Influences on Particle Emissions of Modern 2-Stroke Scooters

Transcription

1 Berner Fachhochschule Haute école spécialisée bernoise Gwerdtstrasse 5 CH-2560 Nidau Tel./Tél. +41 (0) Fax +41 (0) Hochschule für Technik und Informatik HTI Haute école Technique et Informatique HTI Abgasprüfstelle (AFHB) Contrôle des gaz d échappement Influences on Particle Emissions of Modern 2-Stroke Scooters 3rd report of the Project A / 2006 Ordered by: Swiss Federal Office of Environment, Bern, CH Bundesamt für Umwelt, BAFU Bern Swiss Petrol Union, Erdöl-Vereinigung, Zürich, CH Swiss Association of Lubricants Industry, VSS-Lubes, Zürich, CH Report: J. Czerwinski, Dipl. Ing. Dr. techn. P. Comte, Dipl. Ing. ETS T. Neubert, Dipl. Ing. FH Y. Zimmerli, Dipl. Ing. HES P. Wili M. Clénin UNIVERSITY OF APPLIED SCIENCES LAB. FOR IC-ENGINES & EXHAUST EMISSION CONTROL Gwerdtstrasse 5, CH-2560 Nidau / Switzerland Network Leading: TTM, Technik Thermische Maschinen, Niederrohrdorf, CH Partnership with: EMPA CH, JRC Ispra It., VTT Fin., SUVA CH Peugeot Motorcycles F, Piaggio Motorcycles It., Engelhard It., PANOLIN AG CH, Motorex CH, Lubrizol GB, Matter Eng. AG CH, TTM CH Nov. / Dec B 193

2 2-S Scooters / Project A3 / AFHB 06 2 CONTENTS 1. SUMMARY 3 2. INTRODUCTION AND OBJECTIVES 4 3. MORPHOLOGY 5 4. PAH ANALYTICS FOR EMISSION PHASE EMISSION FACTORS Repetitions with Peugeot TSDI & Gilera Runner DI Aprilia DI & Aprilia Carb CATALYSTS SCREENING (Emissions phase 5) Procedure, oils, fuel and mufflers Catalyst ageing Results POTENTIALS OF EMISSION REDUCTION (Emissions phase 6) Standardization of lube oils Combinations of technical measures & test procedure Results CONCLUSIONS ACKNOWLEDGMENTS REFERENCES LIST OF FIGURES ANNEX ABBREVIATIONS 21

3 2-S Scooters / Project A3 / AFHB Summary Further research of emissions of modern 2-S scooters was performed in the present work with special focusing on particle mass and nanoparticles. The present report is a continuation of the 1 st report [1] and 2 nd report [2], from which the general informations about the project A and the descriptions of probands and of measuring apparatus are not repeated. The most important topics of this work were : Research on particle morphology (EMPA) Repetition of PAH-analytics (JRC) Repetitions of emission factors Catalysts screening Potentials of combined technical measures. As principal results it can be pointed out : Morphology Particles investigated for TSDI catalyst consist mostly of soluble (liquid) materials, while the particles after catalyst show mostly a soot-like structure. Under electron beam the particles are instable a stochastic partial evaporation reveals, that they are volatile materials. The treatment of sampled aerosol by hot thermodesorber shows similar effects, like treatment by the oxidation catalyst: elimination of liquid fraction, more dry carbonaceous structure. The microstructure of the soot like particles seems to be less graphitized than diesel engine soot particles. The presence of Ca and S in some particles gives clear indication on the contribution of lubrication oil. PAH analytics In general the emissions of PAH correlate well with PM-emissions. In dynamic real world driving there is generally higher PM- and PAH-emission. Dynamic driving implies other composition of PAH (fingerprint), than steady state driving. Due to different PAH-composition the effect on toxicity equivalence TEQ can be strongly modified. WFC modifies the PAH; in particularly it lowers PAH & TEQ at full load. The repeatability of results in dynamic driving cycle is very good. The reproducibility of results, comparing to the previous measuring series, is in part not sufficient. These differences are mainly due to the differences (instability) of emissions. Emission factors The reproducibility of emission results is sometimes poor considerable variations of emissions can occur due to changes of oil & fuel dosing and efficiency of aftertreatment. Vehicle taken randomly from the fleet showed an inactive catalyst. With growing speed the emission of particle mass (PM) increases.

4 2-S Scooters / Project A3 / AFHB 06 4 Catalysts screening Dummy a substrate without washcoat and without catalytic coating shows generally no catalytic conversion with the highest values of CO, HC, PM & DC. The emitted particles are in a bigger size spectrum. Thermally aged catalyst (1000 C, 24h, air) shows generally lower conversions with higher CO, HC, PM & DC values. For some applications (Carburettor with highest t exhaust ) there are no differences between thermally aged and mass production catalyst. The other catalysts, which were supplied specifically for each vehicle show generally similar results like the mass production catalysts. There is a good repeatability of emission results with the catalysts. No conversion (dummy), or weak conversion (thermally aged) are indicated by lower t exhaust after cat. Combinations of technical measures The combinations of technical measures to lower the particle emissions of scooters confirmed the expected effects and showed considerable reduction potentials. These technical measures are: o Higher tier lube oils o Lower oil dosing o Active oxidation catalyst o Supplementary filtration & oxidation devise (WFC). 2. Introduction and objectives Several topics, which were investigated and reported in 2006 are represented in this report: EMPA Laboratory for IC engines performed a research of particle morphology of 2-S scooters by means of Transmission Electron Microscopy (TEM). The results of this work are reported. PAH analytics of PM filters from emission phase 4 ([2] chap. 9), was performed by the EU JRC analytical laboratory with the scope to check the repeatability of results and the influences of stationary and dynamic driving, of different oils and Buck WFC1 on the PAH composition and on TEQ. One of the continuous objectives is collecting data for emission factors including the data of particle emissions, which are still unlimited for scooters. Supplementing the results from [1] (chap. 6) some repetitions tests with both DI-scooters (Peugeot & Piaggio), as well as measurements of two other vehicles (Aprilia) from the market fleet were performed. In the present part of project A, of which the whole matrix is reminded in annex A1, the principal objective was to perform a systematical screening of different catalysts on 4 available vehicles (Peugeot & Piaggio). During these test the effects on emissions during cold start, warm-up and stationary warm operation were studied. Another interesting point was to demonstrate the potentials to reduce the particle emissions by means of different technical measures and their combinations. These measures were: reduced lube oil treatment, higher tier oil, improved exhaust gas aftertreatment and special fuel.

