(Particle) Emissions of Small 2-& 4-Stroke Scooters with (Hydrous) Ethanol Blends

Transcription

1 (Particle) Emissions of Small 2-& 4-Stroke Scooters with (Hydrous) Ethanol Blends Copyright 2010 SAE International Czerwinski, J.; Comte, P. AFHB, University of Applied Sciences, Biel, CH Reutimann, F. BAFU, Federal Office of Environment, CH Makkee, M. Catalysis Engineering, Faculty Applied Sciences, Technical University Delft, NL ABSTRACT The objectives of the present work are to investigate the regulated and unregulated (particle) emissions of a classical and modern 2-stroke and a typical 4-stroke scooter with different ethanol blend fuels. There is also comparison of two different ethanol fuels: pure ethanol (E) *) and hydrous ethanol (EH) which contains 3.9% water and is denatured with 1.5% oline. Special attention is paid in this research to the hydrous ethanol, since the production costs of hydrous ethanol are much less than those for (dry) ethanol. The vehicles are with carburettor and without catalyst, which represents the most frequent technology in Eastern Asia and offers the information of engine-out emissions. Exhaust emissions measurements have been performed with fuels containing ethanol (E), or hydrous ethanol (EH) in the portion of 5, 10, 15 and 20% by volume. During the test systematical analysis of particle mass (PM) and nano-particles counts (NP) were carried out. The most important results are: there are no significant differences of results between the blends with pure ethanol (E), or hydrous ethanol (EH), except of some cases, where EH improves slightly the emissions (CO, HC, PM, NP) and reduces the fuel consumption, addition of ethanol to the oline provokes a leaner tuning of the engine operation, for the investigated newer 2-S scooter with leaner tuning the irregularities of combustion and increased emissions of PM & NP were remarkable with higher ethanol content, there was a poor driveability, the older 2-S scooter showed good performances and reduction of CO and of fuel consumption up to E15; no impact on or reduction of (nano-) particles and reduction of particle mass emissions with growing ethanol content, the operation of 4-S scooter was without problems, the leaning by ethanol caused: lowering of CO, HC & fuel consumption, increase of NOx, no effect on PM and reduction of nanoparticles count concentrations especially at transient operation, there are no significant differences of results between the blends with pure ethanol (E), or hydrous ethanol (EH). The present investigations did not concern the durability of parts exposed to the chemical influences of ethanol. Also the cold start, particularly in extreme conditions and the lube oil dilution were not addressed. *) Abbreviations & References see at the end of paper - 1 -

2 INTRODUCTION The growth to sustainability of the transportation fuels is highly stimulated by the EU directives and blending of 5.75 % fuel derived from biomass should be achieved in 2010 and the legislation is aiming for 20 % into the year Ethanol produced by fermentation of corn (the first generation) and biomass (the second generation) and methyl esters of fats and oils (the first/second generation bio diesel) are the renewable fuel. Several countries have objectives to substitute a part of the energy of traffic by ethanol as the renewable energy source. Laboratories for IC-Engines and Exhaust Emission Control (AFHB) of the University of Applied Sciences, Biel, Switzerland are involved since 2000 in several research projects about emission factors and possibilities of reduction of (nano)particle emissions of 2- wheelers. A special attention was paid to the 2-stroke scooters, which have much higher particle emission, than the 4-strokers. In an international network project, treating about analytics and possibilities of reduction of (nano)particle emissions from 2-S scooters, several topics were investigated, [1, 2, 3, 4, 5, 6] and the combinations of technical measures to lower the particle emissions of scooters confirmed the expected effects and showed considerable reduction potentials. These technical measures were: o Higher tier lube oils o Lower oil dosing o Active oxidation catalyst o Supplementary filtration & oxidation device (wire-mesh filter-catalyst WFC) o Special fuel. The special fuel used in those tests was Alkylat Aspen oline with a uniform HC-matrix (mostly isooctane) and no aromatics. This special oline is recommended in Switzerland for the hand-held machines in professional application. The idea of using ethanol blends was known, but not applied before in the research of scooters. - adaptation of engine construction in regard to a changed thermal stress of combustion chamber parts, - adaptation of spark plugs and injectors, - fuel injection system, - wear of valves, pistons, rings and liners, - polymer materials and sealings, - crankcase ventilation and oil dilution, - software of engine ECU, new or flexible parameter settings. Small portions of ethanol E5 are generally accepted for the vehicle fleets without any adaptations. Very useful information about handling of olineethanol blends up to 10% v/v is given in the CONCAWE report No. 3/08, [12]. The objectives of the present work are to investigate the limited and the unregulated emissions of typical 2-stroke and 4-stroke scooters 50 cc with different ethanol blend fuels, with pure and hydrous ethanol. The vehicles are with carburettor and without catalyst, which represents the most frequent technology in Eastern Asia and offers the information of engine-out emissions. Ethanol fuel specifications worldwide traditionally dictate use of anhydrous ethanol (less than 1% water) for oline blending. This result in additional costs, energy usage and environmental impacts associated with the extra processing step required to dehydrate the hydrous ethanol produced via distillation (4-5% water) to meet the current anhydrous ethanol specifications. For saving costs and energy the hydrous ethanol may be effectively used in most ethanol/oline blending applications, reducing or eliminating the need for anhydrous ethanol production and distribution, [13], [14]. The hydrous ethanol is predisposed to be a subject to standardisation and acceptance by major stakeholders in the fuel and automotive industries. During the tests a systematical analysis of particle mass (PM) and nanoparticles counts (NP) was performed. INVESTIGATED SCOOTERS Ethanol is used for passenger cars for a long time (Brazil). In the last years, due to the increasing prices of crude oil, there is a growing interest for ethanol. Several countries have objectives to substitute a part of the energy of traffic by the renewable energy. On the other hand there are interferences with the prices of food in certain regions. Some manufacturers offer FFV (flex fuel vehicles), which is particularly challenging for high ethanol content (E85) in countries, like Sweden with colder climatic conditions. There are several technical problems to resolve to guarantee the long live operation of the engine with E85, [7, 8, 9, 10, 11]: Fig. 1a: Investigated 2-S Scooter: Piaggio Typhoon 2-S 50 cc

