Otimização de uma turbina eólica Darrieus H baseada em metodologias de superfícies de resposta e CFD

Abstract

The ongoing energy transition has resulted in an increased focus on renewable energy systems and their viability for Distributed Generation (DG) applications. In this context, small-scale wind turbines have drawn significant attention due to their compact and adaptable dimensions, making them well-suited for water pumping, irrigation, and energy generation in domestic settings, whether in urban or remote areas. In this way, the main objective of this study was to develop a reliable methodology for optimizing the aerodynamic performance of a small-scale Vertical Axis Wind Turbine (VAWT), specifically the Darrieus H type. To accomplish this, a flexible, accurate, and efficient Computational Fluid Dynamics (CFD) methodology was devised to predict the Power Coefficient (CP) of various configurations of Darrieus H VAWTs. This methodology was calibrated using experimental data from the literature and adapted to minimize the number of simulations related to spatial and temporal discretization independence studies. Furthermore, the CFD methodology was integrated into an optimization procedure based on Response Surfaces (RS), which was constructed using Design of Experiments (DoE) techniques and explored within their domains through Optimization Algorithms (OA) execution. The results of this study provided valuable insights into the configuration of a CFD approach for small-scale Darrieus H VAWTs, with a maximum percentage error of 12% in CP compared to experimental data. Additionally, the optimization procedure resulted in an 11% increase in nominal CP, where the typical CP value for small-scale Darrieus H VAWTs was 0.30, and the maximum percentage error of the CFD methodology was taken into consideration. The maximum value of 0.3778 for solidity (σ) was achieved when σ was equal to 0.24, and the installation angle (β) was equal to -4.1°. In conclusion, the integration of CFD simulation methodologies and RS-based optimization can provide reliable results in a relatively short time, thus serving as a powerful tool in the design of Darrieus H VAWTs..