Study On Coking Tendency And Products Distribution Of Heavy Oil Thermal Reaction

In recent years, the thermal processing techniques has turned into the primary treatment of upgrading the heavy oils. However, the refineries have encountered many difficulties in the practical thermal processing, because of lacking basic data, which would be available for design and estimation. Therefore, it is important to research the trendency of coking in thermal reaction and the distribution of products, providing reference for the optimizing of the industrial thermal process and the specific operating condition.The vacuum residues of W and K were used as experimental feeds. First of all, the main properties of residua have been analyzed. After that, the coking tendency of the two feeds at light thermal conversion degree as well as the affection of reaction temperature and feestocks'properties was studied through autoclave experiments. Furthermore, the coking of heavy oils has been deeply studied from the perspective of colloidal stability. The results demonstrated that the higher the reaction temperature was, the shorter the coke-induction period was and the faster the coking rate was. The yield of coke was higher under the same reaction time as well. In addition, the coke-induction period was closely related to the factors including SARA content of heavy oil system, H/C ratio and condensation index. The coke-induction period of WVR was shorter than that of KVR and the coking rate behavior was contrary. It was also manifested that within the coke-induction period the colloidal stability of two feeds decreased as the reaction time increased without apparent coke formation; the coking went fast and the colloidal stability index fell and became stable. The starting point of becoming stable was corresponding to the point at which the massive coking began. Thus the coking of heavy oil could be effectively delayed through the promotion of colloidal stability.In addition, the data of thermal reaction product distribution of heavy oil under different conditions were collected and the kinetic model with 6 lumps for thermal reaction of heavy oil was built up on the basis of practicability and analysis circumstance available in the laboratory. The classical forth-order Runge-Kutta was applied as the method for the modeling formulation. The Marquart was used as the approach of fitting. The related kinetic parameters were solved. At the end of research, the model proved to be reasonable from the perspective of statistical test and the yield of products including the tendency of product selectivity with the operation condition was predicted based on the aforementioned model.