Loading...
Please wait, while we are loading the content...
Similar Documents
Effect of alcohol blending on real driving emissions of particulate matter from ordinary gasoline automobile engines: a comparison of ethanol, n-butanol and isobutanol
| Content Provider | Semantic Scholar |
|---|---|
| Author | Vojtisek-Lom, Michal Beránek, Vít Klír, Vojtěch Pechout, Martin Mazáč, Martin Dittrich, Luboš |
| Copyright Year | 2015 |
| Abstract | This paper summarizes the recent and ongoing work o n eal driving emissions of several automobiles with ordinary, non-flexible-fuel spark ignition engines, powered by alcoholgasoline blends with higher concentrations of ethan ol, n-butanol and isobutanol. On a Ford Focus automobile with a direct injection EcoBoost e ngine, powered by gasoline and its blends with 15% ethanol, 25% n-butanol and 25% isobutanol, particle size distribution were measured with an on-board fast mobility particle sizer along a 55 km route. Particle emissions were moderately reduced by ethanol and considerably by b oth utanol blends. On a Škoda Fabia and a Škoda Felicia cars with indirect injection engine s, powered by blends with higher concentrations of ethanol, n-butanol and isobutanol , particle emissions measured by a miniature on-board system were examined over a 13 km route. B lends of 30% and 50% of butanol had no or slightly positive effect on particle emissions. Blends of 70% ethanol and 85% n-butanol and 85% isobutanol, used with an auxiliary engine contr l unit, had no to slightly positive effect on particle mass, and reduced total particle length (r oughly corresponding to lung deposited surface area) by about one half. Introduction This study evaluates the emissions performance of o rdinary in-use gasoline engines when operated on higher concentrations of ethanol, n-but anol and isobutanol, with focus on real-world particulate matter emissions (real driving emission ). Replacement of fossil automotive fuels with renewab le, low carbon footprint, domestically produced fuels and reducing exhaust emissions of pr imarily particulate matter and secondarily nitrogen oxides are among the main challenges autom obile engines are currently facing. A large variety of fuels have been examined, out of which s everal have obtained larger market penetration: natural gas in gaseous and liquid form , liquified petroleum gas, ethanol, and biodiesel. Of these, ethanol and biodiesel are prod uce from renewable resources, with ethanol being used primarily in spark ignition engines, and biodiesel virtually exclusively in compression ignition engines. Ethanol is an oxygenated compound with 35% of oxyge n by weight. For this reason, more ethanol (both by weight and by volume) is needed, c ompared to gasoline, to form a stoichiometric mixture with a given amount of air. Therefore, on any engine calibrated to run on gasoline, the quantity of the fuel delivered must b e increased when running on ethanol. There are therefore two strategies to use ethanol: either blended in small concentrations (up to around 10%) with gasoline for the general use, or in high concentrations in designated engines. The current practice in the Czech Republic, where E85 ( spark ignition engine fuel containing 7085% of ethanol) is widely available at filling stat ions, while the number of flexible fuel vehicles certified to run on this fuel is rather small, sugg ests that ethanol is used in higher concentrations in the existing vehicle fleet. Assuming that the fu l does not lead to adverse performance (otherwise it would not be used by the public), the remaining question is the effect of such practice on exhaust emissions. The effects observed during laboratory studies are reviewed in [1-3] and in the previous works by the authors [4-6 ]. It has been, however, known that the emissions under realistic driving conditions are of t n higher than during standardized typeapproval laboratory tests. Therefore, the question of the effects of higher concentrations ethanol blends on real driving emissions was sought to be a nswered by real driving emissions tests. Also, as ethanol is known to be hygroscopic and agg ressive to many elastomers and other materials found in the fuel systems [7,8]. For this reason, additional alcohols which could also be produced from biomass were considered. Two isome rs of butanol, n-butanol (1-butanol) and isobutanol (2-methyl-propan-1-ol), have the potenti al to be commercially produced from biomass [9-12] at costs and fossil energy inputs co mparable to ethanol [9]. Compared to ethanol, both n-butanol and iso-butanol have higher energy d ensity, lower hygroscopicity, higher viscosity, better lubricity, lower vapor pressure [ 13], and are less aggressive towards many materials commonly used in vehicle fuel systems. Bo th isomers of butanol have been used in spark ignition engines, both port fuel injection [1 -5,14-19] and direct injection type [6,20-22], with encouraging results, yet without a universal c onsistent conclusion as to the effect on the emissions. In the recent past, the performance of butanol blen ds has been investigated by the authors on several engines, including throttle body injection, port injection and direct injection automobile gasoline engines, and several small carbureted engi n s used in garden equipment