Lumped Kinetic Model for the Hydrodeoxygenation of Full-Range Middle-Low Temperature Coal Tar

Wang Yijie,Yanchang Petroleum Co. Ltd.,Tian Feng,Dong Huan,Fan Xiaoyong,Du Chongpeng,Cui Louwei,Zheng Huaan,Teng Haipeng
DOI: https://doi.org/10.1007/s11144-021-02053-1
2021-01-01
Abstract:Based on the experimental data of full-range middle-low temperature coal tar (LTCT) hydrodeoxidation (HDO), the phenolic compounds (PC) in feedstock were divided into three lumps, four lumps or five lumps according to different reaction activities. Three corresponding lumping kinetic models of LTCT HDO were established. The Levenberg–Marguardt algorithm was used to calculate the kinetic parameters of each model. The fitting results of the three models are in good agreement with the experimental data, as the average relative error is less than 1.01%. Three types of lumped dynamic models are compared in detail, and the five-lumped model has an average relative error of less than 0.91%. Hence, this model can reliably predict PC removal and showed a good extrapolation performance. The effects of liquid hourly space velocity (LHSV), initial hydrogen pressure (P) and reaction temperature (T) on the HDO process of LTCT were analyzed by five-lumped model calculation, and the reaction law of HDO was obtained. The results show that the order of influence of various factors on the oxygen content of hydrogenated oil is LHSV > P ≥ T. Moreover, the hydrogenation process conditions were optimized by simulation: T = 663 K, P = 14 MPa, LHSV = 0.3 h−1 and V(H2)/V(Oil) of 1100:1, under which the expected oxygen content was 468.28 μg g−1. The results of this study can also provide a scientific understanding of HDO mechanism, greatly reduce the number of LTCT industrial tests, reduce the oil inspection frequency, and help to improve the industrial hydrogenation process.
What problem does this paper attempt to address?