5 2-S Scooters / Project A3 / AFHB Morphology In August 2005 a measuring campaign concerning morphology was performed by EMPA laboratory for IC engines in cooperation with AFHB. EMPA report about this results is given in annex A2. For this research the scooters Peugeot TSDI and Carb. were used und the emissions were analysed at stationary, warm operating conditions (v = 40 km/h). PM-sampling for microscopic research was performed from tail pipe by means of a partial flow dilution tunnel (see [1], Fig. 3). Simultaneously to that SMPS analysis of nanoparticles was repeated with varying temperature of thermodesorber (TD). Due to technical problems with the PM-morphology-samples only the results from Peugeot TSDI are represented in A2. General conclusions of this work (for TSDI) are: Particles investigated before catalyst consist mostly of soluble (liquid) materials, while the particles after catalyst show mostly a soot-like structure. Under electron beam the particles are instable a stochastic partial evaporation reveals, that they are volatile materials. The treatment of sampled aerosol by hot thermodesorber shows similar effects, like treatment by the oxidation catalyst: elimination of liquid fraction, more dry carbonaceous structure. The SMPS results with TD thermal sample treatment confirm the results from [1] (chap. 8) and show much evaporation of the particles as seen from the present images of Transmission Electron Microscopy (TEM). EMPA performed in 2006 further research on this subject using similar scooters: Peugeot TSDI (50cc, 2002) and Kymco Carb. (50cc, 2001) and reported about it in [3] with following conclusions: Particles emitted from 2-stroke scooters were highly volatile. The relative volume loss of the exhaust particles upstream of the catalyst is higher than their downstream counterparts. This is, most likely, due to the removal of highly volatile particles by the catalytic converter. No solid core was detected within the size range of the SMPS No significant difference could be identified between particles sampled from the two different scooters (TSDI, carburettor). Volatile particles observed under cryo-electron microscopy are in agreement with the SMPS measurement. The microstructure of the soot like particles seems to be less graphitized than diesel engine soot particles. The presence of Ca and S in some particles gives clear indication on the contribution of lubrication oil. All those results are perfectly according with the experiences from previous works of the Swiss Scooter Network, [1] and [2]. Only the similarity of particle structure between TSDI-and Carb. versions must be regarded with certain reserve. It has been shown, that the particle emission level of Carb. scooter can vary considerably for different vehicle types according to their mixture tuning, oil dosing exhaust gas temperature and secondary air dosing all conditions, which influence the intensity of particle postoxidation in the exhaust system.

6 2-S Scooters / Project A3 / AFHB 06 6 The modern carburettor scooters in the present project A have approx. 10x lower particle emission (PM) than TSDI, thanks to a very intense postoxidation (SAS & t exh = C). The older-type Carb. scooters in contrary had 2 3x higher PM-emission than TSDI. Also the condition of catalyst (ageing, masking) provokes differences of particle emission even for the same vehicle and the same engine operating point. An example of astonishing differences for Peugeot Carb. is given in Fig. 1, where the influences of carburation, catalyst and SAS are the possible reasons, which can superimpose on each other. In reality the air-fuel-dosing of the usual carburetor is not stable and this is the primary reason, which influences the function of Ox. Cat. & SAS. This question will be further investigated. From all this information can be assumed, that also the morphology of the particles varies according to the intensity of oxidation. Several variants of vehicles, lube oils, oil dosing, fuels, exhaust gas aftertreatment with WFC and different operating conditions offer a large field for the further research of morphology. 4. PAH analytics for emission phase 4 Several PM measuring filters from emission phase 4 ([2], chap. 9) were ananlysed at the EU JRC Ispra analytical laboratory for PAH and TEQ. The filter data are given according to the topic of research in following annexes: repetition of oil screening - [2] annex A13 dynamic driving cycles - [2] annex A16 repetition with Peugeot Carb. and Buck 1 (WFC) present report annex A3 repetitions of emission factors present report annex A6. The results are represented in the JRC report, annex A4. Following conclusions are to be mentioned: In general the emissions of PAH correlate well with PM-emissions. In dynamic real world driving there is generally higher PM- and PAH-emission. Dynamic driving implies other composition of PAH (fingerprint), than steady state driving. Due to different PAH-composition the effect on toxicity equivalence TEQ can be strongly modified. WFC modifies the PAH; in particularly it lowers PAH & TEQ at full load. The repeatability of results in dynamic driving cycle is very good. The reproducibility of results, comparing to the previous measuring series, is in part not sufficient. These differences are mainly due to the differences (instability) of emissions. 5. Emission factors 5.1. Repetitions with Peugeot TSDI & Gilera Runner DI Due to some political discussion a question about PM-emissions of Peugeot TSDI scooter arose from the Italian partners.

