PLS regression on spectroscopic data for the prediction of crude oil quality: API gravity and aliphatic/aromatic ratio

This work describes a chemometric approach for predicting quality parameters of crude oils by using the information present in spectroscopic data as Fourier Transform Infrared-Attenuated Total Reflectance (FTIR-ATR) absorption and Synchrounous Ultra Violet Fluorescence (SUVF). Using multivariate analysis such as Partial Least-Square (PLS) analysis, the predictive ability of spectroscopic techniques has been explored to estimate the American Petroleum Industry (API) gravity usually determined using standard physical methods and infrared structural/functional indices characterizing the repartition of aliphatic and aromatic structures present in crude oils. Giving global information on chemical compounds present in oil, FTIR-ATR also appears to be a rapid analytical method for quantifying changes in abundances of aliphatic and aromatic structures with the help of the infrared indices calculated from area ratio of specific bands. Then, a PLS model based on MIR data allows to predict this aliphatic/aromatic ratio for various crude oils and avoid time-consuming step of infrared peaks integration. (C) 2012 Elsevier Ltd. All rights reserved.