CFD-based parametric study of venturi-shaped roof optimization for wind energy harvesting by building-integrated wind turbines in an idealized high-rise building

The Venturi-shaped roof creates advantageous wind conditions for building-integrated wind turbines (BIWTs) to harvest wind energy in urban areas. However, no study has optimized Venturi-shaped roof designs to maximize wind energy harvesting by BIWTs. This study developed an optimization framework for a Venturi-shaped roof design by integrating global sensitivity analysis, CFD simulations, surrogate modeling, and multi-objective optimization under parametric constraints. The global sensitivity analysis revealed that the length of the roof supports and the fillet radius of rounded corners have the largest effects on wind velocity and turbulence intensity (TI) in the tunnel. A surrogate model was developed based on an Artificial Neural Network (ANN) model to predict wind speed and TI. The Non-Dominated Sorting Genetic Algorithm II was utilized for multi-objective optimization of six design parameters, including roof support structures, tunnel shape, and rounded corners, thereby maximizing wind speeds and minimizing TI in the roof. Three optimized Venturi-shaped roof designs were compared with the basic roof without optimization. The results showed that, compared with the basic roof, the wind acceleration area in the optimized roof tunnel increased by 1.35 times, and the area with TI > 15% decreased by about 35%. As a result, the optimized roofs produced 1.5 to 1.8 times more wind energy than the basic roof. Our study confirmed that the optimized Venturi-shaped roofs generate higher wind speeds and smaller TI in wind directions ranging from 0° to 45°, producing more wind energy and prolonging the service life of BIWTs.