油品分馏塔产品闪点模型

Modeling for Flash Point Data of the Distillate from Crude Oil Tower

采用某炼厂的实际操作数据,建立了闪点与操作参数关联的数学模型,建立了闪点与20℃时油品的相对密度和常压恩氏蒸馏数据相关联的数学模型,并对闪点与操作参数相关联的预测模型进行了验证。建立的常压塔产品闪点预测模型,含有6～8个参数、8～14个系数,方程形式简单,在预测计算中没有任何迭代计算,也不用求最优点,更不用进行分馏塔的复杂严格模拟计算,只需要很少数据,就可以进行快速计算和预测。对于常压塔,模型回归的算术平均误差为2.496℃和3.238℃,模型的复相关系数分别为0.9534和0.9758。

Operating data of atmospheric column obtained from a certain refinery were used to establish predictive modeling between flash point and operating data, which also were used to establish modeling among flash point, ASTM-D86 data and standard relative density. It was proved that the predicting modeling between flash point and operating data can agree well with the data. There were 6-8 parameters and 8-14 coefficients in the model equation. The prediction was rapid and simple because the complex iterations, the calculation of the optimum point and the rigorous simulations were avoided. For the atmosphere column, the equation＇s average regressive errors were 2.496℃ and 3.444℃, and the correlation coefficients were 0.9534 and 0.9758. resoectively.