Mathematical Model of Virtual Quality Analyzer for Selective Petroleum Oil Refining

Petroleum oil without undesirable impurities, so-called refined petroleum, is the basic resource for the production of a wide range of consumer lubrication oils. The undesirable impurities to be re-moved include resinous substances and polycyclic aromatic hydrocarbons. Selective oil refining is a widely used raw oil purification method that involves the extraction of impurities with a selective sol-vent. The efficiency of this process depends on the extraction temperature and extraction agent con-centration for a specific type of raw material. The instability of its hydrocarbon composition along with the lack of constant target quality control for refined petroleum is one of the key problems. In this context, selecting and maintaining optimal production process parameters become difficult to achieve. This article presents a mathematical model of a virtual analyzer to control the quality of refined petroleum. The refraction index was selected as a quality indicator for the finished product. The model was based on the results of pressure, temperature, and flow measurements at specific production setup points, and it can assess the quality of the produced refined petroleum in real time to generate prompt control responses. The comparison of the model data obtained on an operating setup and the laboratory analysis data showed that the error of refraction index assessment does not exceed 1%.