Desenvolvimento de uma nova metodologia para determinar a densidade API em destilados de petróleos obtidos a partir de unidades de destilação manual e automática
Data
Título da Revista
ISSN da Revista
Título de Volume
Editor
Resumo
Density is one of the most important physical-chemical properties regarding the characterization of petroleum and its derivatives. Tightly related to this property, the True Boiling Point (TBP) curve is appliedas one of the parameters for economic purposes, such as price assessment. Nowadays, the TBP curve is generally obtained through automatic distillation in laboratorial scale, in which multiple distillation fractions of the process are characterized throughprimary methods. In certain cases, this automatic systemare not the best optionto process, once those automatic units do not allow interruptions during the distillation, as well as not coupling glassware and other instruments for evaluation and optimization of the process. In this context, a new methodology was performed in this work to evaluate the association of samples of fractions obtained through manual and automatic laboratorial distillation, based on ASTM D 2892, to estimate the API gravity.In order to give data to support a future implementation of anin locodensity online reading systemduring distillation.Firstly, it was performed distillations in the manual units to verify their behavior through references TBP curves obtained by automatic distillation of the same crude oils. Subsequently, models based on Near-infrared and Mid-infrared Spectroscopy (NIR and MIR) data was developed by using Partial Least Squares Regression (PLS Regression) in order to test the viability of estimating the API gravity of the samples of fractions from manual distillation through a database of samples previously obtained from automatic distillation. Hence, aiming a small number of distillations necessary for calibration, thus diminishing the usage of reagents, as well as increasing the results reliability by using data from a certified automatic unit. The models of the multivariate calibration developed for the group of samples for both techniques of infrared showed satisfactory results accords of specifications applied in the petroleum industry
