Regulation EC/443/2009) Specific emissions of CO 2 = 130 + a x (M M 0 ) i.e. translated value for the gradient (a) And for the vehicle reference mass, which in Regulation EC/443/2009 is the mass in running order (as defined in Directive 2007/47/EC
27 Option 2(b)(ii) - Adjust the NEDC emission target formulae to minimise change Methodology for providing outputs: For gradient of emissions target formulae (a) The relative changes in manufacturers distances to the target are analysed, and it is seen whether adjusting the slope of the gradient can reduce the changes caused by the introduction of the WLTP. I.e. function to be minimised = S Changes in differences from the target lines Manufacturers An illustrative example might help explain this.
28 Option 2(b)(ii) - Adjust the NEDC emission target formulae to minimise change CO 2 emissions (g/km) M1 + x + x Scaled NEDC target formulae rescaled to average WLTP CO2 emissions and adjusted to minimise change M3 Scaled NEDC target formulae rescaled to average WLTP CO2 emissions M2 + x + x + x + x + x + x M4 + x M5 + Scaled NEDC target formulae as specified in Regulation 411/2009/EC x Vehicle mass (kg) 1,372
29 Option 2(b)(ii) - Adjust the NEDC emission target formulae to minimise change Manufacturer NEDC WLTP Av CO 2 Diff from target line Av CO 2 Diff from target line Change in Diff from target lines Revised change in Diff from target lines Manufacturer 1 112.6-7.4 xxxx - 7.9-0.5 +0.5 Manufacturer 2 133.0-0.1 xxxx 0.0 0.0 0.0 Manufacturer 3 135.1 + 1.5 xxxx + 2.0 +0.5 0.0 Manufacturer 4 161.0 + 15.2 xxxx + 22.6 +7.4 +3.7 Manufacturer 5 156.9 +7.4 xxxx + 15.3 +7.9 +2.0
30 Option 3: Mixed approach Advantages and disadvantages Advantages: As for Option 1 for the transition period For short/medium term, i.e. during the transition period, confirms the existing regulatory framework. Legislation linked to the CO 2 targets at European (e.g. reporting and monitoring requirements), national (e.g. UK vehicle tax bands) and local (e.g. CO 2 based charge exemptions) levels remain valid. As for Option 2 following the introduction of WLTP The WLTP will be introduced and the conversion from NEDC to WLTP will be required at some date in the future Range of sub-options available for simple direct translation of targets, including consideration of technology evolution, and reducing the effects of NEDC testing rather than technology optimisation
31 Option 3: Mixed approach Advantages and disadvantages Disadvantages: It would be wrong to indicate that the mixed approach has no disadvantages. However, some disadvantages noted, e.g. for Option 1 that the WLTP will be introduced at some point and sticking with NEDC base target translation will be required at some date in the future, are substantially reduced by this option. Many of the challenges noted for Option 2(a) will remain and need to be addressed.
32 Conclusions and discussion Three different translation options have been presented: Option 1: Maintain NECD based targets Option 2: Introduce WLTP based targets Translating the fleet target Translating the specific emission target formulae Option 3: Mixed approach At a high level, do you agree these three options cover the scope of the target translation options adequately? At a more detailed level do the sub-options and the possible methodologies cover the scope of the target translation options adequately?
Thank you John Norris Ricardo-AEA Ltd The Gemini Building Fermi Avenue Harwell, Didcot, OX11 0QR T: E: W: +44 (0)1235 753685 John.norris@ricardo-aea.com www.ricardo-aea.com www.ricardo-aea.com