Mathematical Modeling for the Delayed Coking Unit Using Molecular Information

The Delayed Coking Unit is a thermal conversion process, known in the petroleum industry as “bottom of the barrel”. This unit pays a key role in the conversion of vacuum residue in products of higher added value, providing the refining scheme with an increase in profitability. Market competition, restrictions on product specifications and operational bottlenecks require better control of production planning. Therefore, developing new strategies and mathematical models focused on operational optimization of industrial processes and formulations of products is essential to achieve better yields and a more accurate product quality monitoring.

The difficulty in obtaining modeling and simulation in good adherence to the industrial unit, which is highly empirical, is in the data quality and degree of detail of the models used.This work is based on the experimental analysis and mathematical modeling, which aims to discuss the feedstock characterization, highlighting the use of mass spectrometry for asphaltenes and evaluation of green coke quality. From more detailed information about the molecular characteristics of the feedstock it is possible, through correlations with bulk properties, commonly measurable, to develop a mathematical model for the process in order to optimize the operating variables, to maximize the conversion with lighter products besides producing coke with the desired quality. The model is based also in a lumped kinetic scheme and vapor-liquid equilibrium conditions in the furnace and coke drum. The thermal cracking is consequently well predicted.