Fuel composition impact on heavy duty diesel engine combustion & emissions

Abstract

The Heavy Duty Diesel or compression ignition (CI) engine plays an important economical role in societies all over the world. Although it is a fuel efficient internal combustion engine design, CI engine emissions are an important contributor to global pollution. To further reduce engine emissions and improve fuel efficiency, modification of the fuel composition is an interesting option. First of all, a (partial or gradual) change over to synthetic fuels can reduce the dependency on fossil fuel, which enables reduction of CO2 emission if renewable feedstock is used. Secondly, modification of the fuel composition potentially can benefit the CI engine's emission tradeoff between particulate matter (PM) and nitric oxides (NOx). The aim of the present work is to give further insight into fuel composition impact on CI engine combustion and resulting emissions. It is shown that fuel composition changes are a powerful instrument to change the PM - NOx tradeoff on existing CI engine designs. Modern CI engines inject liquid fuel into the combustion chamber of the engine. This happens typically at high hydraulic pressure (up to 250 MPa), using a fuel injector with typically 6 to 9 injector holes during a very short time (<3 ms), very close to the end of the compression stroke of the engine. To create fundamental understanding of the CI engine combustion characteristics, the fuel spray vaporization and combustion phenomenology needs to be studied in detail. To do so, an optically accessible constant volume cell EHPC (Eindhoven High Pressure Cell) is used in this work. To facilitate the investigations, the EHPC was further developed to create engine like conditions using the pre-combustion method. This method uses a premixed combustion and by adjusting the mixture composition of the premixed gases, the residual gas mixture composition and properties can be tuned. This means that at the moment of fuel injection, during the cool down period after the pre-combustion event, non-reacting or reacting sprays can be studied. This method allows coverage of ambient conditions in this work up to a gas density of 32 kg/m3 for a temperature range of 750 to 1200 K and oxygen concentrations between 0 and 21 volume percent. To characterize the fuel spray using the EHPC, optical diagnostics and image analysis methods have been developed. Schlieren is used to investigate non-reacting spray penetration and angle dispersion. The liquid core behavior of vaporizing sprays is measured using MIE scattering. To validate the results, regular automotive CI engine fuel is used (EN590). Comparing the non-reacting EN590 fuel spray results with literature data, shows that in general the spray behavior as function of ambient conditions is in line with the trends reported. Detailed analysis reveals that, for the data captured in this work, the often used power law to describe spray penetration is not accurate enough. The spray dispersion angle at one condition is out of the expected range and more experimental data is required to get more understanding and better statistics. This outlier also slightly affects the validation of the liquid length measured using laser MIE scattering at this specific condition. When the measured data is compared with a predicted liquid length using a mixing-limited vaporization model, the liquid length is shorter than expected. The general trend in the results is however still present. For reacting fuel spray characterization the ignition delay is measured using a pressure measurement and laser line of sight extinction is use to qualify the soot presence during quasi-steady state combustion. The results achieved with both methods are in line with results from literature. It can be concluded that the EHPC, the optical diagnostics used and image analysis methods are successfully implemented and validated using EN590 fuel. With this setup and tools at hand, two additional fuels were characterized: SMDS, which is a synthetic gas to liquid fuel, and SMDS blended with 29.3 mass percent Tripropylene G lycol Monomethyl Ether (TPG ME), called fuel S-TP-9. The latter results in an oxygenated fuel with 9.4 9 mass percent oxygen. B oth fuels show differences compared to EN590, especially for reacting fuel sprays. The higher Cetane Number of both fuels results in an, expected, reduced ignition delay. B oth fuels produce less soot during quasi-steady state combustion. O xygenated fuel S-TP-9 clearly shows an even lower soot production compared to SMDS. At reduced ambient temperatures and oxygen concentration soot production for all tested fuels is reduced, but the additional soot production reduction for S-TP-9 is less significant compared to the results at an ambient oxygen concentration of 21 volume percent. Additional to the fuel spray combustion investigations using the EHPC, a broader range of fuels was tested on Heavy Duty CI engine test rig to quantify the effects of fuel composition. The results show that oxygenated fuels up to 15 mass percent oxygen can reduce engine-out PM emission up to one order of magnitude at reduced in-cylinder temperatures and oxygen concentration. The other effects of fuel composition are limited to a minor impact on injection duration, apparent heat release and NOx emission. The results of engine and EHPC combined suggest that the beneficial influence of additional oxygen in the fuel on PM emissions, becomes even more effective after end of combustion, during the late expansion phase of the engine. How this process evolves up to the moment the PM emissions are measured in the exhaust system of the engine, should be part of further investigation.