Aplicação de técnicas quimiométricas na determinação de parâmetros fisicoquímicos para combustiveis

Abstract

Petroleum-derived fuels, known as non-renewable fuels, constitute the primary global energy source. Aviation kerosene (Jet A-1), depending on the standard) is a fuel utilized in aircraft turbines, obtained through the direct distillation of petroleum within a temperature range of 150 to 300 °C, and is predominantly composed of paraffinic hydrocarbons containing 9 to 15 carbon atoms. Within the class of renewable fuels, we have biodiesel (B100). This biofuel has been increasingly blended with mineral diesel and is commercially available for the operation of diesel engines. Fuels, whether renewable or non-renewable, adhere to specifications established by the ANP (Agência Nacional do Petróleo, Gás Natural e Biocombustíveis), which aims to set quality standards. A major issue involving these fuels is non-compliance with quality regulations. For biodiesel, this problem is related to the formation of solids resulting from its oxidation, a phenomenon increasingly common as the biodiesel content increases in diesel/biodiesel blends. Regarding Jet A-1, the problems are more associated with physicochemical parameters such as the Freezing Point and contamination by different solvents. Currently, standardized fuel analysis methods exist; however, these methods are generally expensive, consume large quantities of sample, and require specific equipment. In this context, the objective of this work was to develop supervised and unsupervised classification models and calibration models to identify and predict non-conformities related to these fuels. The use of Synchronous Fluorescence Spectroscopy (SFS) in conjunction with chemometric methods such as PLS and PLS-DA enabled the development of models capable of predicting the Oxidative Stability of B100 with a validation R2 of 0.822 and quantifying the content of contaminants in Jet A-1 with a validation R2 above 0.9 for samples contaminated with gasoline (GA), aviation gasoline (GAV), and illuminating kerosene (QL). For the contaminant commercial kerosene (QC), the validation R2 was 0.741. Furthermore, the classification models for both the types of contaminants and the Freezing Point of Jet A-1 achieved an efficiency rate above 80% for the test set of the models. Finally, the utilization of Synchronous Fluorescence Spectroscopy associated with chemometric techniques enabled the analysis of two types of fuels without the need for any sample preparation, in a rapid and effective manner when compared to the methods established by the ANP.