Hydrodeoxygenation of stearic acid to produce diesel-like hydrocarbons: kinetic modeling, parameter estimation and simulation

There has been an increasing search for renewable sources to produce alternative biofuels due to the growth in the global energy demand and the environmental impacts caused by fossil fuels. Green diesel obtained via hydrotreatment of vegetable oils and fats is a promising option, mainly due to the possibility of being integrated into an existing refinery. Here, we investigate the kinetics of hydrodeoxygenation of stearic acid as a model compound of triglyceride raw materials over NiMo/Al2O3 catalyst to obtain hydrocarbons in the diesel range. Kinetic models based on different assumptions to the adsorption step, such as Eley-Rideal and Langmuir-Hinshelwood mechanisms, were evaluated through nonlinear regression to parameter estimation using a hybrid algorithm combining Particle Swarm Optimization and Nelder- Mead method. All proposed models fitted well to the experimental data, presenting correlation coefficients higher than 90% for the fatty acid and the final hydrocarbons. However, the surface reaction limited model based on competitive adsorption with non-dissociate H2 adsorbed showed the best results and was used to simulate the process. (C) 2022 Elsevier Ltd. All rights reserved.

Automated summary

Due to the increasing global energy demand and the environmental impact of fossil fuels, there is a growing search for renewable sources to produce alternative biofuels. Green diesel obtained from vegetable oils and fats through hydrotreatment is a promising option that can be integrated into existing refineries. In this study, the kinetics of hydrodeoxygenation of stearic acid, a model compound for triglyceride raw materials, over a NiMo/Al2O3 catalyst was investigated to obtain diesel-range hydrocarbons. Different kinetic models based on assumptions about the adsorption step were evaluated, and the surface reaction limited model based on competitive adsorption with non-dissociate H2 adsorbed showed the best results and was used for process simulation.