Flash point of binary and ternary mixture of biojet blends: Experimental data and modeling

The flashpoint (FP) holds significant importance in the fire risk mitigation and prevention strategies within the fuel industry, specifically when considering biojet blends. This property directly impacts the safety aspects associated with the handling and transportation of biojet fuel, moreover the FP of biojet blnds exhibits variability based on the specific type of biofuel utilized and the percentage incorporated into the blend. Hence, it is crucial to employ a precise model for the estimation of the FP in newly developed biojet blends, ensuring compliance with safety standards governing their handling and transportation. Determination and modeling of such property is important, considering that the presence of other compounds (e.g. ethanol, esters) in a fuel mixture can change its FP. This study emphasized the sensitivity of fuels' FP regarding ethanol concentration and the applicability of activity coefficient models to predict the FP of mixtures. The FP of mixtures containing ethanol, hydrocarbons (dodecane, tetradecane and hexadecane) and esters (methyl palmitate, ethyl palmitate and biodiesel) as aviation fuel surrogates were experimentally determined from 0.0 to 1.0 molar fraction range for both binary and ternary mixtures, totalizing 11 binary mixtures and 5 ternary mixtures. For all system, an abrupt decrease in the FP temperature was observed with small additions of ethanol. The FP drop is less expressive for the binary mixtures formed by hydrocarbon esters, due to the similarities between the compounds and FP of the pure compounds. Besides, the Liaw's approach combined with UNIFAC, UNIFAC-Dortmund, UNIQUAC, and NRTL models as well as COSMO-RS, were used to calculate the FP data of the studied mixtures, resulting in global root-mean-square deviations (RMSD) with an average deviation ranging from 0.24 K to 12.03 K. The predictions provided by all models were more accurate than the ideal approach (RMSD = 17.42 K) for the studied mixtures. These results highlight the significance of accounting for non-ideal effects when estimating FP data for such systems. Additionally, the COSMO-RS model and UNIFAC proved to be an advantageous model to predict the shape of the FP curve and helped the selection of suitable surrogates for the biodiesel system and provide essential information for accurate process safety and fire risk assessment in the biojet fuel industry.