Posters - WindEurope Technology Workshop 2022
Resource Assessment & Analysis of Operating Wind Farms 2022
23-24 June • Brussels


Come meet the poster presenters to ask them questions and discuss their work

Check the programme for our poster viewing moments. For more details on each poster, click on the poster titles to read the abstract.

PO098: Reduction of wind speed spatial variation uncertainties in a large on-shore wind farm array utilizing mesoscale modelling

Nikolaos Daniil, Senior Manager, Technical - Renewables, AMEA Power


AMEA Power is developing a 500MW wind farm project in Ras Ghareb region in Egypt. The site is large and consists of flat desert terrain. High concentration of stable and unstable atmospheric conditions have been observed from analysis of the on-site measurement datasets with mesoscale flow effects (Coriolis effect, Coastal Jets, etc) influencing the flow in the region and prevailing in the micro-scale. AMEA Power has performed an evaluation of different wind flow models to aid in the decision and select of the model that achieves the lowest wind flow modelling spatial variation uncertainties for the project. AMEA Power's methodology for reducing the spatial variation uncertainties is performed utilizing a combination of very low wind speed uncertainty measurement campaign and wind flow modelling approaches. In regards to configuration of the measurements campaign: 6 x 83m IEC 61400-12-1 ed2 & MEASNET compliant masts (average campaign data availability: ~99.6%) + 3 x 82-84m IEC 61400-12-1 ed1 & MEASNET compliant mast datasets (average campaign data availability: ~99.3%) have been procured. The masts are well located and a very close distance with respect to the proposed turbines in terms of elevation, exposure and spatial representativeness, reducing the the chance of large spatial extrapolation errors in the wind farm mean wind speed and direction estimation. Different modeling approaches have been tested and ranked towards achieving the lowest possible wind speed spatial variation uncertainties. More specifically, the variation in wind speed over the wind farm site has been predicted using mesoscale modelling provided by Vortex and in house WRF-ARW (ERA5) simulations, WAsP and ZephyTOOLS computational flow models. The Vortex mesoscale flow model (Vortex FARM) was calibrated with measurements and long-term corrected wind flow model was established and compared against simulation results from WAsP and ZephyTOOLS software packages. The calibrated Vortex FARM is considered to be the most suitable type of flow model for this site. In regards to the cross prediction model performance, considering the distances between the masts, the results show good agreement between the measured and modelled all-directional wind speeds at the measurements for the Vortex model and in house WRF-ARW (ERA5) simulations, but poor agreement for the WAsP and CFD model. This is as expected, as the wind flow in this region is dominated by mesoscale effects which are not captured by micro-scale models such as WAsP and neutral stability CFD models such as ZephyTOOLS k-epsilon turbulence model variations. The Vortex prediction error between the southern mast and the two northern masts is as expected considering the large distance (~15 km) between the masts. The reasonable quality of the cross predictions adds confidence in the ability of the Vortex flow model to accurately capture the variation in wind speed across the site with the spatial extrapolation wind speed uncertainties reduced from 3% to 1% (confirmed by specialist consultancy team who performed the model validation) when more measurement data (6 vs 14 months of data) were utilized in the mesoscale model calibration.