3 The research of emissions with alcohol was performed on the Piaggio Typhoon 2-stroke and with an older type 2-stroke Kreidler Florett small motorcycle (lube oil directly mixed with the fuel). The investigated 4-S Scooter was Honda Zoomer with carburettor and without catalyst, Fig. 2. The most important data of this vehicle represents Table 2. Fig. 1 shows both vehicles in the laboratory and Table 1 represents the most important data. Fig. 1b: Investigated 2-S scooters :Kreilder Florett 2-S 50cc Piaggio Kreidler vehicle identification Typhoon K54/511 Florett RS model year transmission no. variomat m5 of gears km at beginning 670 km 8316 km engine: type displacement cm 3 2 stroke 50 2 stroke 50 number of cylinders 1 1 cooling Air cooled Air cooled rated power kw rated speed rpm idling speed rpm max vehicle speed km/h weight empty kg mixture preparation carburettor with automatic oil pump carburettor blend 2% lube oil in oline SAS (secondary yes no air system) catalyst dummy no Table 1: Data of the investigated 2-S scooters. Fig. 2: Investigated 4-S scooter: Honda Zoomer 4-S 50 cc Honda vehicle identification Zoomer NPS 50 model year 2004 transmission no. of variomat gears km at beginning 74 km engine: type displacement cm 3 4 stroke 50 number of cylinders 1 cooling rated power kw rated speed rpm idling speed rpm max vehicle speed km/h weight empty kg mixture preparation SAS (secondary air system) catalyst Water cooled carburettor yes dummy Table2: Date of the investigated 4-S scooter All vehicles use simple conventional carburettors with a cable-controlled throttle body and needle

4 Fuels As a basic fuel a standard oline, lead-free, RON 95, Swiss market quality was used. At the beginning of network projects about the particle emissions of 2-S scooters 2004 a large batch of this oline was purchased to perform all research with the same fuel. The sulphur content of this oline was analysed and no sulphur was found (detection limit < 2 ppm). The investigated fuel blends contained ethanol (E), or hydrous ethanol (EH) in the portions of 5, 10, 15 and 20% by volume. Pure ethanol is C 2 H 5 OH and the hydrous ethanol contains: 94.56% vol ethanol, 3.94% vol water and 1.5% oline. The most important parameters of the fuels used are summarized in the Table 3. It can be seen, that with increasing the ethanol ratio the stoichiometric air requirement of the blendfuel decreases. That means by an approximately equal air flow rate the air-fuel-mixture will be leaner. Furthermore there is less heat value in ethanol. The boiling point at a fix temperature and the high latent heat of evaporation of ethanol may cause serious problems of cold starting. Gasoline Ethanol E15 E20 EH10 EH15 EH20 C 2 H 5 OH density [g/cm 3 ] stoichiometric air / fuel ratio lower calorific value Table 3: Parameters of fuels used. *) according to [9] Lube oils [-] [MJ/kg] *) boiling point [ C] *) Research Octane Nbr. *) latent heat of evaporation [-] [kj/kg] The lube oils used were according to the requirements of vehicle manufacturers: for Piaggio Typhoon fully synthetic Selenia HI- Scooter 2 TECH oil, product code: 1050, Piaggio 2966, API TC ++ for Kleider Florett fully synthetic oil Motorex Nbr. 4, [15]. For Honda Zoomer Motorex 10W-40, API SG. MEASURING APPARATUS & PROCEDURES CHASSIS DYNAMOMETER roller dynamometer: Schenk 500 G5 60 driver conductor system: Zöllner FLG, 2Typ. RP d, Progr., Version 1.4 CVS dilution system: Horiba CVS 9500T with Roots blower air conditioning in the measuring hall (intake and dilution air) automatic, temperature: o C, humidity: g/kg The measuring set-up on a chassis dynamometer is represented in Fig. 3. TEST EQUIPMENT FOR REGULATED EXHAUST GAS EMISSIONS This equipment fulfills the requirements of the Swiss and European exhaust legislation 70/220/EEC /76/EC; 97/24/EC - chap /51/EC. eous components: exhaust measuring system Horiba MEXA- 9400H CO, CO 2 infrared analyzers (IR) HC IR... only for idling HC FID... flame ionization detector for total hydrocarbons NO/NO X... chemoluminescence analyzer (CLA) O 2... Magnos (paramagnetic analyzer) The dilution ratio DF in the CVS-dilution tunnel is variable and can be controlled by means of the CO 2 - analysis. measurement of the particulate mass (PM): sampling from the full-flow dilution tunnel filter temperature 52 o C conditioning of filter: 8-24 h (20 o C, rel. humidity 50%) scale: Mettler, accuracy ± 1 μg PARTICLE SIZE ANALYSIS In addition to the gravimetric measurement of particulate mass, the particle size and counts distributions were analyzed with following apparatus: SMPS Scanning Mobility Particle Sizer, TSI (DMA TSI 3081L, CPC TSI 3772) NanoMet System consisting of: - PAS Photoelectric Aerosol Sensor (Eco Chem PAS 2000) - DC Diffusion Charging Sensor (Matter Eng. LQ1-DC) - MD19 tunable minidiluter (Matter Eng. MD19-2E, see Fig. 3)

