Controle de qualidade de biodieseis de pinhão manso e crambe usando espectrometria no infravermelho médio e cartas de controle multivariadas

Abstract

Biodiesel is a renewable source of fuel that can be used in pure form or in blends with diesel. In Brazil, from November 2014, the percentage of biodiesel in blends with diesel was set at 7% (v/v), through Law No. 13.033, the mixture being called B7. However, during the transport process, storage, distribution and marketing these mixtures may be adulterated by vegetable or mineral oils and other products, in order to forge or undesirably contaminate the product, making the product quality out of specification. Thus, to ensure the quality of biodiesel/diesel blends is necessary to develop methodologies that provide rapid and efficient responses to control the quality of this fuel. In this work was applied Mid Infrared Spectrometry (MIR) associated with Multivariate Control Charts based on Net Analyte Signal (NAS) to develop methodologies able to monitor compliant and non-compliant samples of Jatropha and Crambe biodiesels in mixtures with diesel. The biodiesels studied were produced from their vegetable oils using methyl and ethyl routes. In each model three charts were constructed: (i) NAS chart referring to the analyte of interest (biodiesel); (ii) Interference chart referring to the contribution of all other components in the sample (diesel) and (iii) Residual chart referring to any non-systematic variations in the spectra (instrumental noise). In calibration step 61 samples were used within the quality specifications, while the validation were used 665 samples ―in‖ and ―out‖ of specifications not used in the calibration step. This validation was done in relation to biodiesel content B7 blends and in relation to the presence of adulterants by: (i) partial substitution of biodiesel by soybean and waste frying oils; (ii) partial substitution of diesel by residual automotive lubricant oil and gasoline, and (iii) direct addition of each adulterant to B7 blends. All samples prepared within quality specifications were correctly classified as in control samples, because they were within the statistical limits of all the charts; and the samples prepared outside the quality specifications were classified as out of control samples, because they flag out of at least one of the charts limits. Therefore, the developed methodologies were efficient for the quality control of these types of biodiesels.