Abstract: Due to the complexity of feedstock, it is challenging to build a general model for light olefins production. This work was intended to simulate the formation of ethylene, propene and 1,3-butadiene in alkanes pyrolysis by referring the effects of normal/cyclo-structures. First, the pyrolysis of n-pentane, n-hexane, n-heptane, n-octane, n-nonane, n-decane, cyclohexane, methylcyclohexane, n-hexane and cyclohexane mixtures, and n-heptane and methylcyclohexane mixtures were carried out at 650-800℃, and a particular attention was paid to the measurement of ethylene, propene and 1,3-butadiene. Then, pseudo-first order kinetics was taken to characterize the pyrolysis process, and the effects of feedstock composition were studied. It was found that chain length and cyclo-alkane content can be qualitatively and quantitively represented by carbon atom number and pseudo-cyclohexane content, which made a significant difference on light olefins formation. Furthermore, the inverse proportional/quadratic function, linear function and exponential function were proposed to simulate the effects of chain length, cycloalkane content and reaction temperature on light olefins formation, respectively. Although the obtained empirical model well reproduced feedstock conversion, ethylene yield and propene yield in normal/cyclo-alkanes pyrolysis, it exhibited limitations in simulating 1,3-butadiene formation. Finally, the accuracy and flexibility of the present model was validated by predicting light olefins formation in the pyrolysis of multiple hydrocarbon mixtures. The prediction data well agreed with the experiment data for feedstock conversion, ethylene yield and propene yield, and overall characterized the changing trend of 1,3-butadiene yield along with reaction temperature, indicating that the present model could basically reflect light olefins production in the pyrolysis process even for complex feedstock.

Key words: Empirical model, Normal-alkane, Cyclo-alkane, Pyrolysis, Light olefins

摘要： Due to the complexity of feedstock, it is challenging to build a general model for light olefins production. This work was intended to simulate the formation of ethylene, propene and 1,3-butadiene in alkanes pyrolysis by referring the effects of normal/cyclo-structures. First, the pyrolysis of n-pentane, n-hexane, n-heptane, n-octane, n-nonane, n-decane, cyclohexane, methylcyclohexane, n-hexane and cyclohexane mixtures, and n-heptane and methylcyclohexane mixtures were carried out at 650-800℃, and a particular attention was paid to the measurement of ethylene, propene and 1,3-butadiene. Then, pseudo-first order kinetics was taken to characterize the pyrolysis process, and the effects of feedstock composition were studied. It was found that chain length and cyclo-alkane content can be qualitatively and quantitively represented by carbon atom number and pseudo-cyclohexane content, which made a significant difference on light olefins formation. Furthermore, the inverse proportional/quadratic function, linear function and exponential function were proposed to simulate the effects of chain length, cycloalkane content and reaction temperature on light olefins formation, respectively. Although the obtained empirical model well reproduced feedstock conversion, ethylene yield and propene yield in normal/cyclo-alkanes pyrolysis, it exhibited limitations in simulating 1,3-butadiene formation. Finally, the accuracy and flexibility of the present model was validated by predicting light olefins formation in the pyrolysis of multiple hydrocarbon mixtures. The prediction data well agreed with the experiment data for feedstock conversion, ethylene yield and propene yield, and overall characterized the changing trend of 1,3-butadiene yield along with reaction temperature, indicating that the present model could basically reflect light olefins production in the pyrolysis process even for complex feedstock.

关键词: Empirical model, Normal-alkane, Cyclo-alkane, Pyrolysis, Light olefins