摘要
装置馏出口质量数据大多仍来自于每 4h或 8h一次的离线分析。对装置调整操作的滞后有时会造成产品质量不合格或质量的严重过剩。在线质量预测是根据确定的数学模型在装置计算机上运行来计算馏出口质量的一种技术 ,装置的原料及质量控制指标不变时与在线质量仪表作用相当。通过对抚顺石油二厂北蒸馏常二线的操作参数值与柴油凝点测定值进行回归分析 ,确定了操作参数与柴油凝点之间的数学模型 ,并对该模型进行了验证 ,预测误差不大于 2℃ ,在此基础上进行了工业应用 ,在线质量预测误差不大于 3℃ ,符合GB/T51 0有关再现性的规定。
The quality data of plant distillate mostly come from off-line analysis at intervals of 4 h or 8 h, sometimes, this lag to process operating can make products disqualified or excessive. On-line quality prediction is a kind of technology that the quality of distillate is calculated by a mathematical model in computer, it can take a same role as on-line analyzers when plant feedstock and the target of quality control are unchangeable. The relation model was built through regressive analysis between the operating parameters of atmospheric second line oil and its measured solid point of North Distillation Unit of Fushun No.2 Refinery, the error of this model was not greater than 2 ℃. When it was put into plant use, the model error of on-line running was not greater than 3 ℃, this result is coincident with the reproducibility of GB/T510.
出处
《抚顺石油学院学报》
2000年第4期36-39,共4页
Journal of Fushun Petroleum Institute
关键词
炼油
蒸馏
在线质量预测
操作参数
凝点
回归分析
On-line quality prediction
Operating parameter
Solid point
Regressive analysis
Mathematical mo(
