Presentations

Abstract

Meso-scale coupled large-eddy simulation (LES) is becoming an essential ingredient of the energy yield assessment (EYA) of large wind farms. The holistic, physics-based description of meso-scale flow effects, global blockage and wind turbine wakes provide e.g. aerodynamic loss estimates without (site-specific) assumptions or tuning. However, offshore wind farms will seldom operate in isolation, so their energy production is impacted substantially by wake and blockage effects induced by surrounding wind farms. Despite the technical feasibility of resolving all wind turbines of the surrounding wind farms in a turbine-scale LES, the computational costs of such simulations make this an economically less attractive option, especially in the earlier stages of the EYA process. Furthermore, it might not be needed to apply LES precision to surrounding wind farms of which only the external aerodynamic effects impact the wind farm of interest.    With this in mind, we present an alternative modelling approach in which the surrounding wind farms are parametrized in a meso-scale model which, in turn, provides boundary conditions to an LES in which the wind farm of interest is modelled down to the turbine scale. This approach balances the cost-effectiveness and accuracy of meso-scale models in capturing the cumulative wake and blockage effect of wind farm (clusters), while exploiting LES precision for representation of the internal aerodynamic losses of the wind farm of interest.    The proposed methodology is demonstrated with a year-long simulation of a full built-out scenario of the IJmuiden Ver cluster in the North Sea, including the current and future surrounding wind farms. By incrementally increasing the size of the LES domain, hence including more of the surrounding wind farms within the LES, the added uncertainty of using a meso-scale model for representing external wind farm wakes will be assessed. Key metrics for evaluation are the total, internal and external aerodynamic losses, as well as the first-row blockage-related losses.