7 2-S Scooters / Project A3 / AFHB 06 7 The open question concerned the emission values from 1 st report [1]: for the constant speed 40 km/h PM-values were between 0.07 and 0.13 g/km ([1], Fig. 40-1, 42-1), while for the emission factors in real world driving PM resulted between and g/km ([1], annex A11-1). It was decided to repeat the emission factors tests with both DI scooters Peugeot and Piaggio. The results of these measurements are given in tables annex A5, the lists of PM measuring filters see annex A6. Figures 2 & 3 represent a graphical comparison with the previous results from 2004 (driving cycles see [1]). It can be stated, that there are mostly little differences of CO-emissions, but clearly more HC, and much more PM for both scooters in the repetition measurement. Further results in 2 nd report [2] concerning TSDI confirmed these repetitions with PM-values: 0,056 0,093 g/km at 40 km/h; 0,135 0,193 g/km at full load and 0,095 0,150 g/km in real world cycles. The very low level of PM-emissions for both scooters in the emission factors measurements in [1] must be regarded as a result of conditions, which created higher temperature in the exhaust system and a more intense postoxidation. On the other hand, as signalized by the HCconversion the catalyst can have lost a part of its activity. Considerable variations of emission factors for Peugeot TSDI were already found in [1] (chap. 6.3, Fig. 17) Aprilia DI & Aprilia Carb. These two used vehicles were randomly taken from the fleet. The main data of the Aprilia scooters are given in Table 1; Fig. 4 shows theses scooters and the equipment for NP-analytics in the measuring laboratory. Aprilia (DI) Aprilia (Carb.) vehicle identification SR 50 water ditech SR 50 model year transmission no. of gears variomat variomat engine: type 2 stroke 2 stroke displacement cm number of cylinders 1 1 cooling liquide liquide rated power kw rated speed rpm idling speed rpm max vehicle speed km/h weight empty kg mixture preparation direct injection with automatic carburetor with automatic oil pump oil pump catalyst no yes Table 1: Data of the investigated Aprilia scooters

8 2-S Scooters / Project A3 / AFHB 06 8 Emission factors are represented in Figures 5 & 6, the numeric data see annex A7. Comparing with the emission factors in [1] (Fig. 15 & 16) it can be remarked, that the Aprilia scooters have much higher CO and HC Aprilia DI has no catalyst and leaner tuning, Aprilia Carb. has clearly rich tuning and inactive catalyst. The PM-emissions nevertheless are similar, as for Peugeot and Piaggio, which indicates low oil dosing. The used lube oil is Motorex 2-stroke oil given with the vehicles. There is no further information about this oil. Additionally to the emission factors particle mass and nanoparticles were measured at constant speeds 30 km/h, 40 km/h and full load (data see annex A8). Fig. 7 shows the results for DI with increasing speed and the repeatability at 40 km/h. Fig. 8 shows the same for Aprilia Carb. The measurements were performed first as screening in state of delivery at 40 km/h. After that all the real world cycles for emission factors (A6) and finally the constant speeds in growing sequences were driven. It results: with growing speed the PM-emission increases for both scooter; the growing tendency is confirmed by nanoparticles counts and summary surface of the aerosol, the repetition at 40 km/h at the end of measuring program shows for both scooters lower particle emissions (mass and counts) due to the conditioning of engine and catalyst. Direct comparison of particle emission of both scooters, Fig. 9, shows a lower mass emission and higher count emission of the Carb. variant at all speeds. This is similar like for Peugeot scooters ([2], Fig. 9 & 10). The Carb. variant has richer mixture tuning and stronger oxidation in the catalyst (no SAS) than DI. 6. Catalysts screening (Emission phase 5) 6.1. Procedure, oils, fuel and mufflers The measuring procedure was similar, as in [1] (chap. 9): cold start (20-25 C) acceleration to 40 km/h and v = const = 40 km/h. It was decided to keep this speed to guarantee the catalyst light off with all researched combinations vehicle-catalyst. After v = const period ECE 47 (warm) was performed. During the first 4 min after start the measurement of NP was performed with NanoMet and with CPC (over time) to register the warm-up phase. Since the 9 th minute after start the SMPS (DMA + CPC) was used, performing usually 6 scans. During the ECE47 cycles it was switched again to CPC. The sampling for nanoparticle analysis took place at tailpipe through MD 19 (heated to 120 ), as in the previous works [1], [2]. The represented results are averages of 6 scanns. The temperature and CO after catalyst were time-measured to see the light off. The stationary warm operation was prolonged until 20 min to get enough mass on the measuring filters for the analytics of PAH. For each measuring configuration gravimetry was performed in the cold part (1 st 4 min) and in the hot part (until 20 min) and ECE47.

9 2-S Scooters / Project A3 / AFHB 06 9 The gravimetric measurements of PM were performed at the CVS tunnel (with the same method as for Diesel cars). The matrix of these tests is represented in Fig. 10 (see also annexes A8 & A9): The used oils were ([2], chap. 3.1, Fig. 2): for both Peugeot Scooters Motorex oil nbr. 5, semi-synthetic (150 ppm S) for both Piaggio Scooters Motorex oil nbr. 3, fully-synthetic (310 ppm S) As fuel the standard market gasoline unleaded, 95 RON was used (normdata see [1] p. 31). At the beginning of the project A 2004 a big charge of this gasoline was purchased to perform all research with the same fuel. The investigated mufflers are enumerated in Table 2, which results from the available mufflers [2] (p.5, Tab. 2) by living away the mufflers, which were already investigated in previous research. Table 2: Investigated mufflers for cat. screening Peugeot TSDI Peugeot Carb. Gilera DI Typhoon Carb. 3rd mass prod. cat (comparison same oil) X X X X mass prod. cat (dummy) X X X X mass prod. cat thermally aged X X other cat 1 X X other cat 2 X X Dummy is a substrate without washcoat and without catalytic precious metal. About the thermally aged catalysts see following chap Some repetition measurements were performed: with Gilera DI, dummy, one repetition, with Gilera DI, other / cat., two repetitions, with Typhoon Carb., other cat., two repetitions 6.2. Catalyst ageing For both Piaggio scooters (DI & Carb.) mufflers with thermally aged mass production catalysts were supplied by Engelhard. The ageing of those catalysts ran in an oven at constant temperature 1000 C in air for 24 hours. A thermal ageing implies sintering of the catalytic component, sintering of the washcoat and also interactions (chemical reactions) between carrier- and catalytic species. All these effects reduce the activity of the catalyst, [4], [5].

