Posters

Abstract

Uncertainty quantification plays an essential role in wind resource assessment. The uncertainty to a high degree defines the bankability of the windfarm project. Therefore, the siting report typically includes a detailed assessment of various uncertainty components such as measurement uncertainties, uncertainties on losses, as well as uncertainties related to horizontal and vertical interpolation by numerical models. The main objective of the “Simulation uncertainties for the detailed assessment of wind energy yield” - SUnDAY project is to systematically quantify the uncertainty of the numerical modelling involved in the siting process. Currently, here often single experience based values are taken throughout multiple projects. In the presented work, a non-intrusive polynomial chaos methodology is applied to the micro-scale CFD (computational fluid dynamics) model. Unlike other methods such as Monte-Carlo, a limited number of simulations is required only, which is important in case of CFD model due to the computational cost. The required number of simulations is depending on the order of the chosen polynomial function. Furthermore, the non-intrusive approach does not require any changes to the CFD code base, which makes it suitable for any existing CFD model as a postprocessing step. The newly developed method can be applied to any model parameter such as mesh resolution, domain size, but also, physical parameters such as forest height or thermal stability parameter etc. In the presented work, two parameters were investigated, namely mesh resolution and forest height in complex terrain at several real-world wind energy sites. The uncertainty of the wind speed profiles is calculated at various positions with respect to each parameter. The presented method is not only limited to wind speed profiles but can also be extended to the entire domain or terrain following slice resulting in an uncertainty map for the entire site.