Identification of petroleum profiles by infrared spectroscopy and chemometrics

• Published in: Fuel (Guildford) 2019-10, Vol.254, p.115670, Article 115670
• Main Authors: Lovatti, Betina P.O., Silva, Samantha R.C., Portela, Natália de A., Sad, Cristina M.S., Rainha, Karla P., Rocha, Julia T.C., Romão, Wanderson, Castro, Eustáquio V.R., Filgueiras, Paulo R.
• Format: Article
• Language: English
• Subjects: Chemometrics; Crude oil; Decision making; FTIR; Infrared spectroscopy; kNN; Model testing; Oil; Oil exploration; PCA; Petroleum; Petroleum industry; Physicochemical properties; Properties (attributes); Refining; Spectroscopy; Spectrum analysis; Transportation
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2019.115670

[Display omitted] •Chemometrics methods used to verify the similarity between petroleum samples.•PCA was used to reduce the dimensionality of the data.•From FTIR spectra it is possible to determine the profile of petroleum.•FTIR spectra reach the same conclusion as physical-chemical property assay. The petroleum industry has the interest to know the physicochemical properties of the crude oil before its exploration in order to estimate its costs with processing, transportation, storage and refining. The profile of the recovered oil is determined by the analysis of some physicochemical properties, which may require a long time or a large volume of the sample, delaying decision-making on production and refining. Spectroscopic techniques allow the acquisition of a large amount of information about the sample at a molecular level, in a short time, and with a small amount of sample. This paper aims to identify petroleum profiles using a spectroscopy technique in the middle infrared region associated with methods of standards recognition PCA and kNN. We used 81 oils, divided into two groups: 69 samples of the training set (model construction) and 12 samples of the test set (model performance verification). For all samples, MIR spectra were acquired and measurements of 13 physicochemical properties were obtained. The results show that the methodology using MIR spectroscopy with chemometrics makes it possible to obtain equivalent conclusions about the most similar crude oils, that is, with similar behaviors, when we analyze a database with 13 physicochemical properties of the crude oils.