关键词: 在线近红外光谱分析, 吸附脱硫, 汽油, 辛烷值, 密度, 馏程, 蒸气压, 烃组成

Abstract: In order to apply the on-line near infrared spectroscopy (NIR) analysis technology to the on-line analysis of gasoline adsorptive desulfurization (S Zorb) unit, based on a large number of representative S Zorb unit feedstock gasoline and product samples, their near-infrared spectra were collected, and a near-infrared spectral analysis model was developed to predicting the material density, RON, vapor pressure, boiling range and hydrocarbon types composition of gasoline of S Zorb adsorptive desulfurization unit using the chemometric method. There is good consistency between the near infrared spectroscopy analysis method and standard method. Based on on-line FT-NIR analysis system and application of material multi-properties NIR analysis model of S Zorb, the on-line NIR technology was applied in S Zorb unit to perform real-time analysis of density, vapor pressure, RON, boiling range and hydrocarbon type composition of gasoline. The on-line analysis of 65 properties, including the density, octane number, vapor pressure, distillation range and hydrocarbon composition of 3 streams of the unit, can be completed within 1 min. The standard deviations of S Zorb desulfurization gasoline density (20 ℃), RON, olefine content, aromatic content, vapor pressure are 1.5 kg/m³, 0.2, 0.46 percentage point, 0.51percentage point, 1.5 kPa, respectively. The predicted standard deviations of initial boiling point, 10% distillation temperature, 50% distillation temperature, 90% distillation temperature and final boiling point are 2.0, 1.7, 2.1, 2.3, 2.8 ℃, respectively. The accuracy and detection speed of the on-line NIR method can meet the needs of rapid analysis in the industrial process.

Key words: on-line near infrared spectrum analysis, S Zorb, gasoline, octane number, density, boiling range, vapour pressure, hydrocarbon types composition