5 for constant speed for driving cycles The stationary warm operation was prolonged until 20 min to get enough mass on the measuring filters for the analysis of PM. The driving resistances of the test bench were set according to the Swiss exhaust legislation for motorcycles. (reference mass 170kg; a/b/c= 10/0/245; absorbed power RESEARCH PROCEDURES The basic investigations with ethanol-based fuels were performed with each variant of fuel according to the same procedure: 5 min conditioning at full load legal test cycle - for Piaggio Typhoon limited to 45 km/h ECE 47, 1) - for Kreidler and Honda unlimited speed ECE 40, 2) constant speed 40 km/h - first 5 min conditioning - further 10 min measurements of PSD s with SMPS - last 3 min last scan SMPS. The legal driving cycles are represented in Fig. 4. Fig. 3: Sampling and measuring set-up for nanoparticulate analysis of scooters. SMPS enables to count nanoparticles according to their size distribution (10-400nm). PAS (photoelectric aerosol sensor) is sensitive to the surface of particulates and to the chemical properties of the surface. It indicates the solid particles with carbonaceous surface. This type of particles is rarely measurable in the 2-S exhaust and the results of PAS are not represented in this paper. DC (diffusion charging sensor) measures the total particle surface independent of the chemical properties. It indicates the solids and the condensates. The results of DC correspond very well with the measured particle mass PM. The sampling and measuring set-up during the tests is shown in Fig. 3. The sampling position for NP-measurements is at tailpipe (TP) for basic investigations at constant speed and for driving cycles the sampling position from CVS (according to PMP) is used. The minidiluter MD19 is heated to 120 C. All NP-measurements in this work were performed without the thermoconditioner (TC). speed [km/h] speed [km/h] Ph 1 cold ECE time [s] Ph 1 cold ECE 40 Fig. 4: Legal driving cycles used for the investigated vehicles. 1) according to 97/24/EC chap. 5 annex 1 2) according to 97/24/EC chap. 5 annex 2 Ph 2 hot Piaggio Typhoon 2-S Ph 2 hot Honda Zoomer 4-S & Kreidler Florett 2-S time [s] - 5 -

6 RESULTS Piaggio Typhoon The first test were performed with original tuning of the air-fuel ratio and without catalyst. Before and after tests the mixture tuning was controlled at idling (without SAS). There were following values: before tests CO idl. 2.0 % n = 1770 rpm after tests CO idl. 2.2 % n = 1770 rpm - leaning of mixture by increasing ethanol portion, - lowering the combustion peak temperatures and NOx formation with increasing ethanol portion, - retarded combustion, increased cyclic irregularities of combustion and increasing HC-emissions with ethanol and with growing ethanol share. The hydrous ethanol increases further the effect of leaning. According to the driver there is remarkable roughness of operation and weak acceleration aptitude with ethanol, even, to the extend, that it would be unacceptable for the market. Fig. 5 gives an overview of limited emission components in the ECE 47 driving cycle with both ethanol types E & EH. Regarding CO, HC and NO x three overlapping effects are visible: ECE 47 CO [g/km] E CO EH Fig. 6-1: Fuel & energy consumption at 40 km/h with different fuels. Piaggio Typhoon, dummy. HC [g/km] HC NOx [g/km] NOx Fig. 5: Limited emissions with different fuels. Piaggio Typhoon, dummy, CO, HC, NOx bag values Fig. 6-2: Exhaust temperatures (tailpipe) at 40 km/h and maximum speed with different fuels. Piaggio Typhoon, dummy. The volumetric fuel consumption in Fig. 6a increases with growing E-content. This on one hand according to the lower heat value of the blend fuels. On the other hand it must be assumed, that due to the irregular

