Titre de la thèse: Downscaling the wind energy resources in complext terrain using a coupled mesoscale/microscale CFD modeling system including wake effect
Lieu de la soutenance :
Ecole des Ponts ParisTech - Bâtiment Coriolis - Amphi Caquot,
6-8 avenue Blaise Pascal, Cité Descartes,
Champs-sur-Marne 77455 Marne la Vallée
- Pr. François CAUNEAU, École des Mines de Paris, raporteur et président du jury
- Pr. Jeroen VAN BEECK, The Von Karman Institute for Fluid Dynamics (rapporteur)
- Dr. Javier SANZ RODRIGO, CENER (examinateur)
- Dr. Bertrand Carissimo, CEREA/EDF (directeurs de thèse)
- Dr. Eric DUPONT, CEREA/EDF (codirecteurs de thèse)
The development of wind energy generation requires precise and well established methods for wind resource assessment, which is the initial step in every wind farm project. During the last two decades linear flow models were widely used in the wind industry for wind resource assessment and micro siting. But the linear models inaccuracies in predicting the wind speeds in very complex terrain are well known and led to use of CFD, capable of modeling the complex flow in details around specific geographic features. Mesoscale models (NWP) are able to predict the wind regime at resolutions of several kilometers, but are not well suited to resolve the wind speed and turbulence induced by the topography features on the scale of a few hundred meters. CFD has proven successful in capturing flow details at smaller scales, but needs an accurate specification of the inlet conditions. Thus coupling NWP and CFD models is a better modeling approach for wind energy applications.
A one year field measurement campaign carried out in a complex terrain in southern France during 2007-2008 provides a well documented data set both for input and validation data. The proposed new methodology aims to address two problems: the high spatial variation of the topography on the domain lateral boundaries, and the prediction errors of the mesoscale model. It is applied in this work using the open source CFD code Code_Saturne, coupled with the mesoscale forecast model of Météo-France (ALADIN). The improvement is obtained by combining the mesoscale data as inlet condition and field measurement data assimilation into the CFD model. Newtonian relaxation (nudging) data assimilation technique is used to incorporate the measurement data into the CFD simulations. The methodology to reconstruct long term averages uses a clustering process to group the similar meteorological conditions and to reduce the number of CFD simulations needed to reproduce 1 year of atmospheric flow over the site. The assimilation procedure is carried out with either sonic or cupanemometers measurements. First a detailed analysis of the results obtained with the mesoscale-CFD coupling and with or without data assimilation is shown for two main wind directions, including a sensitivity study to the parameters involved in the coupling and in the nudging. The last part of the work is devoted to the estimate of the wind potential using clustering. A comparison of the annual mean wind speed with measurements that do not enter the assimilation process and with the WAsP model is presented. The improvement provided by the data assimilation on the distribution of differences with measurements is shown on the wind speed and direction for different configurations.