No abstract

A kinetic model of steam‐cracking in industrial‐scale steam crackers was developed by a combination of mechanistic and empirical approach. Kinetic parameters of the model were optimized using the data of process experiments, carried out in full‐scale industrial crackers, the product yields being determined by hot‐gas sampling for the usual types of feedstock, including ethane, liquefied petroleum gas, gasoline, atmospheric gas oil, and hydrocracked vacuum distillate. Experimental and simulated results were compared for a wide range of feedstock and reaction conditions. Data on cracker control parameters and feedstock composition were collected by monitoring the whole production unit consisting of ten crackers and simulation results of total production compared to those of process balance.

A computer program for the automated generation of steam-cracking reactions network (RNG) was developed. The RNG algorithms are based on graph theory tools. Kinetic parameters of generated reaction steps are derived using the information about the structure in the neighborhood of the reaction center by the developed group contribution method. Total radical activity in a reaction mixture is assumed to be stationary along the reactor. A method for estimating this activity in dependence on reactant structure and reaction conditions was also developed. Group contribution parameters were optimized using the results of laboratory experiments involving several pure hydrocarbons (n-alkanes, i-alkanes, cycloalkanes). The laboratory pyrolysis reactor was close to isobaric with plug-flow and specified temperature profile. Results of optimized model were in good agreement with experimental data obtained on a laboratory pyrolysis reactor.