A composition based approach to predict density, viscosity and surface tension of biodiesel fuels

Abstract

The variations in biodiesel composition based on the feedstock used for its production result in significant variations in the fuel properties and thereby, engine characteristics. This necessitates the need for adopting a composition based approach to analyze biodiesel fuel properties and engine characteristics. Based on this premise, biodiesel composition based multi linear regression models are proposed in the present work to predict the density, kinematic viscosity and surface tension of biodiesel. Biodiesel composition effects are captured through two new parameters, namely straight chain saturation factor (SCSF) and modified degree of unsaturation (DUm) which represents the effects of saturated and unsaturated methyl ester constituents of biodiesel, respectively. SCSF and DUm are found to better capture the effects of biodiesel composition as compared to the parameters available in the literature, namely long chain saturated factor, modified long chain saturated factor and degree of unsaturation by addressing their limitations. The proposed models also account for the effects of temperature on the variations of biodiesel properties. To train and validate the proposed multi linear regression models, the density and kinematic viscosity in the temperature range of 293 to 373 K and surface tension in the temperature range of 303 to 353 K of seven biodiesels having significantly different composition are measured. For validation, the measured property data of three additional biodiesels not considered in training the models are used. The predictions compared well with measured data with regression coefficients of 0.98, 0.97 and 0.93 and mean absolute percentage error of 6.28, 0.36 and 1.4 percent for kinematic viscosity, density and surface tension, respectively. Statistical analysis carried out using ANOVA revealed a strong correlation between biodiesel composition based parameters, SCSF and DUm and aforementioned properties. Further, DUm is found to have a profound influence on the kinematic viscosity and density of biodiesel as compared to SCSF while its effects on the surface tension are statistically insignificant.