10 2-S Scooters / Project A3 / AFHB In the real application on the vehicle other effects are superimposed on the properly thermal ageing due to the presence of contaminants in the exhaust gas and due to dynamic changes of gas flow and temperature. These effects are: poisoning (chemical reactions with the catalytic component), fouling or masking (deposits blocking the accessibility of gas to the catalytic component), losses of washcoat and of the catalytic component In previous works of this project some effects of real world aging were demonstrated, [1] chap. 7 and [2] chap. 8. Some important remarks were given from Engelhard concerning the catalyst ageing on 2-S scooters: palladium (Pd) is applied frequently (instead of Platinum (Pt)) for the reason of lower costs, nevertheless Pd ages much quicker with sulphur, synthetic oils contain less S and provoke less ageing of catalyst, than the mineral oils, synthetic oils are more thermostable and can be dosed in lower ratio (1:100), than the mineral oils (1:50), silica from gaskets can provoke layers (masking) on the catalytic surface, irreversible chemical poisoning and masking are more dangerous, than a reversible pollution with substances, which can be burned away at high load operation, in the automotive industry there are high quality requirements and very severe ageing protocols of catalysts Results Cold start & light off Figures 11 & 12 show the time-plots of CO after cat. and exhaust gas temperature after cat for all investigated variants scooter-muffler. Figures 13 & 14 show the respective evaluation of light off time and average CO-values in the succeeding phases of warm-up. Following can be remarked for Peugeot scooters: TSDI has a much leaner mixture tuning, than Carb., this is particularly to see on the CO with dummy, the light off time with TSDI is shorter, the final temperature after 25 min depends on catalytic activity (exothermic heating) for dummy this temperature is always lower, than for active catalyst this ΔT is for TSDI approx. 45 C, but for Carb. ~ 250 C. It has to be reminded, that the mufflers for both Peugeot scooters are geometrically equal, only SAS is used for Carb., [1]. Remarks for Piaggio scooters are: DI (Gilera) is also leaner tuned, than Carb. (Typhoon) but the difference is much less, than for Peugeot (Typhoon Carb. is leaner with lower oil treatment), the light off time with DI & Carb. is roughly the same, the final temperature with dummy is also lower than with original cat. this ΔT is for DI ~ 30 C and for Carb. ~ 60 C, the repeatability of results with other cat is for both scooters very good,

11 2-S Scooters / Project A3 / AFHB the thermally aged catalyst shows lower CO-conversion on DI, but no differences on Carb.; the light off times are the same as for other cats this on both scooters. The exhaust systems and mufflers of both Piaggio scooters are different, [1]. DI is w/o SAS and Carb. with SAS (similarly like Peugeot). 40 km/h warm The tables of numeric data in annex A9 collect all emission values of limited and unlimited (PM, NP) components during cold start, constant speed and ECE 47. Annex A10 gives a list and data of all PM-measuring filters (Pallflex) used in these test series. In following figures the results at constant speed 40 km/h (warm) will be regarded. At steady state operating conditions the scanning of nanoparticles size distribution (PSD) with SMPS is possible. At the end of the figure for each scooter a comparative representation of particle emissions (PM, DC & SMPS) is added. Fig. 15 Peugeot TSDI: dummy shows generally less catalytic conversion with clearly higher values of CO, HC, PM & DC. The PAS-values are still in the zero-sensitivity range of this sensor. Lower SMPS count concentrations together with higher emitted mass PM indicate that there are bigger particle sizes with dummy. Fig. 16 Peugeot Carb. : dummy produces the highest values of CO & HC. In the domain of particles the research catalyst other cat. shows higher values. Even, if the absolute differences are very little and could be in the range of measuring dispersion, the values of PM & DC confirm each other for this single measurement. Obviously the other cat. provokes more sulphate production than the 3 rd mass prod. Cat. SMPS PSD spectra confirm it. Fig. 17 Piaggio Gilera DI: dummy and thermally aged catalyst produce clearly higher CO, HC, PM & DC. PAS is lower (or zero), since the condensates (DC) envelope the solids (PAS). (Due to relatively low temperature level the thermally aged cat. is as bad as dummy.) The repeatability of results with other cat. is very good (increase of NO x in 2 nd repetition is due to temporary deviation to higher speed). Fig. 18 Piaggio Typhoon Carb: dummy shows the usual increase of all emission components (CO, HC, PM, DC), while thermally aged cat. causes no deteriorations for CO & HC, but less efficiency for PM & DC. (Due to high temperature level the thermally aged cat. is quite better, than dummy). The repeatability of results with other cat is also very good. The integrated SMPS particle counts with dummy and with aged cat. are much higher due to more spontaneous condensates. Comparisons: cold, warm, ECE 47 Peugeot Fig. 19 Peugeot TSDI: cold start period produces more CO, HC, & PAS (solids), than the constant speed warm. There is less PM & DC with cold start in spite of inactive catalyst during the light off periods. Fig. 20 Peugeot Carb.: comparing with v = const warm cold start produces here also more CO & HC, but also more PM & DC. The influence of sulfates in sense of increasing PM with other cat. over the level of dummy is not visible in the cold start phase and in ECE 47.

