Euro6d RDE EVALUATION METHODS / HYBRIDS > Package#3 -Calculation Methods > Package#4 -Further Analysis of Proposals > 16Oct 17

Transcription

1 Euro6d RDE EVALUATION METHODS / HYBRIDS > Package#3 -Calculation Methods > Package#4 -Further Analysis of Proposals > 16Oct 17 ACEA TF-RDE SUB-Group EVALUATION METHODS / HYBRIDS

2 1. Package #3 Calculation Methods 2

3 EURO6D-RDE PACKAGE#3 CALCULATION METHODS o Progress still required to ensure that everyone can calculate the same results from any given RDE Data Transfer File as per Package#3 Regulations currently in force... o Golden dataset required for any/all calculation methods to be checked against? o ACEA Round-Robin datasets shared in May 17 of any use to help identify required characteristics of golden data? o ACEA Round-Robin exercise identified a list of items within the regulations where we have experienced differences of interpretation supplied to JRC 03Aug17 some items resolved in Q&A / some not (see following 2 slides). Further clarification required. o TNO study Report / 29Sep17 RDE-LDV Presentation mentioned the discovery of potential ambiguities when attempting to calculate results as per the regulatory text. Have these been identified to JRC / are there further items to those noted by ACEA study? 3

4 ACEA ROUND-ROBIN ISSUES IDENTIFIED #1 Initial Data Analysis > Raw Trip > CO2 Ratio CALCULATION STEP Application of Ext=1.6 Calculation of a-res Hanning Filter smoothing of vehicle speed. Corrected / smoothed altitude +ve Altitude Gain for Urban trip part only Altitude difference between start and end point of the trip INFO AnnexIIIA 9.5 appears to indicate that Ext=1.6 is applied at the initial data generation stage, but then there is no mention of it in Appendix4. So not clear if it is already applied to the 1Hz data in the Exchange File, or if it is to be applied at the Results Evaluation phase? (Raised in ACEA presentation at 4/5May17 RDE-LDV) clarification required How should acceleration be calculated in order to perform the a-res check in Appendix 7a On a 1s basis or by using the 2s method as detailed in Appendix 7a 3.1.2? (Discussed at 4/5May17 RDE-LDV) If vehicle speed is smoothed in Appendix7a (&, we assume, used for V.A+ & RPA calcs - and not just in section 3.1.2) should it then also be used as the vehicle speed for all other Results Evaluation calculations in AnnexIIIa & Appendices? (Raised in ACEA presentation at 4/5May17 RDE-LDV) Should the corrected altitude data from Appendix7b Section 5.2 Or the smoothed altitude data from Appendix 7b Section Be used elsewhere in results evaluation - i.e. for Trip start-end delta elevation determination of "extended" altitude conditions in AnnexIIIa 5.2? Or should that assessment still just use the original altitude data from the Exchange File - that has only been subjected to the consistency check in Appendix 4 Section 6? (Question sent to ec-rde@ec.europa.eu 14Jun17) How to calculate +ve Altitude Gain for Urban part of RDE Trip - as included in Package#3? (Raised in ACEA presentation at 4/5May17 RDE-LDV) Calculate from raw altitude signal, or corrected / smoothed altitude data? WLTP total CO2 g/km How is the WLTP total grams of CO2 calculated (to be used x0.5 as CO2 Ref for MAW analysis)? Exchange File is to contain g/km for each value - for MAW CO2 Ref phase and total g/km for whole WLTP... Use actual phase distances for the Type Approval test data on the CoC, or schedule distances for the and for CO2 WLTP - but need to know which version of WLTP the data refers to)? normalisation (ACEA Lessons Learned suggests Add "WLTC CO2 total [g]" in the Data Exchange File) Ki Values How are Ki values to be applied - as introduced in Package#3? +ve &/or multiplicative, to both Total Trip & Urban results? (Included in ACEA Lessons Learned info). 4

