Number of somatic cells (white blood cells) in bovine milk matrix is one of the defining characteristics of milk quality. An increasing Somatic Cell Count (SCC) in milk negatively effects the texture, color and taste of the substance, reducing its shelf life, hence the commercial value. With somatic cells being considered as sparse particles in a continuous medium, disk-stack centrifuges have become the most common industrial application of reducing and controlling the SCC index. In this study, a Computational Fluid Dynamics-Discrete Element Modeling (CFD-DEM) mainframe on Euler-Lagrange coupling basis is developed to determine the separation ability of a clarifying disk-stack centrifuge. The 3D model is simplified from an actual machine while strictly adopting the geometrical features of the disc-stack. The CFD-DEM model yields a 3.76% relative error in separation ability at average particle diameter, when compared to field tests of the machine at steady operation. Simulations show an increasing g-force, combined with reduced feed rates yields the best separation ability for the process. With optimum process parameters generating a 91.05% sedimentation ratio, the upper limit of g-force at 10 ton/h feed rate offers a 95% separation efficiency, while the lower limit of feed rate offers 96% at 7,700rpm.