12 2-S Scooters / Project A3 / AFHB Piaggio Fig. 21 Piaggio Gilera DI: cold start provokes higher CO, HC & PAS, lower PM & DC, than v = const. warm Fig. 22 Piaggio Typhoon Carb.: comparing with v = const warm cold start produces here more CO & HC, but also more PM & DC. The influences of dummy are to remark by higher values of CO, HC, PM for all scooters. Comparing with v = constant operation dynamic driving in the ECE 47 cycle increases generally the limited components CO, HC, NO x and PM, except for Piaggio Carb. where CO & HC decrease. Interesting question is: why DI scooters (both Peugeot & Piaggio) have lower PM & DC in the cold start phase, than in the warm operation at v = const and Carburettor scooters inversely have higher PM & DC in cold start phase? This question is important, since the influences on emissions in the cold start phase (4min) are not dominated by the exhaust gas aftertreatment. Most evidently the light off period (1 to 1,5 min) with no conversion in the catalyst produces the remarked effects and differences. In the previous work [1] & [2] the principal differences between the two concepts (DI & Carb.) were explained: DI-leaner tuning, lower t exhaust, no SAS in opposite to Carb. Also the influences of different oils and fuels on PM-formation were explained by different composition and different behaviour of condensation of the heavy hydrocarbons. For DI the lube oil, which is injected in the intake air gets in contact with the fuel, which is insufflated in the combustion chamber shortly before the ignition (as leaner total mixture). For Carb. the contact oil-fuel is earlier, there is more time for mixing and prereactions and there is higher surplus of fuel. With these considerations it became clear, that the two concepts (DI & Carb.) offer different conditions for combustion of lube oil and finally for the chemistry of resulting hydrocarbons. Additionally to that another effect caused by the deposition (storage) of lube oil in the cold engine carkcase shall be mentioned. During the first period after cold start this deposition is more, or less present and it is surely lower with vehicle type with carburettor. Consquently there is a higher deposition and a lower instantaneous oil dosing arriving in the combustion chamber for DI during the cold start and warm-up. Fig. 23 summarizes again the PM-values of all measured variants at 40 km/h warm and in ECE 47. The effects of dummy and of the lower oxidation intensity with the thermally aged catalyst are well visible. Further figures show the time-plots of the NP-measuring signals in the ECE 47 driving cycle: Fig. 24 & 25 Peugeot scooters: some changes over the time are remarkable. Particularly the reduction of PAS-peak values in Fig. 25 results from growing temperature and growing DC-values during the test. The increased condensates envelope more solids and let disappear a part of PAS.

13 2-S Scooters / Project A3 / AFHB Fig. 26 & 27 Piaggio scooters: particularities of plots for dummy and for thermally aged catalyst are clearly to distinguish. For Typhoon Carb. (Fig. 27-2) there are repeatable effects (DC & CPC) resulting from the growing temperature during the test procedure. 7. Potentials of emission reduction (Emissions phase 6) 7.1. Standardization of 2-S lube oils Similarly, like for other engine applications there are also standard testing procedures of lube oils for 2-stroke gasoline engines with lost-oil lubrication. The different standards are: API classes American Petroleum Institute (API) defined together with CEC following standard classes for 2-S lube oils: API-TA (TSC-1) for light motorcycles (mopeds) API-TB (TSC-2) for motorcycles and scooters API-TC (TSC-3) for high power engines API-TD (TSC-4) for outboard marine engines, according to NMMA TC-W2 The classes API-TA and API-TB are no more in used up to date. Class API-TC (CEC TSC-3) is valid with: test engines: Yamaha CEC 505 and Yamaha 350 M2 test criteria: seize up, autoignition, hindering of piston sticking and piston fouling, protection against lacquer coating on the piston. JASO- and ISO (global) specifications, [6] The Japanese Automotive Standard Organisation established for the Asian 2-S markets following lube oil classes: JASO FA for light performances JASO FB for middle performances JASO FC for middle performances, particularly smoke-free. Since these JASO-classes were not sufficient for European markets, a new classification together with ISO was issued: ISO EGB (=JASO FB) ISO EGC (=JASO FC) ISO EGD (with higher requirements of general engine cleanliness than JASO FC). The test criteria for JASO/ISO are: test engines: Honda DIO AF27 and Suzuki SX800R validation criteria: engine cleanliness, lubricity, fouling of exhaust slots & channels, smoke (especially for ISO EGC & ISO EGD).

14 2-S Scooters / Project A3 / AFHB There are works about further development of test methods. In [7] Yamaha presents three supplementary evaluation methods for lube oils, which are not included in the JASO standards. These methods are: low-temperature detergency, low-temperature start-ability, exhaust port blocking. No use of polybutene with molecular weight bigger than 1000 is recommended. NMMA TC-W3 Natinal Marine Manufacturers Association (US) established this norm for the 2-S outboard engines in USA. Beside the usual test requirements (see above) additional laboratory validation tests of corrosion protection and abilities of mixing and filtering are necessary. In this application segment the engines work continuously at rated power over long time periods, which requres the highest oil performances, [8]. TISI 1040 Thailand Industrial Standards Institute created the TISI 1040 classification of 2-S lube oils for the Thailand market: test engines: Kawasaki KH 125 M for the engine cleanliness requirements, Suzuki SX800R for exhaust smoke. Lube oils investigated in project A Table 3 represents all lube oils, which were investigated in project A, [1], [2] and shows their relationship to the international standards. Lube oil Panolin Synth. Panolin Aqua-Synth. Panolin TS Nycolube 248 Selenia Hi-Scooter 2Tech Motorex N 1 Motorex N 2 Motorex N 3 Motorex N 4 Motorex N 5 API ISO JASO NMMA Ref. Proj. A TC EGD FC TC-W3 TC EGC FC [1], p. 22 TC EGD FC TC++ [1], A10 EGD FC EGD FC EGD FC [2], Fig.2 EGD FC EGD FC Table 3: 2-S lube oils investigated in Project A and their belonging to the international standards.

