Approach for determining WLTPbased targets for the EU CO 2 Regulation for Light Duty Vehicles Brussels, 17 May 2013 richard.smokers@tno.nl norbert.ligterink@tno.nl alessandro.marotta@jrc.ec.europa.eu
Summary Goal Sketch of the overall approach Technical correlation exercise Determination of a generalised correlation function between CO 2 on the NEDC and WLTP Determination of equivalent WLTP-based targets overall target level slope of target line 2
Goal to translate the current NEDC-based fleet average CO 2 targets and the target calculation formulae for 2015 and 2020 into equivalent target levels and target calculation formulae for CO 2 emissions measured on the WLTP Starting point: target function will remain linear and mass-based: CO 2 = target + a (M M 0 ) 3
Overall approach technical correlation exercise This contains two sub-steps: Determining the CO 2 emissions measured on the NEDC and on the WLTP for a range of vehicle configurations and technology packages Determination of a generalised correlation function between CO 2 on the NEDC and the WLTP translation of the NEDC-based target to an equivalent WLTP-based target on the basis of agreed equivalence criteria using insights from the technical correlation exercise 4
5
The technical correlation exercise Two possible approaches: testing vs. modelling / simulations A mixed approach is proposed: mostly based on modelling combined with results from vehicle testing for three purposes: Pros: validation of the model results providing input for the model providing data on aspects of the tests that cannot be simulated (e.g. cold start or impact of variations in certain test conditions) cost effective combines pros of both approaches while avoiding many of the cons improves acceptance of modelling results allows evaluation of aspects of the tests that cannot be adequately simulated Cons: costs somewhat higher than for approach based on modelling alone 6
cons pros testing includes all relevant effects of vehicles, powertrains and test conditions very costly and time consuming can only be performed with existing vehicles using current day technology requires agreement on test conditions to be used within allowable bandwidths on the NEDC and WLTP rewards optimisations of vehicle hardware and software towards the NEDC modelling / simulation very cost effective allows evaluation of vehicles with advanced technologies that could be on the market by 2020 allows variation of a large number of vehicle and test parameters cannot cover all possibly relevant effects of vehicles, powertrains and test conditions requires agreements on model and model inputs (e.g. component modules, engine maps, etc.) requires some inputs from vehicle testing (e.g. coast down data, effect of cold start) 7
Steps in the technical correlation exercise Identification of the main differences between the NEDC test procedure and the WLTP Resulting in a table identifying: relevant aspects of the test procedures the extent to which these can be dealt with on the basis of simulations or additional testing activities. 8
1 Vehicle running-in NEDC < 3,000 km (petrol) < 15,000 km (diesel) WLTP Must have been run-in and driven between 3,000 km and 15,000 km before the test. No differentiation between petrol and diesel vehicles. Impact on NEDC/WLTP correlation (*) Impact reflected in modelling Testing required Data available from VP2 database or other sources? N N N Not relevant 2 Soak Between 20 C and 30 C At 23 C ± 3 C N N N Not relevant 3 Preconditioning test 4 Test procedure 1 NEDC + 2 EUDC (petrol) 3 EUDC (diesel) 4.1 Test mass Kerb mass + 100 kg 4.2 Inertia class Stepwise inertia classes 4.3 Test temperature Between 20 C and 30 C 4.4 RLD (F0, F1, F2) Based on test mass and best case conditions 4.5 Speed profile 4 UDC + 1 EUDC 4.6 Gear shift strategy Fixed speed (unless differently specified by manufacturers) WLTC for that class of vehicle N N N Not relevant TMH and TML, both higher than NEDC ref. mass Step-less approach: inertia class = test mass At start: 23 C ± 3 C During test: 23 C ± 5 C Based on TMH (+ aerodynamic parts) and TML (w.o. aerodynamic parts) L+M+H+exH with downscaling, if necessary. Will depend on the class of vehicle. Obtained from engine speed and power curve of the vehicle Y Y N Y Y Y N Y N N N Not relevant Y Y Y N Y Y Y N (speed profiles of Class 3 vehicles are not those tested in VP2) Y Y Y? Parameter relevant for simulation, set for WLTP Parameter relevant for simulation but not yet set for WLTP 9
