提出了一套采用虚拟组分法来估算润滑油糠醛精制过程中的物理性质和液液平衡的方法。虚拟组分的物性参数由物性实验数据拟合得到,这些参数可满意描述润滑油及其分离物的密度、比重、折光率及含硫量,且可推广应用到其它平均沸点下的润滑油。利用物性估算得到虚拟组分的分子量,液液平衡数据的质量分数可转化为摩尔分数,进而可用NRTL活度系数模型回归计算,最终可应用于工业实际中的润滑油糠醛精制抽提塔的模拟。该估算方法与文献结果相比,在关联的结果上具有一定优势,对于密度、比重和折光率而言,该估算方法得到平均绝对偏差均小于文献偏差的0.5倍,而对于预测结果,两种方法的偏差基本相当。另外,按此方法估算得到的虚拟组分性质更加合理可信,可为模拟工作提供可靠的基础数据。
A pseudo-component method was used to estimate physical properties and liquid-liquid equilibria of lubricating oil refining processes with furfural. Physical properties including density, specific gravity, refractive index and sulfur content can be well estimated with the parameters of the pseudo-components. Parameters were obtained by fitting experimental data, and this method can be extended to other lubricating oils with different average boiling points. The molecular weights of the pseudo-components could be calculated using their properties, and mass fractions from liquid-liquid equilibrium data were converted into mole fractions These data were correlated with NRTL model, and this method was applied to simulate an extraction column for lubricating oil refining with furfural. Compared with calculation methods reported in the literature, this method has advantages in correlation results obtained, especially for density, specific gravity and refractive index, where the average absolute deviations (AADs) in this work are less than half of those reported in the literature. This method has comparable result prediction accuracy as those reported. Furthermore, this proposed method can produce more reasonable pseudo-component properties, which can provide fundamental date for subsequent simulation.