COMPARISON OF EXHAUST AND NON-EXHAUST EMISSION FACTORS IN NORTHERN EUROPE

Transcription

1 COMPARISON OF EXHAUST AND NON-EXHAUST EMISSION FACTORS IN NORTHERN EUROPE SMHI, 4-5 October 25 Matthias Ketzel 1, Gunnar Omstedt 2, Christer Johansson 3, Ingo Düring 4, Lars Gidhagen 2, Achim Lohmeyer 4, Ruwim Berkowicz 1, Peter Wåhlin 1 1 National Environmental Research Institute of Denmark, 2 Swedish Meteorological and Hydrological Institute, 3 Environment & Health Protection Administration, Stockholm, (also at Department of Applied Environmental Research, Stockholm University, Sweden), 4 Ingenieurbüro Lohmeyer GmbH & Co. KG, Radebeul, Germany

2 Vehicle PM emissions - exhaust Emission databases (HbEfa, COPERT,..) rely on laboratory test bench measurements focus so far on the exhaust pipe emissions

3 Vehicle PM emissions - non-exhaust substantial part of the emissions comes from non-exhaust: need for field studies to estimate the emissions at: road tunnel, street canyon, highway, etc. using: NOx / C as tracer or inverse modelling

4 Particle sources Influencing factors primary emissions exhaust emissions tyre wear road wear brake wear clutches... re-suspension of road dust and other deposited primary emissions vehicle type diesel // gasoline brake type tyre type summer // winter studded tyres road surface type / condition sanding salting weather conditions rain, snow, ice

5 Measured variation of PM /NOx ratios street contribution only (background subtracted) direct TEOM results - no correction factor PM 1 / NOx [g/g] PM 1 Summer Summer Summer PM PM1_Luet_DE PM1_FFAl_DE PM1_Horg_SE PM1_Rugb_FI PM1_Hcab_DK.6 PM1_Jgtv_DK PM25_Horg_SE PM25_Rugb_FI.5 PM25_Hcab_DK PM 2.5 / NOx [g/g].

6 Measured variation of PM2.5 /NOx ratios street contribution only (background subtracted) direct TEOM results - no correction factor PM25_Horg_SE PM25_Rugb_FI PM25_Hcab_DK Summer Summer Summer PM 2.5 / NOx [g/g].

7 HCAB - PM2.5 all data 9/23-24 HCAB_PM25str 12 1 d PM y =.394x R 2 = d NOx

8 Measured variation of PM1 /NOx ratios street contribution only (background subtracted) direct TEOM results - no correction factor PM1_Luet_DE PM 1 / NOx [g/g] PM1_FFAl_DE PM1_Horg_SE PM1_Rugb_FI PM1_Hcab_DK PM1_Jgtv_DK.1.5 Summer Summer Summer

9 Jagtvej - PM1 all data JGTV_PM1str d PM y =.79x R 2 = d_nox

10 HCAB PM1 all data 4/22-24 HCAB_PM1str d PM y =.958x R 2 = d NOx

11 Road sensor at HCAB Does road surface conditions explain the higher PM1 values?

12

13 d PM /d NOx d PM /d NOx Rain vs PM/NOx ratio - daily averages SurfWater vs PM/NOx ratio - daily averages y = -.269x R 2 =.123 y = -3E-5x R 2 =.124 JgPM JgPM

14 d PM /d NOx Rain vs PM/NOx ratio - daily averages HcPM1 y = -.94x R 2 = d PM /d NOx SurfWater vs PM/NOx ratio - daily averages y = -.1x R 2 =.431 HcPM

15 d PM /d NOx Rain vs PM/NOx ratio - daily averages HcPM25 y =.53x R 2 = d PM /d NOx SurfWater vs PM/NOx ratio - daily averages y = -2E-5x R 2 =.337 HcPM

16 The Swedish Model: description e tot f = e direct f + e re suspension f Processes Emissionsfactor(mg/vkm) Exhaust pipe 2 Tyre and brake wear 1 Road wear 1 Re-suspension (dry conditions) 1-1 Winter e re suspension = f f q * d * e w int er ( dust), PM ref f re suspension ref summer Summer e f * e, ( dust) f = q f PM f q is the source strength for re-suspension which is related to the moister amount of the dust on the street surface d is the amount of dust on the street e ref ( dust) is a reference emission factor f, PM

17 The Swedish Model PM1 Hornsgatan year 2 PM1 street -PM1 roof (µg/m 3 ) Hornsgatan 2 daily mean measurments model average measurments=24.9 average model=24. r=.77 1-Jan 1-Mar 3-Apr 29-Jun 28-Aug 27-Oct 26-Dec Comparison of measured and modelled concentrations of PM1 (µg/m 3 ).

18 Swedish Model for Jagtvej Constant emission factor of 129 mg/(veh km) SMHIs PM emission model, adapted for Danish conditions PM1 model (µg/m 3 ) Number of data points=344 Average measured=7.4 Average modelled=7.1 r=.84 PM1 model (µg/m 3 ) Number of data points=344 Average measured=7.5 Average modelled=6.6 r= PM1 street -PM1 roof (µg/m 3 ) PM1 street -PM1 roof (µg/m 3 ) Comparison of measured and modelled daily mean local concentrations of PM1 at Jagtvej 23. Model results are based on measured local concentrations of NOx.

