determination model for cetane number of biodiesel at different fatty acid composition: a review

The most accepted definition of biodiesel is stated at the EU technical regulation EN 14214 (2008) or in the USA in ASTM 6751-02. As a result of this highly strict description only methyl esters of fatty acids conform to these definitions, nevertheless the term ‘‘biodiesel’’ is spread out to other alkyl fatty esters. Some countries have adopted bioethanol for replacement of methanol in biodiesel transesterification and thus assuring a fully biological fuel. Of course, such position brings some problems in fulfilling technical requirements of EN 14214 or ASTM 6751-02. Biodiesel is actually a less complex mixture than petrodiesel, but different feedstock origins and the effect of seasonality may impose difficulties in fuel quality control. Since biodiesel is an alternative diesel fuel derived from the transesterification of triacylglycerol comprised materials, such as vegetable oils or animal fats, with simple alcohols to furnish the corresponding mono-alkyl esters, its composition depends on the raw material used, the cultivated area location, and harvest time. The choice of the raw material is usually the most important factor for fluctuations of biodiesel composition, because different vegetable oils and animal fats may contain different types of fatty acids. Important properties of this fuel vary significantly with the composition of the mixture. Cetane number, melting point, degree of saturation, density, cloud point, pour point, viscosity, and nitrogen oxides exhaust emission (NOx), for instance, deserve to be mentioned. One of the most important fuel quality indicators is the cetane number; however its experimental determination may be an expensive and lengthy task. To weaken situation concerning biodiesel, the availability of data in the literature is also scarce. In such scenario, the use of reliable models to predict the cetane number or any other essential characteristic may be of great utility. We reviewed available literature to describe model, which will correlate cetane number of single mono-alkyl esters of fatty acids and their mixtures (biodiesels) with structural characteristics achieved by hydrogen nuclear magnetic resonance spectra. We suggest that this could be done through the application of a three stages process, combining statistical tools, fuzzy logic and artificial neural networks. Both treatments pointed to two major characteristics as determinants for the cetane number values: the number of carbon-carbon double bonds and the structure of alcohol moiety in each fatty ester. Neural Networks predicted cetane number quantitatively. The results are beneficial for the further development of analytical models addressing the quality of biofuels.