7 combustion there is a slight deterioration of the engine effective efficiency. The comparison of energy consumption (bars on the right side) eliminates the influence of heat value and confirms the last assumption. With ethanol there is increased exhaust temperature (measured 30cm after tailpipe) due to a retarded combustion and higher cyclic dispersion. The working cycles with a retarded heat release offer much higher temperature of exhaust. In cycles with failing combustion part of fuel in the exhaust line may cause exothermic heating also without oxycat. The use of ethanol enables lower maximum speed. The hydrous ethanol generally increases the observed effects. Fig. 7 represents the integral NP-emissions with all fuels at constant speed v = 40 km/h. PM [g/km] DC [μm2/cm3] SMPS [ nm] E E E E+05 E E E E+08 E+00 Vconst=40 km/h E EH PM DC NP Fig. 7: Particle emissions with different fuels. Piaggio Typhoon, dummy. Beside the particle mass (PM) there are integrated DCsignals over the last 10 min of constant speed and integrated NP-count concentrations over the particle sizes in the SMPS measuring range [ nm]. The NanoMet signal is converted to the values responding to the undiluted volume concentrations in the exhaust DC (diffusion charging sensor) measures the total particle surface independent of the chemical properties. It indicates the solids and the condensates. In this figure only DC is represented, since PAS indicates zero-values. Very often by the 2-S exhaust aerosol the solids are enveloped by the condensates (SOF) and are not detected by PAS (which is sensitive only to the carbonaceous surface). The total aerosol surface DC indicates all particles and correlates usually very well with the particle mass PM for this type of aerosol. It can be stated, that with increasing share of ethanol the summary surface of NP s (DC) increases. The PM- and NP- emission level with ethanol fuels is generally higher, than with oline. This tendency is at first glance contrary to the expectations since ethanol, as a donator of oxygen should improve the combustion. Nevertheless there are other facts in the complex physico-chemical processes running in the engine, which overcompensate this first influence of oxygen donation: particle mass and nanoparticles in a small 2-stroke engine consist mostly of lube oil as unburned, partially burned, or condensates of heavy HC, the mechanisms of particle production are not only due to the combustion itself, but also to the wall quenching, flow during the scavenging (effects on residual ratio) and condensation speed in the exhaust system, since ethanol generally lowers the process temperatures it influences all the mechanisms mentioned above, during the previous research with Alkylat oline and with different lube oils, [3], it was stated, that the chemistry i.e. the HC matrix in the exhaust plays a role for the condensation effects and influences clearly the NP & PM production, during the tests with Piaggio Typhoon the engine operation with ethanol fuels was not as smooth, as with oline, the ability of acceleration was much weaker and the cyclic dispersion of the engine working cycles (due to leaning by ethanol) is believed to be the most important influence on the higher particle emissions. For SMPS, Fig. 8 the tendencies are not monotone, but the differences are, as for this parameter in the range of (emiting + measuring) dispersion. Furthermore there are balancing effects between nuclei- and accumulation modes of the respective PSD s, which slightly varies from measurement to measurement. Why are the maxima of the NP concentrations with ethanol fuels much higher (up to 3.4 times), than with oline? The maxima are in the size range between approx. 40 and 70 nm, which is for the Diesel exhaust aerosol in the lower part of accumulation mode. In the

8 present case of 2-S engine the mechanisms and the nature of NP aerosol is quite different, than for Diesel. Fig. 8: counts dn/dlog Dp [1/cm 3 ] counts dn/dlog Dp [1/cm 3 ] 1.0E E E E E+04 SMPS PSD - spectra (logarithmic scale) 1.0E d [nm] E E E E E+09 E E E+09 EH10 SMPS PSD - spectra (linear scale) EH15 SMPS particle size distribution spectra at constant speed 40 km/h, warm, with different fuels. Piaggio Typhoon, dummy. Three effects can be suggested for this observed increase of particle counts: - improved oxidation of bigger oil droplets during the combustion with ethanol, which generates more of smaller particles, or some solids in the lowest size range, - store and release effects of some particles or substances from the combustion chamber or exhaust system walls, which offer some precursors of the observed nanoparticles (washing effects, [12]), - changed effects of spontaneous condensation in the exhaust system due to changed HC-matrix and eventually generation of new substances (the first effect is known from the research with different oils and fuels on 2-S scooters [3] and the second effect is know from the research with water emulsion fuels on Diesel engines). The repetitions with Piaggio Typhoon were performed to test the driving quality and the emissions with lower portions of Ethanol blended and with richer basic tuning of the engine. The basic tuning of air / fuel-ratio is set at idling (by measuring CO) and has for this type of simple carburator influence on the air / fuel dosing up to higher part load operation of the engine. E15 E15 EH10 EH15 E d [nm] The mixture tuning was controlled at idling (without SAS): before, after and during the tests. There were following values: original mixture tuning before tests CO idl. 3.0 % n = 1670 rpm after tests CO idl. 2.8 % n = 1800 rpm rich mixture tuning before tests CO idl. 5.3 % n = 1700 rpm after tests CO idl. 5.0 % n = 1340 rpm Comparing the original mixture tuning with the previous values (from page 6) a quite poor stability of idling setting after longer period (3 months) can be remarked. Fig. 9 shows the comparison of limited emissions at constant speed. CO [g/km] HC [g/km] NOx [g/km] Max. exh. temp. [ C] r rich CO HC NOx t exh. Fig. 9: Limited emissions and exhaust temperatures (tailpipe) with different fuels; lean rich. Piaggio Typhoon; dummy; CO, HC, NOx bag values - 8 -