19 German method for PM1 non-exhaust emission factors (Düring et al., 24) Verkehrssituation Tempolimit [km/h] Anteil Konstant fahrt [%] Stand anteil [%] PM1-Auf/Ab [mg/km] PKW inkl. LNF SV AB> AB_ AB_ AB_ AB_ AB_StGo AO AO AO IO_HVS> Tunnel AB_ Tunnel AB_ Tunnel AB_ Tunnel IO_HVS> HVS HVS HVS LSA HVS LSA LSA IO_Kern IO_NS_dicht Recommended emission factors for good road condition High relevance of HDV (ca. factor 1) No significant effect of traffic speed limit Strong dependence on traffic situation (acceleration/deceleration)

20 Danish method for PM1 non-exhaust emission factors (Berkowicz et al., 25) exhaust part adopted from the UK-TLR method non-exhaust part adopted from EMEP CORINAIR Emission Inventory Guidebook ospm.dmu.dk

21 Comparing the different models Table 1: Simplified version of the proposed German PM 1 emission factors for non exhaust emissions (last two columns) in dependence on the traffic situation. Values for a fleet mix with 4% heavy duty traffic and exhaust emissions according to (UBA, 24) are given. For comparison the emission factors are given for specific streets applying the Danish and Swedish methods. Method/ Traffic situation average Speed Share of constant speed driving[%] exhaust emiss. factor (fleet-mix) [mg/km veh] non-exhaust emission factor (fleet-mix) non exhaust emission factor* [mg/km veh] [mg/km veh] cars / vans trucks [km/h] German method: motorways or outside cities tunnel city main road (HVS1)** city main road (HVS2)** city main road (HVS3)** city main road (HVS4)** city traffic lights (LSA2)** city slow traffic (IO_Kern)** Danish method for JGTV / 7 23 Swedish method for HORG *** * Values for good quality of the road surface, flat terrain and conditions of rain as usual in Germany. ** Speed limit = 5 km/h; ***annual average for the year 2

22 Comparing the different models PM emission factors in mg / (veh km) 25 2 resuspension tyre brakes exhaust SE mod DE mod DK mod

23 Comparing the different models PM emission factors in mg / (veh km) 25 2 PM1 resuspension tyre brakes exhaust 15 1 PM2.5 5 SE PM1 SE mod DE mod DK mod DK rec.an. DK mod DK rec.an. Receptor analysis for Stockholm C. Johansson et al. 25 Receptor analysis for Copenhagen P. Wåhlin et al. 24

24 Measured versus modelled emission factors German emission method PM1 [g/(km veh)] calculated H./Göttinger Str., D (21).3 B./FF-Allee, D (2) K./Jagtvej, DK (21, 23).1.5 St./Hornsgatan, S (2). K./HCAB, DK (23) PM1-emission factor (exh.+non exh.) from measurements [g/(km veh)] exhaust only exhaust plus non exhaust mod. EPA-model x=y

25 Conclusions all modells / measurements show consistently large part of PM emissions from non-exhaust (5%-85%) measures relying on reducing the exhaust part of the vehicle emissions can have a very limited effect on ambient PM levels The Swedish model is able to reproduce the observed strong seasonal variation of non-exhaust emissions. less pronounced annual variation for Danish and German conditions might allow the application of simpler methods based on constant emission factors Road wetnes does not show very high correlation with PM/NOx ratios

26 Open Questions Influence of driving speed / driving pattern? German data show higher emissions for lower speeds and more Stop and Go. Influence of rain? Influence of bad road surface conditions? German model gives a high additive term for bad conditions. What is the difference between JGTV and HCAB??

27 Motivation EU PM1 limit value has been exceeded e.g. Munich discussion about effective reduction measures: road pricing? diesel particle filter? environmental zones?

28 The Swedish Model: emissions 1 d relative units a tyre d anskid f q relative units.2 1-Jan 31-Jan 1-Mar 31-Mar 3-Apr index for rain b 12 1-Jan 31-Jan 1-Mar 31-Mar 3-Apr 3-May e f resuspension ( mg/fkm) 8 4 c 1-Jan 31-Jan 1-Mar 31-Mar 3-Apr 3-May

29 The Swedish Model 16 PM1 Hornsgatan year PM1 (µg/m 3 ) model measuredbackground model yearly mean percentile percentile PM1 (µg/m 3 ) measured-background As the last figure but sorted in size

30 Outlook / Open questions apply the detailed Swedish model for DE / DK influence of road material and road quality?? influence of traffic speed and slope of the road?? influence of rain??

31 Acknowledgement Swedish National Road and Traffic Administration (Vägverket) Stockholm County Council (Landstinget) Swedish EPA SNAP Programme (Swedish National Air Pollution and Health Programme) Danish EPA - Funding for the aerosol measurements Danish Research Council Thanks for your attention!