5 ACEA ROUND-ROBIN ISSUES IDENTIFIED #2 Moving Averaging Windowing Power Binning COLD START Duration Window Creation Average calculation CO2 characteristic cuve weighting factor Veline calculations - WLTP accel Vref & Aref weighting factor 3s moving average What is the correct mathematical interpretation of cold start considering "cold start period ends once the coolant has reached 343 K (70 C) for the first time"? CS is when after start: a) Tcool<70 C b) Tcool<=70 C, but includes only the 1st occurence of Tcool=70 C c) Tcool<=70 C, but includes all occurence of Tcool=70 C d) Tcool>70 C see example 1) Which calculation method with i-indices "1" and "2" is behind of "The duration (t2,j t1,j) of the jth averaging window is determined by..." a) calculation by their times b) calculation by the samples see example 2) What is the correct duration of e.g. window lengths with definded start and end time points? a) duration exclude start and end point (intergral interpretation) b) duration excludes end point only c) duration excludes start point only d) duration includes start and end point (stepwise data interpretation) see example 2) Should windows be opened on a strict "one per second" basis, or only when there is data valid for MAW analysis?: > Vehicle Speed >= 1 km/h > Gas Measurement Active (Question sent to ec-rde@ec.europa.eu 03Jul17) What is the correct average calculation with definded start and end time points? a) duration exclude start and end point (intergral interpretation) b) duration excludes end point only c) duration excludes start point only d) duration includes start and end point (stepwise data interpretation) see example 2) CO2 characteristic should be defined as 1.0 x WLTP Ph1, 3, 4 g/km CO2 values, and not the multipliers carried-over from NEDC, as per ACEA Lessons Learned. If Tol1 increased from +25%, in 1% steps up to 30% (50% for NOVC-HEV) for Trip Completeness and Validity determination, are Wj values calculated on the modified values or on the original Tol1=+25% value? (At 4/5May17 RDE-LDV Pierre Bonnel stated that Wj values should always be calculated with Tol1=+25%) How should acceleration be calculated in order to generate the Veline... by using the 2s method as detailed in Appendix 7a 3.1.2? (Confirmed by check with Prof Hausberger and reported back to DG-GROW/JRC, that 2s acceleration calcuation to be used for Veline determination. - s 16May-14Jun17 ) Which v-ref & a-ref values should be being used for Euro6d-Temp (RDE#1 - Packge#3) (66 km/h and 0.44 m/s2 confirmed by P.Dilara s 16May-14Jun17) Which values to use for trip time share normalisation - rounding errors...eg. in Bin3 and Bin9 (see Table1-2 vs Table1 Sum <>100% Should now be using values from WLTP/Package#3 Regs (although Urban shares do not quite sum to 100%)? Appendix 6 section 4 What is meant with "...measured CO2 mass flow in 1 Hz basis..." a) "mgas, 3s, k" = 3-second moving average mass flow of the exhaust gas component gas=co2 in time step k given in 1 Hz resolution [g/s] b) "mgas, i" = Instantaneous mass of the exhaust component gas=co2 at time step i, [g/s] 5

6 URBAN PHASE +VEALTITUDE GAIN CALCULATION o ACEA proposal for calculation of altitude gain for the urban part: (i) Run the 3 step procedure defined in Appendix 7b, and accumulate +Ve Altitude Gain for the Total Trip... Further data in back-up (ii) Back-calculate VS i =1/(tw i -tw i-1 )*60^2/1000 Classify Urban<=60km/h Accumulate +Ve Altitude Gain for the Urban Part 6

7 PACKAGE#3 APPENDIX7C PHEVEVALUATION METHOD o Issue with Package#3 Appendix7c PHEV Evaluation Method as discussed during Jul-Aug17... o Resolution required there was mention, at 03Aug17 ACEA > EuCom telco, of the possibility of including distance ratio in the App7c calculation, ahead of Package#4? 7

8 2. Package #4 Further Analysis of Proposals 8

9 INTRODUCTION o Following 29Sep17 RDE-LDV presentations and discussions... o Further review of data and analysis of the Final Methods to be analysed after September (17) o Focussing on clearly distinguishing between: o Basic Trip requirements. o RDE Boundary Conditions. o RDE Validity Criteria avoidance of Intentionally Biased Driving. o Evaluation of RDE results for a valid test. RDE test > PEMS data processing > Exchange File Trip Requirements as per AnnexIIIa Y Y Y Kinematic Boundary Conditions as per Appendix 7a & 7b Trip Validity check... CO2 based method Results Evaluation... CO2 Ratio CO2 x Distance Ratio/0.85 for PHEV (Emissions/km for high SoC) RDE Test Results N N N Test Not Valid Test Not Valid Test Not Valid 9

10 2.1 Basic Trip Requirements 10

11 BASIC TRIP REQUIREMENTS o Basic Trip Requirements, as defined in Regs-time / distances / average speeds / stop times, etc, o Should be ensured by suitable route design and validation o And driver familiarisation o etc... as per the PEMS Guidance Document. 11

12 BASIC TRIP REQUIREMENTS PHEVCOLD START o Although it has been noticed that the Package#3 Cold Start requirements are likely to cause problems with Basic Trip Requirements for PHEV CD mode RDE testing... o Therefore further consideration required. OEM JRC CD Test CD Test#1 CD Test#2 Time for first engine start s Time for 5mins cumulative engine run s Time for Coolant>70C s Cold Start Ave Spd km/h km/h Cold Start Max Spd km/h < 60 km/h Cold Start Veh Stop Time s < 90s 12