9 Fig. 10 represents the (nano)particle emissions. The variants with richer basic tuning were designed with r. counts dn/dlog Dp [1/cm3] E E E E E E+03 SMPS PSD - spectra 10 d [nm] PM [g/km] E5, r E5, r, r r rich 1.6E E E+05 E+05 E+00 only 3rd sample DC [μm 2 /cm 3 ] E5, r E5, r, r 8.0E E+08 E E+08 E+00 Fig. 10: Particle mass and nanoparticles at constant speed 40 km/h, warm, with different fuels; lean - rich. Piaggio Typhoon, dummy. The richer tuning provokes generally lower particle count concentrations in the nuclei mode. There are more agglomeration effects and the particle mass (which has similar values for all variants) is produced by particles of bigger sizes. Another clear influence of the richer tuning is a satisfactory driveability of the vehicle. The lower cyclic dispersion of combustion has positive impact on HC, which stays nearly constant in spite of the richer mixture. The lower cyclic irregularities influence the mechanisms of NP-production in the combustion chamber and in the exhaust line. The richer tuning increases the tolerance of ethanol application and makes it possible to use with similar values of particle mass and fuel consumption as the original fuel but with lower NO x. With the original, leaner tuning there is increased exhaust temperature due to a retarded combustion and higher cyclic dispersion. There are some differences of shape of the SMPS PSDspectra between pervious measurements (June 08) and the later repetitions (Sept. 08). E5, r E5, r, r SMPS [10-400nm] E5, r E5, r, r The differences are particularly visible in the nucei mode (d < 30 nm), where the spontaneous condensation is a dominant effect. The nuclei mode is a very sensitive indicator of different influences, like: differences of tuning, irregularities of oil dosing and mixing with air, deposits of oil in the crankcase, or in the silencer (from cold starting), irregularities of carubator (dosing and mixture preparation) etc. It was stated in the previous research, that the 2-S scooters are not stable concerning the nanoparticles emissions and the PSD s over longer time periods. Looking on average results of leaner and richer engine setting in Fig. 10 it can be stated that there is no clear difference of PM-emissions, but clearly higher NP-count concentrations for the leaner variant. This higher NPvalues are caused by the higher nuclei mode, which in turn is mostly influenced by the changed conditions of the spontaneous condensation and nucleation in the exhaust system. Kleider Florett CO [g/km] HC [g/km] NOx [g/km] E EH CO 5% 10% 15% HC ECE 40 Fig. 11: Comparison of limited emissions with different fuels. Kreidler Florett, CO, HC, NOx bag values. Before, during and after tests the mixture tuning was controlled at idling. There were following values: rep 5% 10% 15% NOx 5% 10% 15% rep.

10 before tests CO idl.4.36 % n =1630 rpm after tests CO idl % n = 1750 rpm The tuning of engine mixture is very rich and there is no possibility to change it. Fig. 11 gives an overview of all limited eous emission components in the driving cycles. Regarding CO, HC and NO x following effects are visible: - leaning of mixture by increasing ethanol portion (by lowering CO and HC), - increasing the combustion peak temperatures and NO x formation with increasing ethanol portion and strengthening of these effects with hydrous ethanol, (moving from low Lambda, very rich towards Lambda 1, less rich). B [l/100km] Exh. temp. [ C] Max. speed [km/h] v=40 km/h 5% 10% 15% v=40 km/h Fig. 12: Fuel consumption and exhaust tempe-ratures (tailpipe) at 40 km/h and maximum speed with different fuels. Kreidler Florett. According to the driver there was no problem of driveability with the highest ethanol share E15, or EH15. B E EH rep t exh. constant speed 40 km/h, 10 min. 5% 10% 15% v max. 5% 10% 15% The volumetric fuel consumption is reduced due to the leaning of mixture and increased efficiency with the ethanol blend fuels, Fig. 12. There is in particular an advantage of EH15, over E15. Also the exhaust temperature at constant speed generally decreases with higher ethanol rate. With hydrous ethanol this decrease is less pronounced, which is an effect of stronger leaning with EH. Interesting is the influence on the maximum speed (maximum power), which even gives some advantages with hydrous ethanol. In Fig. 13 there are comparisons of: PM-values, integrated DC-signals over the last 10 min of constant speed and integrated NP-count concentrations over the particle sizes in the SMPS measuring range [ nm]. Vconst=40 km/h PM [g/km] DC [mm2/cm3] SMPS [ nm] E E E E E+05 E E+08 E E E E+08 PM 5% 10% 15% DC E rep. 5% 10% 15% E+00 5% 10% 15% Fig. 13: Particle emissions with different fuels. Kreidler Florett. It can be stated, that with increasing share of ethanol the summary surface of NP s (DC) decreases. DC decreases also with addition of water. The same is valid for the PM (total particle mass) and for integrated NPcounts. EH [1/cm 3 ] NP

