Simultaneous determination of hydrocarbon renewable diesel, biodiesel and petroleum diesel contents in diesel fuel blends using near infrared (NIR) spectroscopy and chemometrics

Abstract

Highly polluting fuels based on non-renewable resources such as fossil fuels need to be replaced with potentially less polluting renewable fuels derived from vegetable or animal biomass, these so-called biofuels, are a reality nowadays and many countries have started the challenge of increasing the use of different types of biofuels, such as ethanol and biodiesel (fatty acid alkyl esters), often mixed with petroleum derivatives, such as gasoline and diesel, respectively. The quantitative determination of these fuel blends using simple, fast and low cost methods based on near infrared (NIR) spectroscopy combined with chemometric methods has been reported. However, advanced biofuels based on a mixture of hydrocarbons or a single hydrocarbon molecule, such as farnesane (2,6,10-trimethyldodecane), a hydrocarbon renewable diesel, can also be used in mixtures with biodiesel and petroleum diesel fuel and the use of NIR spectroscopy for the quantitative determination of a ternary fuel blend of these two hydrocarbon-based fuels and biodiesel can be a useful tool for quality control. This work presents a development of an analytical method for the quantitative determination of hydrocarbon renewable diesel (farnesane), biodiesel and petroleum diesel fuel blends using NIR spectroscopy combined with chemometric methods, such as partial least squares (PLS) and support vector machines (SVM). This development leads to a more accurate, simpler, faster and cheaper method when compared to the standard reference method ASTM D6866 and with the main advantage of providing the individual quantification of two different biofuels in a mixture with petroleum diesel fuel. Using the developed PLS model the three fuel blend components were determined simultaneously with values of root mean square error of prediction (RMSEP) of 0.25%, 0.19% and 0.38% for hydrocarbon renewable diesel, biodiesel and petroleum diesel, respectively, the values obtained were in agreement with those suggested by reference methods for the determination of renewable fuels.

Monitoring photolysis and (solar photo)-Fenton of enrofloxacin by a methodology involving EEM-PARAFAC and bioassays: Role of pH and water matrix

“…Hence, implementation of a simple methodology to obtain major trends on by-products formation is needed. In this context, chemometric tools are gaining momentum, as they can extract valuable information from raw complex data (Alves and Poppi, 2013;Boughattas et al, 2019;Karoui and De Baerdemaeker, 2007;Trawiński and Skibiński, 2018). In this sense, and taking advantage of the photophysical properties of FQ-type molecules, fluorescence excitation-emission matrices (EEMs) with the multi-way method, parallel factor analysis (PARAFAC), have been selected to study the ENR degradation.…”

Evaluation of single and multi-feedstock biodiesel – diesel blends using GCMS and chemometric methods

“…The concentration of biodiesel in blended fuels has been measured using various spectroscopic techniques, including infrared [3,4,5,6,7] and fluorescence [8,9] spectroscopy combined with multivariate analysis methods, such as Hierarchical Cluster Analysis (HCA), Principal Component Analysis (PCA), or Partial Least Squares (PLS). Electrospray ionization mass spectrometry has been combined with chemometric methods to analyze different feedstocks and blend compositions of biodiesel [10,11].…”

Simultaneous determination of multiple components in explosives using ultraviolet spectrophotometry and a partial least squares method

“…43,57,58 In general, a residual (3 i ) is dened as the difference between an experimental observation and a predicted value from a regression model, 3 i ¼ y act,i À y pred,i where y act,i is the real value and y pred,i is the predicted value by the regression model. 43,57,58 In general, a residual (3 i ) is dened as the difference between an experimental observation and a predicted value from a regression model, 3 i ¼ y act,i À y pred,i where y act,i is the real value and y pred,i is the predicted value by the regression model.…”

“…In order to detect whether there were outliers in the calibration and independent test samples, residual analysis 43,57,58 was executed in the PLS regression models. There are a total of 211 wavelengths in a spectrum.…”

“…38 Multivariate calibration methods in chemometrics play an important role in multicomponent resolution and quantication 34,[39][40][41][42][43] and have been successfully used to solve problems which exist in the spectral data of complicated mixtures, such those that involve collinearity, band overlaps and interactions. 38 Multivariate calibration methods in chemometrics play an important role in multicomponent resolution and quantication 34,[39][40][41][42][43] and have been successfully used to solve problems which exist in the spectral data of complicated mixtures, such those that involve collinearity, band overlaps and interactions.…”