Fourier Transform Infrared Spectroscopy Models to Predict Cetane Number of Different Biodiesels and Their Blends

The ignition quality of a fuel is described by its cetane number. Experimental methods used to determine cetane number employ Co-operative fuel research (CFR) engine and Ignition quality tester (IQT) which are expensive, have less repeatability and require skilled operation, and hence least preferred. There are many prediction models reported, which involve number of double bonds and number of carbon atoms whose determination is not direct. Using models that relate biodiesel composition to its cetane number is limited by the range of esters involved. Hence, a model to predict cetane number of biodiesels that addresses the limitations of the existing models, without ignoring the influence of factors such as degree of unsaturation and number of carbon atoms, is needed. Fourier transform infrared spectroscopy (FTIR) could be one such method. Five biodiesels with significant compositional variations were prepared from Camelina, Coconut, Karanja, Linseed and Palm oils, and blended in different volume proportions to arrive at 70 samples. The range of cetane number covered was from 42.2 to 65.4. Peak absorbance of different functional groups of these samples and peak ratios were determined using FTIR which were correlated to their cetane number to develop prediction models using regression. These models were validated using biodiesels data that are not used in developing them. Mean absolute deviation and Mean absolute percentage error were the statistical parameters used to compare the proposed model with existing models whose values turned out to be considerably good.