11 It can be said, that for the rich tuning ethanol helps to oxidize the precursor substances of particles (according to the increased blend ratio). The application of hydrous ethanol (EH) increases slightly this tendency thanks to the presence of water. The measurements with SMPS were started after the driving cycles and after 5 min conditioning at the same constant speed (40km/h). Fig. 14 represents some choosen PSD s in logarithmic and in linear scale. It is visible that the application of ethanol moves the maxima of the PSD s to lower counts and to lower median diameters (CMD). The influence of hydrous ethanol comparing to the pure ethanol is not visible, it is overlapped by the emission variability. counts dn/dlog Dp [1/cm 3 ] counts dn/dlog Dp [1/cm 3 ] 1.0E E E+07 Fig. 14: SMPS particle size distribution spectra at constant speed 40 km/h, warm, with different fuels. Kreidler Florett. Honda Zoomer SMPS PSD - spectra (logarithmic scale) 1.0E+05, rep. E5 1.0E d [nm] E E E E+08 E E+08 EH5 EH5 E15 SMPS PSD - spectra (linear scale) EH15 Also for this 4-S scooter the mixture tuning was controlled at idling (with SAS) before and after tests. There were following values: before tests CO idl. 2.5 % n = 2020 rpm after tests CO idl. 3.1 % n = 1950 rpm Figures 15, 16 & 17 display the most important results: The eous components indicate the leaner tuning with increasing share of alcohol reduction of CO & HC,, rep. E15 E d [nm] E5 EH15 increase of NO x, where the effect of hydrous ethanol is slightly stronger. ECE 40 CO [g/km] HC [g/km] NOx [g/km] E EH CO HC NOx Fig. 15: Limited emissions with different fuels. Honda Zoomer, CO, HC, NOx bag values. B [l/100km] Exh. temp [ C] Max. speed [km/h] v=40 km/h E EH v=40 km/h Fig. 16: Fuel consumption and exhaust tempe-ratures (tailpipe) at 40 km/h and maximum speed with different fuels. Honda Zoomer. B t exh. v max

12 The fuel consumption is reduced due to the leaning and increase of the effective engine efficiency. There is a little impact of the used fuels with ethanol on the exhaust temperature (at 40km/h) and on the maximum speed. With hydrous ethanol there is almost no impact. The application of fuels with a lower heat value and the shift to leaner air/fuel ratio does not reduce the maximum power very much, because there is en effect of improved engine efficiency. ECE 40 PM [g/km] E EH PM counts dn/dlog Dp [1/cm 3 ] counts dn/dlog Dp [1/cm 3 ] 1.0E E E E E E E E+06 E+06 SMPS PSD - spectra SMPS PSD - spectra rep. rep. 10 d [nm] rep repetitions rep. rep. DC [μm2/cm3] CPC [1/cm3] E E E+06 E E+06 E+00 DC CPC Fig. 17: Particle emissions with different fuels. Honda Zoomer. The PM-emissions are in average much lower, than for the 2 strokers (10-40 times). In this context the indicated differences between E, EH & (Fig. 17) cannot be regarded as significant. Nevertheless the higher value with (here in ECE 40) belongs to the first measuring series with ethanol, which at the constant speed indicated much higher PM-& NP-values with. The driving cycle was performed after the constant speed and the effects of higher emissions with are not strong any more. During the repetition of this stationary measurement much lower values with were proved, Fig. 18. It became clear, that the higher emissions and the higher NP count concentrations in nuclei mode with (in the primary measuring series) were caused by other effects, than the fuel only E+00 Fig. 18: SMPS particle size distribution spectra at constant speed 40 km/h, warm, with different fuels. Honda Zoomer; rep... repeated test series Spontaneous condensates can originate from the sulfates deposited in the system during previous operation with oline, than released by ethanol (solvent, washing effects) and finally condensated. A supposition of release of solid particles in this size range must be excluded. Apart from this fact it can be stated, that: there is no influence of alcohol-fuels on PM, the NP-emissions at transient operation are much higher, as at steady state operation, ethanol blend fuels help to reduce the NP-emissions especially at transient operation. Regarding the repeatability of all parameters between the original test and the repetitions test (not represented here) it can be stated, that for some parameters, like fuel consumption and particle mass PM the differences measured with different fuels are in the range of measuring dispersion. Comparisons 10 d [nm] Fig. 19 shows the comparison of emissions 2-S 4-S with oline at 40 km/h. Once again the very well known impressive difference of particle emissions is demonstrated. The 2-S engine with lost-oil lubrication has much higher emission of particle mass (PM), particle counts (NP) and summary surface of aerosol (DC); this even if the 2-S in this comparison is with leaner tuning, than the 4-S engine.