and an electric generator. Of these, three automobiles have been te sted under real driving conditions with a portable on-board monitoring system, during which t he emissions of particulate matter were also measured. These measurements are summarized in this paper. Experimental Portable on-board monitoring system The vehicles were fitted with a portable, on-board exhaust emissions monitoring system designed by the first author [23,24]. The system sa ples raw, undiluted exhaust gases via a 6 mm diameter stainless steel tube inserted into the tailpipe, and a 6 mm internal diameter, 5 m long conductive fuel line used as a sample line. Th e sample passes through condensation bowl where condensate is trapped and periodically remove d. The sample is then reheated to approximately 60 C by passing through a resistanceheated copper coil. Concentrations of nitrogen monoxide (NO), carbon monoxide (CO) and ca rbon dioxide (CO2) were measured online with a pair of modified, optimized and tuned BAR-97 grade analyzers, utilizing nondispersive infra-red analyzers (HC, CO, CO 2) and electrochemical cells (NO and NO 2). The response of the NDIR sensor used in this study to e thanol and to hydrocarbon mix during the operation on E85 has not been determined. Specifica tions for an analogous detector [25] show that the sensitivity to ethanol lies between the se nsitivities of propane and hexane, both of which are commonly used to calibrate the automotive NDIR analyzer. Traditionally used flame ionization detector (FID) was not determined to be a r liable reference, as it has cross-sensitivity to oxygenated compounds, resulting in understatemen t of the concentration of oxygencontaining hydrocarbons [26]. Also, as the sample s ystem is not heated, and portion of water vapor in the sample is removed by the condensate, i t can be presumed that ethanol, which is water-soluble, is lost to condensate. Ethanol has b een found to be one of the major constituents of organic species on ethanol-fueled vehicles [27]. The CO and CO2 measurements using the NDIR method are rather straightforward and no adver se issues were anticipated. While the instrument measures both NO and NO 2 using electrochemical cells, only the NO measureme nt is sufficiently dynamic for transient tests, and was e valuated here quantitatively. The volumetric concentrations of total nitrogen oxides (NO x) were assumed to be identical to those of NO in most cases during this study. This overall assumpti on has been verified by extensive comparison tests of the on-board system, and is also in agreem ent with analogous sensors being used, in many regions, in periodic emissions inspections of park ignition vehicles nominally operating at stoichiometric ratio. This is also in agreement with general experience that for engines with no catalytic devices and for engines operating most ly at stoichiometric conditions, the concentrations of nitrogen dioxide (NO 2) are several percent of the total nitrogen oxides (NOx); the only engines known to produce relatively high e missions of NO2 are those equipped with a highly doped oxidation catalyst and operating lean (with excess air). This is also in agreement with the observed range of response of the NO 2 cell, based on which it is not apparent that large r quantities of NO2 (tens of percent of total NO x) were produced. Concentrations of particulate matter were measured online with a forward scattering integrating nephelometer, which, for a given engine and a given setup, tends to provide output proportional to particle mass concentration [4,23]. This measure ment is believed to be possibly affected both by the low light scattering efficiency of smaller p articles, and by the effects of fuel on the particle composition and morphology (mean size, fra ctal dimension, ...) and hence on the ratio of the light scattering efficiency to particle mass . Concentrations of particulate matter expressed as t ot l particle length were measured with a modified industrial building smoke detector equippe d with a measuring ionization chamber utilizing a small radioactive source (241Am, 30 kBq ) to ionize the air. When voltage is applied to the electrodes in the chamber, a small ionizatio n current flows through the chamber. Particles entering the chamber absorb the ions and decrease t he ionization current. The detector was modified so that ionization current can be sensed d irectly and recorded by a data acquisition system. Laboratory comparison tests carried on engi n exhaust by the first author [28] have shown that the system provides a response proportio nal t total particle length concentration (i.e., ft of particles per cu.in., or m.cm-3), that is, the sum of electric mobility diameters of all particles in a unit of volume. On the DISI engine, where particulate emissions wer e anticipated to be the primary issue, particle size distributions and concentrations were measured online with a fast mobility particle sizer (EEPS, Model 3090, TSI), preceeded by a secon dary dilution by a rotating disc diluter (MD-19, Matter Engineering) set to 180:1 dilution r atio; t |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.medetox.cz/wp-content/uploads/2012/04/BIOTRETH-Vojtisek-butanols-real-driving-final.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