13 2.2 Boundary Conditions 13

14 BOUNDARY CONDITIONS o Once an RDE trip has been confirmed to comply with the Basic Trip Requirements, then it needs to be checked to ensure compliance with the Boundary Conditions as defined in the Regulations... following much debate over recent years: Ambient Temperature & Altitude V.A+ & RPA 90% payload +Ve Altitude Gain (Total & Urban) 14

15 +VEACCELCOUNTS FOR THE MOTORWAY PART o However the 150 +veaccel. counts for the motorway part of the trip is unrealistic, physically and mathematically, and is leading to invalid trips from Appendix7a > then the need for biased driving in order to achieve a valid trip as per Lessons Learned input Nov 16 > Jun 17: o Not addressed in Package #3 therefore urgent attention required. 15

16 2.3 Trip Validity Criteria > Avoidance of Intentionally Biased Driving 16

17 AVOIDANCE OF INTENTIONALLY BIASED DRIVING o Once an RDE trip has been confirmed to comply with both the Basic Trip Requirements and the Boundary Conditions then a further Trip Validity assessment is required to check for any Intentionally Biased Driving. o Which could possibly occur anywhere within the valid Boundary Conditions ranges (i.e. is not the same thing as pushing to the edges of the Boundary Condition envelope). o With an assessment based on CO2 proposed as the most appropriate approach with the data available from a legislatively compliant RDE test. 17

18 TRIP VALIDITY CRITERIA WLTP REFERENCE o The existing MAW reference Characteristic Curve based on WLTP CO2 x Scaling Factors, leads to the need for biased driving to achieve a valid trip as per Lessons Learned input: o (Unlike the PB reference Veline which is based on WLTP CO2 values without any scaling). o The reference for a Modified MAW based Trip Validity Criteria should be WLTP values without any scaling factors: Further data in back-up o And for a Combined MAW & PB approach should continue to use a Veline based directly on WLTP values: 18

19 TRIP VALIDITY CRITERIA MODIFIED MAW o Ref. question from 29Sep17 RDE-LDV... Were all of these trips valid for Basic Requirements and Boundary Conditions, and therefore this comparison only considers the difference between existing and proposed MAW based validity checks? o Confirmation = Yes, trips all valid by other criteria at least to Package #1/#2 (some earlier tests may not comply with Package#3 Cold Start criteria). 19

20 TRIP VALIDITY CRITERIA ACEAPROPOSALS MODIFIED MAW / MAW&PBCOMBINED o Further analysis of Trip Validity Criteria method options proposal... o Collate 2 sets of RDE trip data only a limited set of parameters required: o Normal Trips that should pass a validity check. o Biased Trips that should be expected to fail. o Analyse with September methods options: Table 1 16 Engine Rated Power kw 25W Road load parameters from WLTP F0, F1, F2 Plus WLTP ITWC kg 27 Type-Approval CO2 emissions g/km 28 CO2 emissions in WLTC mode Low g/km 29 CO2 emissions in WLTC mode Mid g/km 30 CO2 emissions in WLTC mode High g/km 31 CO2 emissions in WLTC mode Extra High g/km 32 Vehicle test mass (RDE test condition) kg Table 2 Time Vehicle Speed Altitude Exhaust Mass Flow Rate CO2 mass Engine Speed (if available) (or Idle Exh Mass flow rate) s km/h m kg/s g/s rpm kg/s Including consideration of adjusting Tol1 / severity thresholds etc, if required to improve success... 20

21 TRIP VALIDITY CRITERIA ACEAPROPOSALS MODIFIED MAW / MAW&PBCOMBINED o Determine capability of methods: o Retain as high a %age of Normal trips as possible. o Exclude as high a %age of Biased trips as possible... RDE Data Vehicle Type TUG Veh. 1 ICEV OEM X PHEV Route Arzberg Ries tbd Trip ID Normal Subjective assessment of the Trip Supplier!!! Annex IIIA Results of the check per Annex IIIA Appendix 7a & 7b Results of the check per Appendix 7a & 7b Eval. Method XY Results of the check by Eval. Method XY Trip Valid Endresults for Trip validity 89% ( = 1 ) if Subjective assessment = Endresults for Trip validity ( hit rate in %) o To enable a decision on the best method to propose for Package#4. 21

22 2.4 Results Evaluation > By Trip Normalisation 22

23 RESULTS EVALUATION BY TRIP NORMALISATION o Once an RDE test has been confirmed to be compliant with Basic Trip Requirements and Boundary Conditions then the emissions results can be evaluated. o Euro6d-RDE encompasses a wide range of Boundary Conditions vs. WLTP 50%ile nominal and emissions evaluation by Trip Normalisation is the method by which the results can be analysed to determine whether they correspond to those measured at Type Approval. o With CO2 as the best available indicator of cycle energy... o Trip Normalisation by WLTP/RDE CO2 ratio being the most viable option available. o Plus correction by ICE Distance Ratios to compensate for PHEV EV distance driven to enable RDE testing from any initial Battery SoC/ combination of ICE and EV operation. 23

