Andrés Guggeri

As the size of wind turbines increases and their hub heights become higher, which partially explains the vertiginous increase of wind power worldwide in the last decade, the interaction of wind turbines with the atmospheric boundary layer (ABL) and between each other is becoming more complex. There are different approaches to model and compute the aerodynamic loads, and hence the power production, on wind turbines subject to ABL inflow conditions ranging from the classical Blade Element Momentum (BEM) method to Computational Fluid Dynamic (CFD) approaches. Also, modern multi-MW wind turbines have a torque controller and a collective pitch controller to manage power output, particularly in maximizing power production or when it is required to down-regulate their production. In this work the results of a validated numerical method, based on a Large Eddy Simulation-Actuator Line Model framework, was applied to simulate a real 7.7 MNW onshore wind farm on Uruguay under different wind conditions, and hence operational situations are shown with the aim to assess the capability of this approach to model actual wind farm dynamics. A description of the implementation of these controllers in the CFD solver Caffa3d, presenting the methodology applied to obtain the controller parameters, is included. For validation, the simulation results were compared with 1 Hz data obtained from the Supervisory Control and Data Acquisition System of the wind farm, focusing on the temporal evolution of the following variables: Wind velocity, rotor angular speed, pitch angle, and electric power. In addition to this, simulations applying active power control at the wind turbine level are presented under different de-rate signals, both constant and time-varying, and were subject to different wind speed profiles and wind directions where there was interaction between wind turbines and their wakes.

Recently, the Actuator Line Model (ALM) has been evaluated with coarser resolution and larger time steps than what is generally recommended, taking into account an atmospheric sheared and turbulent inflow condition. The aim of the present paper is to continue these studies, assessing the capability of the ALM to represent the wind turbines' interactions in an onshore wind farm. The 'Libertad' wind farm, which consists of four 1.9MW Vestas V100 wind turbines, was simulated considering different wind directions, and the results were compared with the wind farm SCADA data, finding good agreement between them. A sensitivity analysis was performed to evaluate the influence of the spatial resolution, finding acceptable agreement, although some differences were found. It is believed that these differences are due to the characteristics of the different Atmospheric Boundary Layer (ABL) simulations taken as inflow condition (precursor simulations).

Uruguay belongs to the second most prone region of the world to the occurrence of severe convective storms, which also includes the south of Brazil, north-eastern Argentina and part of Paraguay; and to the region of South America that presents the highest frequency of occurrence of intense cyclogenesis events. Both meteorological processes produce high winds in the country. In spite of its strength, the last study at national level of high winds in the whole country had been conducted more than 30 years ago, during the development of the national wind code still in use. The installation along the last years of a network of automatic meteorological stations that measure continuously with a good temporal resolution, geographical distribution and data quality offered adequate information to start a characterization and new analysis of these events. The analysis of their data clearly showed that most high wind events (defined as those with wind gust speeds at 45m height ≥ 22.2m/s) are non-synoptic and usually occur during severe convective activity. Higher winds and more frequent high wind events are found to the northwest of Uruguay. On the other hand, around 70% of the high wind events occur during spring and summer.