Study of pore-scale coke combustion in porous media using lattice Boltzmann method

Abstract

In-situ combustion (ISC) for heavy oil recovery is a high-risk process and thus calls for a deep understanding of the coke combustion front. However, simulations for describing pore-scale coke combustion are challenging because of complicated combustion dynamics. This study proposes a multiple-relaxation-time (MRT) lattice Boltzmann (LB) method to simulate solid coke combustion in porous media at the pore scale. The MRT LB model can enhance numerical stability compared with the single-relaxation-time counterpart. This new LB model offers two major advances over existing LB models. First, thermal expansion effects are considered via temperature-dependent density, satisfying the low Mach number flow condition. Second, conjugate heat transfer and species conservation at the coke-fluid interface are modelled by a source term and a bounce-back scheme, respectively, without resorting to iterative methods. Simulations of pore-scale coke combustion are conducted to validate the superior performance of the developed LB model. The results show that this model can reproduce coke combustion dynamics and capture variations in isotherms from the fluid phase to the solid phase at the pore scale. In contrast, without considering effects of thermal expansion or conjugate heat transfer as in previous models, simulations would underestimate the burning temperature and even fail to predict instability caused by a high driving force. Moreover, using the proposed LB model, a parametric study is performed to assess the impact of the inlet air and porous structure on coke combustion. The findings suggest that inlet air temperature and driving force should be controlled within certain ranges, otherwise their influences may become negligible or even negative. Large porosity accelerates coke combustion and leads to a desired combustion front. Heterogeneous coke distribution affects pore-scale combustion details but not the overall combustion intensity, while a random matrix structure may influence combustion dynamics both locally and globally. As a whole, this study develops a reliable LB model to investigate pore-scale coke combustion, which contributes to advancing the knowledge base for heavy oil recovery using the ISC process.