13 counts dn/dlog Dp [1/cm 3 ] E E E E E E E E E E E+02 2-S CO [g/km] Fig. 19: Comparison of emissions 2-S -> 4-S at 40 km/h, warm. 2-S: Piaggio Typhoon, carb. no ox. cat. 4-S: Honda Zoomer, carb. no ox. cat. Fig. 20 puts together the most important effects of ethanol rate on the three investigated scooters. The leaning of the mixture by ethanol provokes by the lean tuned scooter a reduction of NO x and increase of PM and fuel consumption (due to irregularities of combustion). By the scooters with rich basic tuning the leaning by ethanol increases NO x and reduces PM and fuel consumption. The average basic λ - tuning (calculated according to SAE J1088) was: - for Piaggio Typhoon λ = fore Kreidler Florett λ for Honda Zoomer λ = 0.9 The leaning by use of E15 provokes an increase of the air excess factor Δ λ = 5 to 0.10 CONCLUSIONS SMPS PSD - spectra 4-S 10 d [nm] S HC [g/km] DC [μm 2 /cm 3 ] DC NOx [g/km] Cons [l/100km] 2-S (Piaggio Typhoon - lean) The original tuning of the mixture preparation and dosing is influenced by the ethanol containing fuels in the sense of leaner operation. This caused an irregular operation of the engine and a very much deteriorated load response, which would be unacceptable for the market. Generally a lower ethanol rate has to be recommended. 2-S NOx values multiplied by factor 10! 1.0E E E E E+03 Gasoline RON 95 PM [mg/km] NP SMPS [10-400] [1/cm 3 ] CO [g/km] NOx [g/km] PM [g/km] B [l/100km] Piaggio Typhoon 2-S (lean) ECE 47 CO E15 E20 NOx v=40 km/h E15 E E15 E E15 E20 Kreidler Florett 2-S (rich) ECE 40 E5 E15 E5 E15 v=40 km/h E5 E15 E5 E15 Fig. 20: Influences of ethanol on emission and fuel consumption of the investigated motorbikes. Regarding the legally limited components CO, HC and NO x three overlapping effects are visible: - leaning of mixture by increasing ethanol portion, - lowering the combustion peak temperatures and NOx formation with increasing ethanol portion and strengthening of these effects with hydrous ethanol, - retarded combustion, increased cyclic irregularities of combustion and increasing HC-emissions with ethanol and with growing ethanol share. The volumetric fuel consumption increases according to the lower heat value of the blend fuels and the deteriorated irregular combustion. Particle mass PM and particle counts generally increase with the ethanol fuels, with some tendencies of more increase for hydrous ethanol. 2-S (Kreidler Florett - rich) Honda Zoomer 4-S ECE 40 v=40 km/h The original tuning of the mixture preparation and dosing is influenced by the ethanol containing fuels in the sense of leaner operation. This caused by this quite rich tuned motorbike advantages in regard of emissions and fuel consumption E15 E20 E15 E20 PM E15 E20 B E15 E

14 Regarding the legally limited components CO, HC and NO x following effects are visible: - leaning of mixture by increasing ethanol portion, - increasing the combustion peak temperatures and NOx formation with increasing ethanol portion and strengthening of these effects with hydrous ethanol. The volumetric fuel consumption is reduced due to the leaning of mixture and increased efficiency with the ethanol blend fuels. Particle mass PM and particle counts generally decrease with the ethanol fuels, with some tendencies of more decrease for hydrous ethanol. 4-S (Honda Zoomer - rich) There was a good drivability of the investigated scooter with all ethanol blends. The limited eous components indicate the leaner tuning with increasing share of alcohol reduction of CO & HC, increase of NO x. The fuel consumption is reduced due to the leaning and increase of the effective engine efficiency. There is no remarkable reduction of maximum power in the investigated domain of equivalence ratio. Regarding the PM- and NP-emissions it can be stated, that: there is no influence of alcohol fuels on PM, the NP-emissions at transient operation are much higher, as at steady state operation, ethanol blend fuels help to reduce the NP-emissions especially at transient operation. That means, that alcohol helps to better oxidize the particles. At the beginning of working period with ethanol fuels there is a release ( washing out ) of residues from the previous operation with oline. These residues containing sulfates (from lube oil) offer precursors for the spontaneous condensates in the nuclei mode (nano range), which is sensitively indicated by the NP measuring methods and is also visible in the PM emissions. General From the present results it can be concluded, that the basic tuning of the engine not too lean decides if the influence of leaning by means of ethanol blends has positive, or negative effects. The 2-stroker can be eventually improved by richer basic tuning. The 4-stroker can be deteriorated by the irregularities of combustion, if the basic tuning would be too lean (this can be the case of a modern 4-stokers). Summarizing: the success of ethanol blend fuels depends on: the engine type (2-S, 4-S), the basic tuning of the engine (air-fuel-ratio) and the ethanol content. There are no significant differences of results between the blends with pure ethanol (E), or hydrous ethanol (EH), except of some cases, where EH improves slightly the emissions (CO, HC, PM, NP) and reduces the fuel consumption. The present investigations did not concern the durability of parts exposed to the chemical influences of ethanol. Also the cold startability, particularly in extreme conditions and the lube oil dilution were not addressed. According to the fuel supplier, there is no danger of phase separation of the hydrous ethanol in the used temperature range. ACKNOWLEDGEMENTS The authors would like to express their gratitude for the support of the project to BAFU (Swiss EPA), Dr. M. Schiess and to the Process Design Center B.V.NL and TU Delft. REFERENCES [1] Czerwinski, J.; Comte, P.; Napoli, S, Wili, Ph.: Summer Cold Start and Nanoparticulates of Small Scooters. Report B086 for BUWAL (SAEFL) Bern, Lab. For Exhaust Gas Control, Univ. of Appl. Sciences, Biel-Bienne, Switzerland, Nov SAE Techn. Paper [2] Czerwinski, J.; Comte, P.: Limited Emissions and Nanoparticles of a Scooter with 2-stroke Direct Injection (TSDI). SAE Techn. Paper [3] Czerwinski, J.; Comte, P.; Reutimann, F.: Nanoparticle Emissions of a DI 2-Stroke Scooter with varying Oil- and Fuel Quality. SAE Techn. Paper [4] Czerwinski, J.; Comte, P.; Larsen, B.; Martini, G.; Mayer, A.: Research on Particle Emissions of modern 2-S Scooters. SAE Techn. Paper [5] Czerwinski, J.; Comte, P.; Astorga, C.; Rey, M.; Mayer, A.; Reutimann, F.: (Nano) Particle from 2-S Scooters: SOF / INSOF; Improvements of Aftertreatment; Toxicity. AFHB, JRC, TTM, BAFU, SAE Techn. Paper [6] Czerwinski, J.; Comte, P.; Violetti, N.; Landri, P.; Mayer, A.; Reutimann, F.: Catalyst Aging and Effects on Particle Emissions of 2-Stroke Scooters. SAE Techn. Paper