24 PHEV TRIP NORMALISATION BY CO2 * ICE DISTANCE RATIO o From presentations & discussions through March September 2017 there seems to be a good understanding that PHEVs should also have results evaluation by trip normalisation, to allow testing in any battery SoC/ Hybrid Mode across the full range of RDE Boundary Conditions. o With CO2 * ICE Distance Ratios being the best approach possible, given the data available. o However concern remains w.r.t. the switch to non-normalised results analysis for high SoC/ high EV mode tests at x<0.2? o ACEA s view is that this is not a significant issue / there will not be a fundamental discontinuity in the analysis at one side of the threshold vs. other / any reason to bias PHEV operation modes in order to try and ensure a particular outcome. o Explanation of this on the next slides / further discussion welcome... 24

25 PHEV LOW RDE CO2 & ICE RATIO o PHEV Trip Normalisation is required, to account for testing across all Boundary Conditions, and can be performed by comparing CO2 g/km during ICE operation on the RDE test against the CS WLTP reference: o Therefore the [CO2 Ratio * ICE Distance Ratio/0.85] evaluation should be applied across as wide a range of PHEV RDE tests battery SoC& EV operation as possible. o However as ICE operation reduces, due to high SoC&.: EV operation, the RDE CO2 value becomes a less valid indicator of the overall RDE trip cycle energy and the calculation risks becoming mathematically unstable as both RDE CO2 and ICE-Ratio will tend to small numbers: Small Value Small Value o Therefore x should not be set at too low a value, to avoid introducing variation into results. 25

26 PHEV x<0.2threshold o ACEA s 29Sep17 proposal, based on the time-tested 80:20 rule, is to set x=0.2: o And initial review of available PHEV RDE test data has not indicated that this would be an unsuitable value to use (perhaps even lower than 0.2 is ok for x?)... Further data in back-up o Proposal: Add any further available data to this analysis and finalise value for x. 26

27 Back-Up Data Urban +ve Altitude Gain 27

28 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION Current calculation of altitude gain for the urban part defined in App. 7b

29 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION High Value for Urban compare with slide 7 URBAN Arzb.01 Road Gradient incorrect due to discontinuity. Range [ ] Compare with Total Trip Range [ ] next slide

30 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION TOTAL Arzb.01 Road Gradient for the Total Trip Range [ ]

31 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION Urban Ries 01 -> Fail compare with slide 9 URBAN Ries 01 Road Gradient incorrect due to discontinuity. Range [ ] Compare with Total Trip Range [ ] next slide

32 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION TOTAL Ries 01 Road Gradient for the Total Trip Range [ ]

33 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION ACEA proposal for calculation of altitude gain for the urban part: 1. Run the 3 step procedure defined in Appendix 7b 1.1 Accumulate +Ve Altitude Gain for the Total Trip 2. Back-calculate VSi = 1/(twi-twi-1)*60^2/ Classify Urban<=60km/h 2.2 Accumulate +Ve Altitude Gain for the Urban Part

34 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION Compare with Slide 2 Urban value URBAN Arzb.01 The Urban Range [ ] is within the Range of the Total Trip Range [ ] s. next slide

35 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION TOTAL Arzb.01 Road Gradient for the Total Trip Range [ ]

36 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION Urban Ries 01 -> OK compare with slide 4 URBAN Ries 01 The Urban Range [ ] is within the Range of the Total Trip Range [ ] s. next slide

37 ACEA TF-RDE -EVALUATION METHODS / HYBRIDS SUB-GROUP URBAN PHASE POSITIVE ALTITUDE GAIN CALCULATION Values for Total Trip remain unchanged s. slide 5 TOTAL Ries 01 Road Gradient for the Total Trip Range [ ]

38 Back-Up Data MAW Characteristic Curve 38

39 TRIP VALIDITY CRITERIA WLTP REFERENCE 39

40 TRIP VALIDITY CRITERIA WLTP REFERENCE 40

41 TRIP VALIDITY CRITERIA WLTP REFERENCE 41

42 Back-Up Data PHEV CO2 x Distance Ratio vs. x 42

43 PHEV x<0.2threshold X WLTP CO2/RDE CO2 ICE Ratio CO2 Ratio x ICE Ratio/0.85 t u t u t u t u OEM Gasoline SUV JRC Gasoline TUG Gasoline C-Segment TUG Diesel E-Segment OEM Diesel E-Segment