4.7 Cooling fan 4.8 Cold start 4.9 Humidity Dynamometer road 4.10 load setting Alternative dynamometer road 4.11 load setting equation 4.12 Dynamometer road load: inertia of rotating parts 4.13 Bag analysis NEDC Constant speed (50 km/h) of the fan allowed above 50 km/h Contribution of cold start CO2 emissions higher than in WLTP Shall be measured to ± 1 g H 2 O/kg air Table values allowed ECE-R 83 allows setting coefficients as follows: F=a + b*v² ± ( 0.1 * F80) where F80 is the load at 80 km/h Shall be known and be within ± 20 kg of the inertia class for the test. 20 minutes after end of cycle WLTP Proportional, with a 10% margin between cycle max speed and fan speed Contribution of cold start CO2 emissions less than in NECD Impact on NEDC/WLTP correlation (*) Y? (probably of secondary importance if one assumes same fan for the two test procedures) Y Impact reflected in modelling Y? (probably of secondary importance if one assumes same fan for the two test procedures) N/Y Powertrain models generally only simulate hot test, but experimental data can make modelling of cold start possible Testing required maybe some testing could be useful to check impact of this issue Y Vehicle tests with cold & hot start on NEDC & WLTP can provide input for determining general cold start addition factors Data available from VP2 database or other sources? Measured to ± 5 per cent N N N Not relevant Still to be determined if table values will be allowed N N N Not relevant Not in the GTR N N N Not relevant May be estimated to be 3 per cent of the unladen vehicle mass for permanent 4WD and 1.5 per cent for 2WD vehicles 30 minutes after end of cycle phase (to be determined: end of cycle phase or end of cycle?) Y? N? Y? N N N N Y N Y Parameter relevant for simulation, set for WLTP Parameter relevant for simulation but not yet set for WLTP 10
5 Rechargable Energy Storage System (RESS) No provision NEDC WLTP Monitored and the delta energy balance is taken into account for fuel consumption Impact on NEDC/WLTP correlation (*) Impact reflected in modelling Testing required Data available from VP2 database or other sources? Y Y N? N 6 Cycle allocation No provision WLTP has provisions Y Y Y N 7 CO2 correction with Europe average temperature None The value measured at 23 C will be reported to 10 12 C N? N N N 8 Fuel As WLTP As NEDC N N N N 9 DPF regeneration As prescribed in R101 Not yet clear how this is defined Y N Y? 10 Hybrids 10.1 NOVC hybrids As prescribed in R101, with ΔSOC correction (or using test mode not covered by Not yet clear how this is defined? Y N? procedure) 10.2 OVC hybrids 10.2.1 - test A and B As prescribed in R101 Not yet clear how this is defined Y Y N? 10.2.2 - electric range As prescribed in R101 Not yet clear how this is defined Y Y N? Parameter relevant for simulation, set for WLTP Parameter relevant for simulation but not yet set for WLTP 11
Steps in the technical correlation exercise Definition of a range of vehicle configurations: Starting point is a sales database for Europe for the selected reference year Reference year could be 2009 It is proposed to divide the passenger car market in 6 segments Separately for petrol and diesel For each segment a representative vehicle model will be selected based on the average CO 2, mass and performance in that segment Sales weighted averaging of results per segment yield overall result for fleet average 12
Proposed definition of market segments 13
Proposed definition of market segments The European sales divided in light, medium weight, and heavy vehicles, show a shift and increasing spread of NEDC CO 2 emission with increasing vehicle weight 14
Approach for vans and alternative energy carriers A similar approach will be used for Light Commercial Vehicles In that case it may suffice to only look at diesel vehicles Also the necessary number of market segments may be different Alternative energy carriers (e.g. natural gas) and alternative propulsion systems (specifically plug-in hybrids) will also be assessed. Methods for that are still to be determined 15
Steps in the technical correlation exercise (1) For each reference vehicle configuration define: a baseline powertrain configuration a set of technology packages combining a range of relevant CO 2 reduction technologies minimum performance criteria acceleration, top speed, trailer towing capability, high climbing ability Modelling of the different combinations in a detailed powertrain simulation tool all packages dimensioned to meet minimum performance criteria of reference vehicle possible requirements for tool: see next slide 16
Proposed requirements for simulation tool The advanced powertrain simulation tool to be used for the correlation exercise should be a so called forward calculating physical simulation model, taking account of all relevant mechanical, electrical, and if necessary thermal interactions between powertrain components. The simulation tool to be used for this correlation exercise should be one that is accepted by industry stakeholders. This means that it could be a tool that is widely used by OEMs or their technical consultants and R&D suppliers for actual powertrain R&D and development activities. 17
Steps in the technical correlation exercise (2) Determination of CO 2 emissions of the different combinations on the NEDC and the WLTP: Simulation of the above defined vehicle configurations and technology packages over the NEDC and WLTP Simulations augmented by a limited amount of dedicated vehicle testing activities to: provide inputs to the correlation exercise that relate to aspects of the test procedure that cannot be adequately simulated validate the simulation model 18
Reduction technologies to be included 1. Engine e.g. downsizing, cycle change, valve actuation, petrol direct injection 2. Transmission e.g. automated MTs, dual clutch, high-gears, gear ratios 3. Electrification / hybridisation e.g. start-stop, regenerative braking, boost-power with downsizing, full hybrid 4. Resistance reduction e.g. weight reduction (body-in-white), drive-line components, aerodynamic improvements 5. Novel technologies e.g. waste-heat recovery, thermal management, auxiliary and battery optimisation 19