15 [7] Bergström, K.; Melin, S-A.; Coleman, J. General Motors Powertrain: The New ECOTEC Turbo BioPower Engine from GM Powertrain Utilizing the Power of Nature s resources. 28. Internationales Wiener Motorensymposium 2007, Bd.2, S.47. [8] Bergström, K.; Nordin, H.; Königstein, A. GM Powertrain Europe; D. Marriott, C.D.; Wiles, M. A., GM Powertrain North America: ABC - Alcohol Based Combustion Engines Challenges and Opportunities. 16. Aachener Kolloquium Fahrzeug- und Motorentechnik 2007, Bd.2, S [9] Kawai, T.; Tsunooka, T.; Chiba, F.; Uda, H.; Sonoda, Y.; Toyota Motor Corporation, Japan: Effect of high Concentration Ethanol on SI Engine Cold Startability and Emissions. 16. Aachener Kolloquium Fahrzeug- und Motorentechnik 2007, Bd.2, S [10] DuMont, R. J.; Cunningham, L. J.; Oliver, M. K.; Studzinski, W. M.; Galante-Fox, J. M.: Controlling Induction System Deposits in Flexible Fuel Vehicles Operating on E85. SAE Techn. Paper [11] Galante-Fox, J. M.; Von Bacho, P.; Notaro, C.; Zizelman, J.:E-85 Fuel Corrosivity: Effects on Port Fuel Injector Durability Performance. SAE Techn. Paper [12] CONCAWE... Conservation of Clean Air und Water in Europe, Brussel, [13] Gottschalk, A.: Gasoline Blending with Hydrous Ethanol. 7 th International Colloquium Fuels, Technische Akademie Esslingen TAE, Jan , [14] Hydrous Ethanol Blends. HE Blends B.V. [15] Czerwinski, J.; Comte, P.; Mayer, A; Reutimann, F; Zürcher, D.: Reduction Potentials of Particle Emissions of 2-S Scooters with Combinations of Technical Measures. FISITA, Munich, Germany, Sept , Paper F , Congress Proceedings Vol. IV, p. 100, ATZ / ATZ auto technology, Springer Automotive Media, Wiesbaden, D. ABBREVATIONS AFHB Abprüfstelle der Fachhochschule, Biel CH (Lab. For Exhaust Gas Control, Univ. of Appl. Sciences, Biel-Bienne, CH) BAFU Bundesamt für Umwelt (Swiss EPA) C Carburetor Carb Carburetor CMD count median diameter CPC condensation particle counter CVS constant volume sampling DC diffusion charging sensor DF dilution factor DMA differential mobility analyser E pure ethanol (energy only Fig. 6a) EC elemental carbon EH hydrous ethanol FHB Fachhochschule Biel Gas oline leed-free, RON 95 (base fuel) MD minidiluter ME Matter Engineering, CH NanoMet minidiluter + PAS + DC (+TC) NP nanoparticulates (< 1μm) OC organic carbon PAS photoelectric aerosol sensor PM particulate matter, particulate mass PMP particle measuring program (of the ECE GRPE) PSD particles size distribution r rich basic tuning SAS secondary air system SMPS scanning mobility particles sizer SOF soluble organic fraction TC thermoconditioner, total carbon TP tailpipe TPN total particle number [# / km] 2-S 2 stroke engine 4-S 4 stroke engine