论文部分内容阅读
在实测大庆、胜利馏分油常压气-液平衡常数 K 及相应物性基础上,回归提出六个预测常压 K 的关联式,其误差分别为4.65%,5.34%,7.52%,7.55%,9.85%和9.50%,各式仅需要体系温度、压力、馏分平均沸点、馏分宽度及密度中的部分作为回归参数,具有一定实用意义。用实测数据分别考察了 CS 模型、GS 关联、修正的 CS 模型、SRK 方程,修正的 SRK 方程,PR 方程及四种按理想体系计算的十种气-液平衡常数计算方法对国产石油馏分常压K 预测的适用性。结果表明 Pitzer-Reidel 式和 Góm(?)z-Nieto-Thodos 式最准确,但所有误差均大于本文提出的关联式。
On the basis of the experimental data of Daqing and Shengli distillate at constant pressure, the equilibrium constants of K and their corresponding physical properties, the regression equations of six predicted atmospheric pressure K are obtained by regression. The errors are 4.65%, 5.34%, 7.52%, 7.55% and 9.85 % And 9.50%, all kinds of only the system temperature, pressure, the average boiling point of the fraction, the width of the fraction and the density of the part as a regression parameter, has some practical significance. The CS model, GS correlation, modified CS model, SRK equation, modified SRK equation, PR equation and four methods of calculating the gas-liquid equilibrium constants calculated by the ideal system were respectively investigated with measured data. K predictability of applicability. The results show that the Pitzer-Reidel and Góm (?) Z-Nieto-Thodos formulas are the most accurate, but all the errors are larger than the correlation proposed in this paper.