Comparison of Fuga and RANS in complex terrain

Research output: Book/ReportReportResearchpeer-review

316 Downloads (Pure)


Modeling wind turbine wakes and wind resources in complex terrain is challenging. High fidelity Computational Fluid Dynamics (CFD) methods can produce accurate results of long term averaged speed up factors for flow over complex terrain although user uncertainty can create large differences between CFD model results [1]. Engineering wake models are often not suited to model wakes in complex terrain because they are developed for wind farms in flat terrain [2]. Fuga [3] is a fast linearized CFD model developed to model wake effects in wind farms for flat terrain and uniform roughness lengths. The base flow in Fuga is an atmospheric surface layer following Monin-Obukhov Similarity Theory [4] and the effect of a wind turbine wake is added as a linear perturbation. Recent work has been carried out to develop a linearized terrain flow model in Fuga. The linearity of this terrain flow model implies that the model can only predict first order perturbations of the terrain and cannot predict flow separation and recirculation zones. In unpublished work, a comparison between the terrain flow model of Fuga and a Reynols-averaged Navier-Stokes (RANS) model as implemented in EllipSys3D [5, 6] has been made for the flow over an axisymmetric Gaussian hill with a moderate maximum slope angle of 17◦using eight test cases. This moderate slope angle was chosen to avoid flow separation. The comparison showed that the streamlines of Fuga and EllipSys3D compare well for all eight test cases. Results of this study are presented in Section 3.1 and the comparison is extended with results of speed up factor contours and profiles. In this report, we would like to continue the
comparison between the terrain flow model of Fuga and RANS (without wind turbine wakes) for additional test cases that are designed to answer the following three research questions:
1. How well do the wind turbine wake streamlines follow the terrain streamlines?
2. For which slope angles does the terrain flow model of Fuga start to deviate significantly from RANS in terms of streamlines and speed up factors for an axisymmetric Gaussian hill?
3. How well does the terrain flow model of Fuga compare with RANS for a real complex terrain site in terms of speed up factors and velocity profiles?

The methodology and test cases are presented in Section 2. The results are discussed in Section 3.
Original languageEnglish
PublisherDTU Wind Energy
Number of pages27
Publication statusPublished - 2018
SeriesDTU Wind Energy E

Cite this