A kernel function-based CFD-DEM simulation of fines retention at tight reservoirs

Clogging of fines in reservoir significantly declines the extraction efficiency. Microscale mechanisms and development process of fines clogging and the pressure drop evolution at tight reservoir remain unclear, partially due to the limitation of meshing-particle size-pore throat size. A kernel function-based CFD-DEM method was implemented to address this limitation, with fines type, flow velocity, flow field pattern, and adhesive force as parameters of interest. For non-buoyant fines (i.e., silt), bridging was the main clogging mechanisms, and the decrement of cohesive force had limited effects in declining bridging probability in convergent radial flow condition due to the gradually increment of flow velocity along the radial direction. For buoyant fines (i.e., clay), clogging formed due to the gradual deposition or bridging after surface deposition formed, categorizing pressure drop into two stages with that increasing slightly (deposition accumulation) and dramatically (clogging formation). A two-stage pressure drop temporal evolution model considering the processes of deposition accumulation and stochastic clogging were constructed. The critical retention volume for forming clogging in convergent radial flow condition (0.68%) was smaller than that in unidirectional flow condition (0.88%). Moreover, convergent radial flow enhanced the circumferential motion of fines, and pre-clogged clogging promoted clogging phenomenon was found in this scenario.