Genetic Algorithm optimization of wind turbine shrouds

The efficiency of Horizontal Axis Wind Turbines (HAWTs) can be improved by surrounding the turbine with a shroud. Both computational and experimental studies have shown that turbines with shrouds can exceed the theoretical Betz power conversion limit. We make use of a genetic algorithm to find optimal shroud shapes subject to constraints related to shroud size. The efficiency of the turbine-shroud system is measured via finite element simulations (using the Code-Saturne CFD package), with shroud geometries and meshes generated programmatically. We conduct full 3D simulations, but use the Actuator Disk theory model to simplify the computations.