Computational Fluid Dynamics (CFD) is the science of predicting momentum mass and heat transport by numerically solving a set of fundamental governing equations in one, two and three dimensions. Because of the computational intensity associated with CFD simulations, application of the code has historically been limited to the academic environment. However, over the past two decades, the cost of computing power has decreased substantially while the processing speed has increased exponentially. These developments have now made the application of CFD in the commercial environment feasible. Many commercial software models are available for CFD modelling. These include, but are not limited to Fluent, CFX and STAR-CD and have different degrees of sophistication in terms of graphical user interfaces and numerical solution techniques.
In this study the commercial CFD code, Fluent, was applied in the design of three pump stations. Pump stations are usually designed using the heuristic method, but when the application of these methods becomes vague it may be necessary to employ CFD to reduce uncertainty in predicting performance efficiency. Typical heuristic design limits may include prediction of flow patterns on approach to the pump intakes, the minimum drawdown levels for safe operation, extent of air entrainment and the formation of potential dead spaces. Simulated outputs from the case studies showed that the CFD code provided a functional estimation of free surface and submerged vortex formation, excessive pre-swirl of flow entering the pump, non-uniform distribution of velocity on the impeller eye and entrainment of air bubbles which was then used during the design process to reduce operating costs and risk as well as improve efficiency.