Simulation of vehicle configurations and technology packages over the NEDC and WLTP reduction potential of technology / package of technologies on WLTP may be different from that on NEDC 20
Simplified analysis carried out in parallel to starting up simulations Two different simplified approaches: theoretical analysis of power at the wheels + analysis of available test results (JRC) using CO 2 emission factor modelling (TNO) Serves two purposes: insight in basic mechanisms / determinants useful for scoping detailed simulations delivering a draft correlation function useful as first input to process of determining the translation algorithm 21
Determination of a generalised correlation function between CO 2 on the NEDC and WLTP A statistical analysis of results from the technical correlation exercise to determine a generalised correlation function ( meta-model ) between CO 2 on the NEDC and WLTP. The correlation factor can be a function of e.g.: vehicle mass CO 2 or ΔCO 2 (relative to reference vehicle) fuel type applied technologies other parameters 22
Determination of equivalent WLTP-based overall target level and slope of target line Defining overall target levels for 2015 and 2020 As a starting point for defining equivalent overall fleet average target levels for 2015 and 2020 it is proposed to consider as the equivalence criterion: that relative reductions of the WLTP-based fleet averages, relative to the value in a reference year, are the same as for the NEDC-based targets. For the 2020 target it could be justified to apply an upward or downward adjustment to the value so obtained, on the basis of: further evaluation of appropriate equivalence criteria insights obtained in the technical correlation exercise, specifically from simulations of advanced technologies for meeting 2020 target Defining the slope of the target calculation formulae for 2015 and 2020 23
Defining overall target levels for 2015 and 2020 Fleet average CO 2 emission CO 2 (NEDC) ref in the reference year is translated to an equivalent WLTP-based fleet average CO 2 emission CO 2 (WLTP) ref in the reference year Based on the insights gathered in the technical correlation phase, the The WLTP-based target for 2015 is then determined as: CO 2 (WLTP) 2015 = CO 2 (WLTP) ref x 130 / CO 2 (NEDC) ref The WLTP-based target for 2020 is then determined as: CO 2 (WLTP) 2020 = CO 2 (WLTP) ref x 95 / CO 2 (NEDC) ref 24
Proposed starting point for defining equivalent overall target levels for 2015 and 2020 Determine WLTP equivalent of NEDC average in reference year apply relative reductions based on NEDC average and targets 25
Defining the slope of the target calculation formulae for 2015 and 2020 Determination of an appropriate slope of the target calculation formula CO 2 = target + a (M M 0 ) The appropriate methodology should be defined taking into account the findings of the technical correlation phase and will be subject to further stakeholder consultation 26
TA CO2 emission [g/km] Example of options for defining equivalent slope of target calculation formula NEDC-based target line WLTP-based target line ΔCO 2 on NEDC equivalent ΔCO 2 on WLTP avg.reference mass NEDC avg. reference mass WLTP NEDC-based target equivalent WLTP-based target OEM NEDC value in reference year OEM WLTP value in reference year translated OEM target Reference mass [kg] based on determining equivalent required CO 2 reduction on WLTP per OEM or per vehicle model or per size class target line is sales-weighted fit through translated targets 27
Example of options for defining equivalent slope of target calculation formula equivalent required CO 2 reduction on WLTP could be: equal CO 2 reduction as required for NEDC CO 2 reduction required for NEDC corrected for generalised change in effectiveness of technology packages on the WLTP based on correlation function derived from technical correlation exercise CO 2 reduction that can be achieved at same effort as required reduction on NEDC takes explicit account of possible changes in applied technologies for meeting a WLTP-based target 28
Questions / discussion issues Collect stakeholder responses to the simulation-based approach proposed for the technical correlation exercise Collect stakeholder feedback on details of the proposed approach for the technical correlation exercise, including: requirements for the simulation tool the market segmentation further provisions for defining reference vehicles selection of technology packages to be included in the simulations in view of the switch from NEDC to WLTP essential differences between NEDC and WLTP to be included in the simulations or to be covered by additional testing 29
Questions / discussion issues Assess willingness of stakeholders to contribute to the technical correlation exercise, specifically by means of supplying test data Discussion with stakeholders on the overall approach suggested for determination of equivalent WLTP-based overall target level and slope of target line, including: the choice of reference year the proposed starting point for defining equivalent fleet average target levels for 2015 and 2020 30