Search

Mathematical model of n-heptane pyrolysis under thermal cracking and steam cracking conditions was developed. The objective was achieving good generalization by combining two different data sources for model identification. The data sources included experimental pyrolysis data for many structurally different hydrocarbons measured at reference conditions as well as the data for n-heptane measured in different reactors at different reaction conditions. The model was developed using the automated reaction network generation. The generated network included hydrogen abstraction, β-scission, radical isomerization and recombination reactions, the kinetic parameters of which were expressed as functions of 26 group contribution factors. Molecular reactions and radical additions were substituted by 12 formal shortcut molecular reactions. The initial values of kinetic parameters and group contribution factors were adopted from previous study, where they were obtained by regression based on the experimental data involving wide range of linear, branched and/or cyclic hydrocarbons. More experimental data on n-heptane pyrolysis measured under different conditions, including the use of two different reactors, were added to the data set used for the parameter optimization. The model simulations showed good agreement with the experimental data even for the validation samples. Thus, the proposed approach proved the ability to produce models achieving good generalization. [ABSTRACT FROM AUTHOR]

Abstract: Pyrolysis experiments aimed at measuring product yields and conversion of various hydrocarbons were carried out in a micro-pyrolysis laboratory reactor. Standard pyrolysis conditions were used for all samples—810°C, 400kPa and 0.2–0.4s residence time in the reaction zone. Analysis of products was performed by on-line GC/FID/TCD. All pyrolysis products from hydrogen to pyrolysis oil were detected either individually or by fractions. Detailed compositions of the pyrolysis products of 56 hydrocarbons were obtained and are included completely in the article. The study involved hydrocarbons, which represent the wide structural variability, including linear, branched, cyclic, aromatic and unsaturated. Measured data were used to identify important relationships between structural parameters and product yields. Trends of important product yields, depending on the carbon number, molecule saturation, substitution, ring presence and substitution or a substituent size are presented. [Copyright &y& Elsevier]

• Dependence of products yields on the blending ratio is non-linear. • Co-pyrolysis effects are responsible for this behavior. • Light naphtha donates active radicals to improve heavy naphtha cracking. Naphtha feedstock used for the steam-cracking process is typically blended from several hydrocarbon source streams of composition varying in a broad range. This study aims at evaluation of the potential yield improvement, which could be obtained by separate steam-cracking of light and heavy sub-streams using tailored process parameters for each of the sub-streams. Laboratory pyrolysis reactor was used for both separate and blend processing of the naphtha feedstocks in the temperature range of 760–810 C, and different blending ratios. At maximum temperature, pyrolysis of light naphtha provides ethylene yield 24.1 wt. %, while only 13.6 wt. % of ethylene was obtained from heavy naphtha. Both, ethylene and propylene yield express decreasing trend with increasing content of heavy naphtha in blended feedstock. Obtained results suggest possibility of mutual effects between components resulting in non-linear dependence of products yields on the blending ratio, therefore a principle of additivity cannot be assumed. A potential for improvement of ethylene yield by 1.1 wt. % is shown on example of processing scenario. [ABSTRACT FROM AUTHOR]