Optimization of Savonius turbine clusters using an evolutionary based Genetic Algorithm

Abstract

Abstract VAWTs benefit from constructive coupling effect that significantly improves efficiency. This work aims at optimizing the configuration of turbine clusters with three Savonius turbines. To maximize the aggregate power output of a turbine cluster, two-dimensional (2D) computational fluid dynamics (CFD) is utilized, along with an evolutionary based Genetic Algorithm (GA). Four scenarios are investigated considering different combinations of rotational directions. For the optimal cluster obtained, a highest 37% increase in averaged power coefficient ( C P ¯ )over an isolated Savonius turbine is achieved at Tip Speed Ratio (TSR) = 0.8. The substantial increase in efficiency for turbines in the optimal cluster is essentially due to constructive coupling in the flow field. Two major effective coupling mechanisms are proposed based on flow structures to explain the observed performance increase. Furthermore, a method to expand the optimal cluster into compact wind farms is proposed. It is verified that for a compact Savonis-type VAWT wind farm that consists of 12 turbines, a 30% increase in C P ¯ is obtained, showing the practicality of constructing efficient VAWT wind farms in areas with limited land.