15 2-S Scooters / Project A3 / AFHB A clear and repeatable result of worldwide investigations treating with the emissions of 2-S little gasoline engines is, that the lowering of oil dosing rate is the most desirable objective to limit the emissions. It is necessary to minimize the oil treatment and to approach as near as possible the durability limit of the engine. In this aspect the recommendations of the engine manufacturer have to consider the whole scale of oil qualities present on the market and as consequence the recommended oil dosing rate has a security margin for the lower tier oils. (It can be assumed, that the majority of users in this segment prefers the low cost solutions). There is a big variance of oil qualities on the market and to attain the objective of minimization of emission it is worth to strive for better unification and narrowing of the variance of oil quality parameters. The results of project A give incitation to this worthwhile goal, which is interesting for all market actors Combinations of technical measures & test procedure The possibilities of emissions improvements were combined in steps on two vehicles: Peugeot TSDI and Peugeot Carb. to demonstrate the potentials of reductions. The investigated variants were: no catalyst (dummy), lower tier oil, overdosing, gasoline no catalyst (dummy), lower tier oil, 100% dosing, gasoline no catalyst (dummy), higher tier oil, 100% dosing, gasoline no catalyst (dummy), higher tier oil, 50% dosing, gasoline ox. catalyst (series), higher tier oil, 50% dosing, gasoline ox. cat. + WFC, higher tier oil, 50% dosing, gasoline ox. cat. + WFC, higher tier oil, 50% dosing, Aspen The lower tier oil (worst) was supplied from the army stocks, the data from analysis of this oil see annex A11. This is mineral oil with an older formulation, with additive package from Lubrizol. There is a high content of sulphur (1,25% m) and also of phosphorus, calcium and zinc (compare A 11-1 and [2], Fig. 2). The used higher tier oils (best) are the semi-synthetic oil Motorex nbr. 5 for TSDI and fullysynthetic oil Motorex nbr. 4 for Carb. (see [2], Fig. 2). The tests of comparison of different technical measures were called simply best / worse case and they were performed with warm engine at constant speed 40 km/h and in the driving cycle ECE 47. The lube oil dosing of Peugeot scooters is done by means of electronic injection in the intake air. The overdosing was simulated by mixing some additional oil to the fuel. To provoke a lower oil treatment, than the original, the lube oil in the oil recipient was diluted with fuel. The wiremesh used in this test was coated by Engelhard with precious metal loading 50 g/ft 3, 1/0/0, and wash coat type OEX-546. The WFC was new and had dimensions D60 x L49 mm.

16 2-S Scooters / Project A3 / AFHB Results The results of Peugeot TSDI are represented in Figures 28, 29 & 30 and the numeric data are given in annex A12 & A13. The influence of oil quality alone is to see between the variants 1 and 4. There is a clear reduction of PM with the oil best. Further reduction of oil dosing, application of ox.cat. and ox.cat + WFC bring the expected reduction of PM. The potential of Aspen fuel, elimination of aromates, is not visible in the global PM-emissions. Measurements with WFC show a higher temperature level at tail pipe due to intensified exothermic oxidation. The repeatability check at constant speed (variant 7 & 8) shows a considerable dispersion of PM-results. The results of Peugeot Carb. are represented in Figures 31, 32 & 33 and the numeric data are given in annex A14 & A15. The influence of oil quality alone is to see between the variants 1 and 3 with an PMadvantage of the oil best. With the introduction of the further measures also the expected emission reductions are attained. The exothermic heating (higher t exhaust ) is strongly pronounced already with the oxidation catalyst alone. The repeatability check at constant speed (variant 5 & 6) shows a considerable dispersion of PM-results. 8. Conclusions Following principal conclusions are to be mentioned: Morphology Particles investigated for TSDI catalyst consist mostly of soluble (liquid) materials, while the particles after catalyst show mostly a soot-like structure. Under electron beam the particles are instable a stochastic partial evaporation reveals, that they are volatile materials. The treatment of sampled aerosol by hot thermodesorber shows similar effects, like treatment by the oxidation catalyst: elimination of liquid fraction, more dry carbonaceous structure. The microstructure of the soot like particles seems to be less graphitized than diesel engine soot particles. The presence of Ca and S in some particles gives clear indication on the contribution of lubrication oil.

