Abstract:

Unconverted oil from ebullated bed hydrogenation (UCO) has significant advantages as coking feedstock for preparation of low sulfur petroleum coke. The model was established to forecast the coking product yield and quality of UCO based on multiple stepwise regression method, and the differences of coke rule between UCO and residue were analyzed. The results indicated that the yields of coke and liquid product were single variable linear function of CCR content of UCO in the delayed coking process of UCO. Meanwhile, the sulfur content of liquid product was the single variable linear function of sulfur content of UCO. However, there had bivariate function between the sulfur content of coke and properties of UCO, and the two variables were sulfur and CCR content of UCO. The model prediction effect was good, and the average relative error was less than 2.0%. Compared with coking rule of residue, the sulfur content of coke in UCO coking process was relatively higher, basically remaining at around 60%, and coking tendency of UCO was higher. Based on the results of multiple stepwise regression, and using linear programming method, the linear programming diagram was drawn to optimize the UCO properties for preparation of low sulfur petroleum coke, which could be used to attain value range of UCO properties under the different demand for petroleum coke yield and sulfur content of coke.

Key words: multiple stepwise regression, ebullated bed hydrogenation, unconverted oil, low sulfur petroleum coke, coking rule

摘要：

沸腾床加氢未转化油（UCO）作为焦化原料制备低硫石油焦具有显著优势，本文采用多元逐步回归法建立UCO焦化产物收率和性质的预测模型，剖析UCO和渣油焦化规律差异。结果表明，UCO焦化过程中，焦炭收率和液体产品收率可表示成残炭（CCR）的单变量线性函数，液体产品硫含量为UCO中硫含量的单变量线性函数，而焦炭中硫含量为UCO原料硫含量和残炭的双变量函数。模型预测效果良好，各项指标相对误差的平均误差小于2.0%。与渣油焦化相比，UCO焦化过程焦炭中硫占比较高，基本保持在60%左右，同时UCO生焦倾向更高。基于多元逐步回归结果，采用线性规划方法，绘制出低硫石油焦制备过程UCO性质优选线性规划图，可推出不同石油焦收率和硫含量下的UCO关键性质的取值范围。

关键词: 多元逐步回归法, 沸腾床加氢, 未转化油, 低硫石油焦, 生焦规律