Correlating the cloud point of biodiesel with its fatty acid methyl ester composition: Multiple regression analyses and the weighted saturation factor (wSF)

Abstract

Biodiesel is an alternative fuel that is typically composed of fatty acid methyl esters (FAME) from transesterification of fats or plant oils with methanol. When blended with conventional diesel fuels, biodiesel improves combustion quality and reduces harmful exhaust emissions. On the other hand, biodiesel has poor cold flow properties that can compromise its use during cold weather in moderate temperature climates. In the present study, nine mathematical models were proposed to calculate the cloud point (CP) of biodiesel based on the concentrations and melting properties of the FAME components. Experimental data were compiled for four biodiesel fuels made from canola, palm and soybean oils and yellow grease (CaME, PME, SME and YGME) plus 24 binary admixtures of these fuels. Four models were inferred from multiple regression analysis of CP as functions of Xi = (MPixi; MPi = melting point and xi = mole fraction of the FAME species). These models yielded high correlation coefficients (R2 ≥ 0.997) and low absolute average deviations (AAD) ≤ 0.60 K from analysis of calculated (CPcalc) versus measured data. The results helped to identify species that influenced the CP of FAME mixtures. Five additional models based on the weighted saturation factor (wSF) index as an independent variable for correlating the CP of biodiesel were inferred. The indices were calculated as weighted sums of Xi terms for species present in the FAME mixtures. Both an arbitrarily assigned weight factor (wi) and the MPi of the pure FAME species were used to calculate the wSF indices. The wSF index-based models were developed from regression analysis of FA profile and experimental CP data for the 28 FAME mixtures. Four models yielded R2 = 0.968–0.977 and AAD ≤ 0.73 from analysis of CPcalcversus measured CP data.