17 2-S Scooters / Project A3 / AFHB PAH analytics In general the emissions of PAH correlate well with PM-emissions. In dynamic real world driving there is generally higher PM- and PAH-emission. Dynamic driving implies other composition of PAH (fingerprint), than steady state driving. Due to different PAH-composition the effect on toxicity equivalence TEQ can be strongly modified. WFC modifies the PAH; in particularly it lowers PAH & TEQ at full load. The repeatability of results in dynamic driving cycle is very good. The reproducibility of results, comparing to the previous measuring series, is in part not sufficient. These differences are mainly due to the differences (instability) of emissions. Emission factors The reproducibility of emission results is sometimes poor considerable variations of emissions can occur due to changes of oil & fuel dosing and efficiency of aftertreatment. Vehicle taken randomly from the fleet showed an inactive catalyst. With growing speed the emission of particle mass (PM) increases. Catalysts screening Dummy a substrate without washcoat and without catalytic coating shows generally no catalytic conversion with the highest values of CO, HC, PM & DC. The emitted particles are in a bigger size spectrum. Thermally aged catalyst (1000 C, 24h, air) shows generally lower conversions with higher CO, HC, PM & DC values. For some applications (Carburettor with highest t exhaust ) there are no differences between thermally aged and mass production catalyst. The other catalysts, which were supplied specifically for each vehicle show generally similar results like the mass production catalysts. There is a good repeatability of emission results with the catalysts. No conversion (dummy), or weak conversion (thermally aged) are indicated by lower t exhaust after cat. Combinations of technical measures The combinations of technical measures to lower the particle emissions of scooters confirmed the expected effects and showed considerable reduction potentials. These technical measures are: o Higher tier lube oils o Lower oil dosing o Active oxidation catalyst o Supplementary filtration & oxidation devise (WFC). In scope of minimizing the particle emissions the general oil dosing in the fleet should be as much as possible reduced. To better attain this goal a stricter standardization and a very narrow variance of the oil quality parameters on the market will be necessary. To guarantee the satisfactory efficiency of oxidation catalysts in the 2-wheeler application the introduction of production conformity control and periodical emission check is urgently necessary. There are several materials, which can help to better filtrate and oxidize the exhaust gas components. Promoting of further R&D, as well as introduction to the market through the appropriate regulations or incentives are necessary.

18 2-S Scooters / Project A3 / AFHB Acknowledgements The authors would like to express their gratitude for the support and realisation of the project to: BAFU (Swiss EPA), Mr. F. Reutimann, Mr. D. Zürcher Swiss Petrol Union, Erdöl-Vereiningung, Dr. A. Heitzer Swiss Association of Lubricants, VSS-Lubes, Dr. J. Fiala For the information about TSDI-technology and help with the test material thanks to: Peugeot Motorcycles France, Mr. M. Bonnin, Mr. G. Althoffer, Mr. M. Soler Piaggio Motorcycles Italy, Mr. D. Cundari, Mr. L. Zappala For informations and contribution of lube oils thanks to: Bucher AG, Motorex, CH, Mr. O. Sedello, Mr. M. Kurzwart PANOLIN AG, CH, Mr. P. Lämmle, Mr. R. Fanelli Lubrizol Ltd, GB, Mrs. M.-C. Soobramanien For informations and support concerning catalysts thanks to: Engelhard Srl, Italy, Mrs. N. Violetti, Mr. P. Landri, Mr. G. Rickert (D) For support of the nanoparticle analytics to: Matter Engineering AG, CH, Dr. M. Kasper, Mr. Th. Mosimann For support of analytics of substances thanks to: EMPA Analytical Laboratory, CH, Dr. M. Mohr, Mrs. Dr. A. Ulrich SUVA Analytical Laboratory, CH, Dr. R. Wolf, Mrs. S. Waller EU-JRC Analytical Laboratory, It, Dr. B. Larsen, Mr. G. Martini, Mrs. Dr. M. Astorga-Llorens, Mrs. M. Rey For leadership of the project network special thanks to: TTM, CH, Mr. A. Mayer 10. References [1] Influences on Particle Emissions of Modern 2-Stroke Scooters. 1 st report of the project A/2004. Lab. for Emission Control, Univ. of Appl. Sc. Biel, CH ; Matter Engineering ; B150 Nov. 04 / May 05. [2] Influences on Particle Emissions of Modern 2-Stroke Scooters. 2 nd report of the project A/2005. Lab. for Emission Control, Univ. of Appl. Sc. Biel, CH ; B163, June/Oct [3] Etissa, D. ; Schreiber, D. ; Mohr, M. : Characterization of particles emitted from modern 2-stroke scooters by electron microscopy and tandem DMA. EMPA Federal Laboratories, 10 th ETH-Conference on Combustion Generated Nanoparticles; Zurich, Aug [4] Heck, R.M.; Farrauto, R.J.; Gulati, S.T.: Catalytic Air Pollution Control: commercial Technology. John Wiley & Sons, Inc. New York, 2002, 2 nd edition, ISBN [5] Russ, G.; Flierl, R.; Kairies, D.: Katalysatorvergiftung von Ottomotoren durch Motorenöl. MTZ 6/2005, Jahrgang 66, S. 496 [6] / jaso web

19 2-S Scooters / Project A3 / AFHB [7] Kawabe, H.; Takahashi, K.: Evaluation Technique of Two-Stroke Engine Oil for Motorcycles. YAMAHA, Japan: SAE Paper , SETC, Nov , 2006 [8] Timar, J.; Nygaard, W. C.; Shrout, M. A.; Castile, K. S.: Development and Testing of Optimized Engine Oils for Modern Two-Stroke Cycle Direct Fuel Injected Outboard Engines. Chevron; Southwest Research Institute; SAE Paper , SETC, Nov , List of figures Fig. 1 Variations of particle emissions of Peugeot Looxor Carb. in different measuring periods 2005 Emission factors Fig. 2 Emission factors for Peugeot TSDI comparison Fig. 3 Emission factors for Piaggio Gilera DI comparison Fig. 4 Fig. 5 Fig. 6 Fig. 7 Fig. 8 Fig. 9 Aprilia DI & Aprilia Carb. in the measuring laboratory Emission factors for Aprilia DI Emissions factors for Aprilia Carb. Particle emissions Aprilia DI different speeds, repeatability Particle emissions Aprilia Carb. different speeds, repeatability Comparisons of particle emissions Aprilia DI Aprilia Carb. Cat. Screening Fig. 10 Procedures of catalysts screening emission phase 5 Fig. 11 CO & Temp. after cat., cold start, Peugeot TSDI & Carb. Fig. 12 CO & Temp. after cat., cold start, Piaggio DI & Carb. Fig. 13 Catalyst light off & CO, cold start, Peugeot TSDI & Carb. Fig. 14 Catalyst light off & CO, cold start, Piaggio TSDI & Carb. Fig. 15 Limited & unlimited emissions, 40 km/h warm, Peugeot TSDI Fig. 16 Limited & unlimited emissions, 40 km/h warm, Peugeot Carb. Fig. 17 Limited & unlimited emissions, 40 km/h warm, Piaggio Gilera DI Fig. 18 Limited & unlimited emissions, 40 km/h warm, Piaggio Carb. Fig. 19 Limited & unlimited emissions; cold start, 40 km/h, ECE 47, Peugeot TSDI Fig. 20 Limited & unlimited emissions; cold start, 40 km/h, ECE 47, Peugeot Carb. Fig. 21 Limited & unlimited emissions; cold start, 40 km/h, ECE 47, Piaggio Gilera DI Fig. 22 Limited & unlimited emissions; cold start, 40 km/h, ECE 47, Piaggio Carb. Fig. 23 Comparison of all PM-values, 40 km/h & ECE 47

