No abstract

Modeling selectivity of ethylene and propylene generated by Fe‐based catalyst in a fixed‐bed reactor based on the pressure and ratio of H2/CO in the range of 0.8‐4 MPa and 0.25‐4, has been studied. The neural network was used for network training of experimental data to determine the selectivity model. Then, response surface methodology was applied to determine the cubic polynomial models of selectivity of ethylene and propylene. The effect of various operating condition on the selectivity of products was investigated with statistical approaches. Analysis of variance (ANOVA) for modeling selectivity of propylene and ethylene indicates that mathematical models are significant.