As tar sand and low-yield field production becomes more prevalent, performance of equipment used to separate water, gas, and solids from crude oil becomes critical to the profitability of production operations. As computational hardware and multiphase modeling have improved, the commitment of industrial engineers to virtual prototyping has made computational analysis more prevalent in the design and optimization of separation equipment.

One such piece of equipment is the hydrocyclone. In this study, the coupling of CFD to optimization algorithms for a high level of design automation has been demonstrated. Three design parameters are allowed to vary, these being the inlet diameter, the vortex finder diameter, and the vortex finder length. The specified goal is to achieve a minimum D50, the particle size for which half of the particles entering at the inlet are captured. A Design of Experiments tool chooses appropriate conditions to test with CFD. A response surface is then generated and used by the optimization algorithm to determine the optimum design parameters.