Book Description

Biofuels, such as biodiesel, offer benefits as a possible alternative to conventional fuels due to their fuel source sustainability and their reduced environmental impact. Before they can be used, it is essential to understand their combustion chemistry and emission characterizations due to a number of issues associated with them (e.g., high emission of nitrogen oxides (NOx), lower heating value than diesel, etc.). During this study, emission characterizations of different biodiesel blends (B0, B20, B50, and B100) were measured on three different feedstocks (soybean methyl ester (SME), tallow oil (TO), and waste cooking oil (WCO)) with various characteristics, while an ultra-low sulfur diesel (ULSD) was used as base fuel at low-temperature combustion (LTC). A laboratory combustion chamber was used to analyze soot formation, NOx emissions, while real engine emissions were measured for further investigation on PM and NOx emissions. For further study, carbon emissions (CO, CO2, and CH4) were also measured to understand their relations with feedstocks' type. The emissions were correlated with fuel's characteristics, especially unsaturation degree (number of double bonds in methyl esters) and chain length (oxygen-to-carbon ratio). The experimental results obtained from laboratory experiments were confirmed by field experiments (real engines) collected from Toledo area regional transit authority (TARTA) buses. Combustion analysis results showed that the neat biodiesel fuels had longer ignition delays and lower ignition temperatures compared to ULSD at the tested condition. The results showed that biodiesel containing more unsaturated fatty acids emitted higher levels of NOx compared to biodiesel with more saturated fatty acids. A paired t-test on fuels showed that neat biodiesel fuels had significant reduction in the formation of NOx compared with ULSD. In another part of this study, biodiesel fuel with a high degree of unsaturation and high portion of long chains of methyl esters (SME) produced more CO and less CO2 emissions than those with low degrees of unsaturation and short chain lengths (WCO and TO, respectively). In addition, biodiesel fuels with long and unsaturated chains released more CH4 than the ones with shorter and less unsaturated chains. Experimental results on soot particles showed a significant reduction in soot emissions when using biodiesel compared to ULSD. For neat biodiesel, no soot particles were observed from the combustion regardless of their feedstock origins. The overall morphology of soot particles showed that the average diameter of ULSD soot particles was greater than the average soot particle from biodiesel blends. Eight elements were detected as the marker metals in biodiesel soot particles. The conclusion suggests that selected characterization methods are valuable for studying the structure and distribution of particulates. Experiments on both PM and NOx emissions were conducted on real engines in parallel with laboratory study. Field experiments using TARTA buses were performed on buses equipped with/without post-treatment technologies. The performance of the bus that ran on blended biodiesel was found to be very similar to ULSD. As a part of this study, the toxic nature of engine exhausts under different idling conditions was studied. The results of the PM emission analysis showed that the PM mean value of emission is dependent on the engine operation conditions and fuel type. Besides, different idling modes were investigated with respect to organic carbon (OC), elemental carbon (EC), and elemental analysis of the PMs collected from public transit buses in Toledo, Ohio. In the modeling portion of this work, a simplified model was developed by using artificial neural network (ANN) to predict NOx emissions from TARTA buses via engine parameters. ANN results showed that the developed ANN model was capable of predicting the NOx emissions of the tested engines with excellent correlation coefficients, while root mean square errors (RMSEs) were in acceptable ranges. The ANN study confirmed that ANN can provide an accurate and simple approach in the analysis of complex and multivariate problems, especially for idle engine NOx emissions. Finally, in the last part of the modeling study, a biodiesel surrogate has been proposed and main pathways have been derived to present a simple model for NOx formation in biodiesel combustion via stochastic simulation algorithm (SSA). The main reaction pathways are obtained by simplifying the previously derived skeletal mechanisms, including saturated methyl decenoate (MD), unsaturated methyl 5-decanoate (MD5D), and n-decane (ND). ND is added to match the energy content and the C/H/O ratio of actual biodiesel fuel. The predicted results are in good agreement with a limited number of experimental data at LTC conditions for three different biodiesel fuels consisting of various ratios of unsaturated and saturated methyl esters. The SSA model shows the potential to predict NOx emission concentrations, when the peak combustion temperature increases through the addition of ULSD to biodiesel. The SSA method demonstrates the possibility of reducing the computational complexity in biodiesel emissions modeling. Based on these findings, it can be concluded that both alternative renewable fuels (biodiesel blends) as well as the LTC condition are suitable choices for existing diesel engines to improve the sustainability of fuel and to reduce environmental emissions.