Emission factor modeling in Istituto Motori - KEM. Maria Vittoria Prati, Livia Della Ragione, Giovanni Meccariello, Maria Antonietta Costagliola

Transcription

1 Emission factor modeling in Istituto Motori - KEM Maria Vittoria Prati, Livia Della Ragione, Giovanni Meccariello, Maria Antonietta Costagliola This presentation is dedicated to the memory of Mario Rapone who was keen to see this project brought to fruition. Expert Meeting on the Development of Emission Factors for Road Transport Sector JRC Ispra, October 2008

2 SUMMARY Background Modeling approach The core of the Model Base The development of an User Interface...in progress References

3 BACKGROUND Emission evaluation can be performed for different purposes and uses at different space and time scales Macro scale models are able to define emission inventories and calculate overall average emission estimation based on a few input parameters such as fleet composition and average speed Meso scale models are necessary to develop region and city transit plans. They require detailed representation of the area traffic and transit network into links. For each link average emissions are generally evaluated for each individual vehicle as a function of vehicle flow and average speed Micro scale models are istantaneous emission models and they require more detailed analysis of vehicle driving behaviour data represented by the time series of vehicle speed on the link

4 MODELING APPROACH The kinematic model approach was developed by Istituto Motori in ARTEMIS context. This kinematic model (Kinematic Emission Model KEM) is defined as a meso-scale emission model It predicts average emissions on a driving pattern obtaining a detailed description of the velocity profile using a multidimensional approach. It is based on the analysis of an existing huge emissions database, different fleet composition, different driving sub-cycle and laboratories involved in the emissions measurements, coming from the ARTEMIS EU project.

5 MODELING APPROACH A Decision Support System (DSS) approach for modelling and predicting vehicle real-world emissions as a function of driving behavior kinematics DSS components are the emission data warehouse, the model base and the grafical user interface (GUI) At the present the KEM model has been only developed for cars in function of tecnology, type approval stage and engine capacity. Only for hot conditions.

6 Data Base SYSTEM STRUCTURE

7 The Model Base A consistent set of kinematic parameters is used to represent the real-world driving behavior of vehicles in any traffic situation The response considered is the measured unit emission mass of CO, HC, NOx, CO 2 (expressed in g/km), measured in a driving cycle (DC), that is a portion of a trip Principal Component Analysis and the Partial Least Squares (PLS) based on PCA variables statistical methods are implemented in the system.

8 The Model Base The explicative variables characterize the kinematics of driving cycles and are related to the dynamic vehicle equation, plus idling time to consider stand-still phase emission production M P(t)dt t (a 0 a 1 *v(t)a 2 *v 2 (t)m v *a(t))*v(t) dt ln Y a a ij 4 6 = a MVA Tidle 0 _ a POS a 1 7 MV a Invdist a 5 2 MV Trunning ε ij 2 a 3 Where: M total mass of exhaust, P(t) engine power, v(t) speed, v(t) acceleration, M v mass of the vehicle MV 3

9 Artemis database for cars Diesel IMPORTANT: in similar test conditions TYPE-APPROVAL STAGE/ ENGINE DISPLACEMENT Euro 1 Euro 2 Euro 3 Euro 4 < 2000 E1 D 2000 E2 D 2000 E3 D 2000 n.a. Gasoline TYPE-APPROVAL STAGE/ ENGINE DISPLACEMENT Euro 1 Euro 2 Euro 3 Euro E1 GAS E2 GAS E3 GAS E4 GAS E1 GAS E2 GAS E3 GAS > 2000 n.a. n.a. E3 GAS 2000 n.a.

10 Artemis-Driving cycles mv OSCAR.H1 OSCAR.H2 OSCAR.H3 OSCAR.G2 Napoli.10_23 Handbook R4 - LE6 tr1 StGoHW tr2 OSCAR.G1 Napoli.15_18_21 inrets.urbainfluidecourt Artemis.LowMot_urbdense Artemis.HighMot_urbdense TRL.Urban Artemis - Urban (921 Artemis.urban_incl_start Artemis.LowMot_urban Artemis.HighMot_urban OSCAR.F OSCAR.D1 OSCAR.D2 Artemis.HighMot_freeurban OSCAR.E modem.urban5713 OSCAR.C Handbook R3 - LE2u tr1 LE3 tr2 TUG.Ries_RoadGradient 0perc TRL.Suburban&Rural modemhyzem.road Napoli.6_17 Artemis.rural_incl_pre_post Artemis.rural LDV_PVU.CommercialCars.road Artemis.HighMot_rural Handbook R2 - A4 tr1 LE1 tr2 Artemis.motorway_130_incl_pre_post Artemis.motorway_150_incl_pre_post TRL.Motorway Handbook R1 - AE1 tr1 AE2 tr2 Artemis.motorway_130 Bundesautobahn(EMPA.BAB) Artemis.LowMot_motorway Artemis.motorway_150 Artemis.HighMot_motorway 42/160 Useful driving cycles

11 MODEL RESULTS NOX g/ Km NOx [g/km] Black: mean speed model 0.3 KEM model, including or not including High emitters v_overall An example: Euro 3 Gasoline cars cc Hot emissions

12 Datawarehouse THE USER INTERFACE Model Base GUI PM (g/km) Cycle 7 Cycle 5 Cycle 6 Cycle 4 Art-Urban Art-Road Art-Motorway Urban_Cold Road_Cold Data base Data mart

13 THE USER INTERFACE Easy module: able to predict emission factors based on the input of several parameters The input variables are: - Driving cycle profile; -Homologation and displacement; The core of the tool is a set of regression equations applying an empirical model which calculates from the PLS approach the vehicle emissions in real-world hot driving conditions

14 CONCLUSIONS AND DEVELOPMENTS It is in progress the elaboration of a graphical user interface (GUI) for the production of reports This tool of DSS has to be able to interact with user Continuous data warehouse update The full development of DSS is in progress Include two wheelers Consider Cold Start Include Particulate emissions in terms of particle mass (PM) and number (PN)

15 REFERENCES M. Rapone, L. Della Ragione, G. Meccariello, An integrated knowledge base for modelling and predicting vehicle real-world emissions as a function of driving behaviour kinematics, Proceedings of ITI Conferences, 2008 M. Rapone, M.V. Prati, M.A. Costagliola, L. Della Ragione, G. Meccariello, Emission Factors Determination of Euro III 1,200- to 1,400-cc Petrol Passenger Cars with a PLS Multivariate Regression Model, Environmental Modeling and Assessment, Vol.13(3), pp , Springer Netherlands, 2008 Joumard R., Andre J.M., Rapone M., et alii Emission factor modelling and database for light vehicles, Artemis deliverable 3, 2007

16 Data Quality Approach Data quality approach Data source acquisition Source Data Analysis Case study MAD Analysis Data warehouse Artemis DataBase Identification and Case Study extractions Filter Petrol case Diesel case MAD Outliers pruning All cases Data warehouse