Por meio do método lattice Boltzmann (LBM), este projeto de pesquisa propõe a simulação computacional da extração dos compostos diterpênicos do café verde em sistemas com fluido pressurizado. Para isso foi adotado um modelo fenomenológico dinâmico e de primeira ordem quanto à dependência espacial. Usando a rede D1Q2, o LBM foi implementado com duas funções de distribuição de partículas, acompanhadas das respectivas funções de distribuição de equilíbrio. Cada função refere-se à concentração em uma dada fase Using the lattice Boltzmann method (LBM), this research project proposes a computer simulation of diterpene compounds extraction from green coffee beans in high pressure fluid systems. In order to do so, a dynamic phenomenological model with first-order spatial dependence was adopted. LBM was implemented on D1Q2 lattice, using two particle distribution functions, followed by the respective equilibrium distribution functions. Each function refers to the extract concentration in each phase, namely solid and fluid. Prediction of extract yield curves relies on process parameters such as internal (i.e., intraparticle) mass diffusivity, volumetric partition coefficient and axial mass diffusivity. Values of aforesaid parameters were fitted against experimental data via descendent gradient technique coupled to the LBM computer code and numerically simulated result were consistent. Besides modeling and simulation, this project performed the influence analysis of those process parameters. To this end, a user-friendly graphical interface for the original computer code was created. While fostering value to be added to coffee, this research project develops skills in modeling and simulation of chemical species extraction through numerical methods.