20 2-S Scooters / Project A3 / AFHB Fig. 24 Time plots of PAS, DC & CPC in ECE 47, Peugeot TSDI Fig. 25 Time plots of PAS, DC & CPC in ECE 47, Peugeot Carb. Fig. 26 Time plots of PAS, DC & CPC in ECE 47, Piaggio Gilera DI Fig. 27 Time plots of PAS, DC & CPC in ECE 47, Piaggio Carb. Best / worse case Fig. 28 Limited & unlimited emissions, best / worst case, Peugeot TSDI Fig. 29 Comparison of all PM-values, best / worst case, Peugeot TSDI Fig. 30 Exhaust gas temperatures, best / worst case, Peugeot TSDI Fig. 31 Limited & unlimited emissions, best / worst case, Peugeot Carb. Fig. 32 Comparison of all PM-values, best / worst case, Peugeot Carb. Fig. 33 Exhaust gas temperatures, best / worst case, Peugeot Carb. 12. Annexes A 1 Test Matrix Project A, June A 2 EMPA report: Morphology PM 2-S Scooters Biel, Aug A 3 List of measuring filters (P) Peugeot Carb. & Buck1 (phase 4) A 4 JRC PAH analysis for emission phase 4 ([2], chap. 9) A 5 A 6 A 7 A 8 A 9 A 10 Emission factors for Peugeot TSDI & Gilera DI numeric values List of measuring filters (P) emission factors TSDI / DI Emission factors for Aprilia DI & Aprilia Carb. numeric values Limited and unlimited emissions with Aprilia Scooters at constant speeds All emission values in phase 5, cat screening: tables of data List of PM-measuring filters (P) in phase 5, cat. screening A 11 A 12 A 13 A 14 A 15 Data of the lower tier lube oil All emission values in phase 6, best / worse case, Peugeot TSDI List of PM-measuring filters (P) in phase 6, best / worse case, Peugeot TSDI All emission values in phase 6, best / worse case, Peugeot Carb. List of PM-measuring filters (P) in phase 6, best / worse case, Peugeot Carb.

21 2-S Scooters / Project A3 / AFHB Abbreviations AFHB AMF API ANCMA BAFU BAT BfE C Carb CEC CMD CPC CVS DC DEA DF DI DMA DPF EC ECU EGR EMPA EPA ETHZ EV FL GRPE IEA INSOF ISO JASO JRC ME NanoMet NMMA NMOG NP OC OP P PAH Pan PAS PC PM PMP PN PSD Q SAS SMPS SOF Abgasprüfstelle der Fachhochschule, Biel CH (Lab. For Exhaust Gas Control, Univ. of Appl. Sciences, Biel-Bienne, CH) Advanced Motor Fuels American Petroleum Institute Associazione Nazionale Ciclo Motociclo Accessori, Milano, It. Bundesamt für Umwelt (Swiss EPA) best available technology Bundesmat für Energie, CH Carburetor Carburetor Commission of European Community count median diameter condensation particle counter constant volume sampling diffusion charging sensor Deutsche Erdöl AG experimental oil w/o any additive packages dilution factor direction injection differential mobility analyser Diesel Particle Filter elemental carbon electronic control unit exhaust gas recirculation Eidgenössische Materialprüfungs- und Forschungsanstalt Environmental Protection Agency Eidgenössische Technische Hochschule Zürich Erdöl Vereinigung, CH full load Groupe Rapporteur Pollution et Energie International Energy Agency insoluble fraction International Standard Organization Japanese Automotive Standard Organization Joint Research Center, Ispra It. Matter Engineering, CH minidiluter + PAS + DC (ev. + TC, or TD) US National Marine Manufactures Association non methan organic gases nanoparticulates organic carbon ozon potential Pallflex polycyclic aromatic hydrocarbons Panolin (Swiss lube oil manufacturer) photoelectric aerosol sensor particles counts particulate matter, particulate mass particle measuring program particles number particles size distribution Quartz secondary air system scanning mobility particles sizer soluble organic fractions

22 2-S Scooters / Project A3 / AFHB SUVA T TC TD TEQ ThC TISI TSDI TP TPN TTM VOC VOF VTT WFC WMTC Schweizerische Unfall Versicherungs Anstalt (Swiss Occupational Insurance) TSDI thermoconditioner, total carbon thermodesorber Toxicity Equivalence Factor thermoconditioner Thailand Industrial Standards Institute Two Stroke Direct Injection tailpipe total particle number Technik Thermische Maschinen, Niederrohrdorf, CH volatile organic compounds volatile organic fraction Technical Research Center of Finland wiremesh filter catalyst Worldwide Motorcycle Test Cycle