Back to the programme printer.gif Print

Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'How does the wind blow behind wind turbines and in wind farms?' taking place on Tuesday, 11 March 2014 at 16:30-18:00. The meet-the-authors will take place in the poster area.

Carsten Albrecht AL-PRO GmbH & Co. KG, Germany
Carsten Albrecht (1) F P Anselm Grötzner (2) Didier Delaunay (3) Céline Bezault (3) Mareike Kohlert (1) Lukas Pauscher (4)
(1) AL-PRO GmbH & Co. KG, Großheide, Germany (2) CUBE Engineering GmbH, Kassel, Germany (3) Meteodyn, Nantes, France (4) Fraunhofer IWES, Kassel, Germany

Printer friendly version: printer.gif Print




Considering thermal stability in wind power assessment is becoming more and more popular, especially with regard to increasing turbine hub heights and therefore the need to model the wind conditions in the upper boundary layer. But what is the best way to determine the in-situ-stability and how to introduce it in CFD models? Investigations have shown that comparing stability calculated from wind measuring data and mesoscale data often brings diverging results.
This paper proposes some solutions by analyzing reanalysis data and in-situ measurements as well as a new approach for the modeling of stability effects in CFD codes.


Regarding the determination of the in-situ-stability, the idea is to use several datasets available for a braod range of sites, like meso scale and reanalysis datasets. This will be the modern-era retrospective analysis for research and applications (MERRA), ConWX, which has a very fine resolution of 3x3 km, Vortex and ERA Interim. Data from 200 m met towers at two sites will be used as well. One site is located at Cabauw in the Netherlands, the terrain is simple structured. The other site is located in Wolfhagen near Kassel in Germany where the terrain is complex structured.
Regarding the numerical modeling of the stability effects, the approach is to solve the RANS equations using a k-L model in which the turbulent length scale is a function of the atmospheric stability, considering several layers in the Stable Boundary Layer. The upper height where the Monin-Obukhov Similarity Theory is valid constitutes a characteristic of the thermal situation.
Simultaneously a statistical method giving joint distributions of wind speed, wind direction and stability has to be defined for the application to wind resource assessment. The main point of this method is to make a realistic classification of stability classes, considering few parameters: wind speed, wind direction and height of the MOST layer. This methodology should be able to eliminate some problems regarding wind profiles at large heights and the horizontal model transfer for stable stratified situations, including strong stability cases. After implementation in Meteodyn WT, validation of the methodology is conducted by comparison with measurements up to 200 m height at several sites.

Main body of abstract

At first, this paper presents new results in determining stability classes and related wind profiles from flux measurements at 200 m met towers and data from reanalyses and mesoscale models (MERRA, ConWX, Vortex and ERA Interim). We show the capability of reanalyses data in reproducing stabilities being obtained by real measurements and how to use the reanalyses to steer a CFD-model in cases where no in-situ stability data are available.
Solving the steady RANS equations is at present the most efficient approach for the micro-scale flow computation in wind resource assessment. Generally, the turbulent fluxes are linked to the gradients of the mean variables via the notion of eddy viscosity, product of a wind speed scale, generally the square root of the turbulent kinetic energy (TKE) and a turbulent length scale LT.
The TKE is computed via a transport equation, including a dissipation term which depends on LT . At present, this approach fails to reproduce strong stability cases at the level of high wind turbines, because of the underlying hypothesis of the Monin-Obukhov Similarity Theory (MOST) which applies in the limit of a critical Richardson number.
LT can be deduced either from a TKE dissipation transport equation or by a semi-empirical parameterisation (k-L model). Here, a new multi-layer approach is proposed, taking benefit of the k-L model which allows a great flexibility of adaption to real atmospheric flows including the fitting with experimental data. Four layers are considered:
1. The dynamic layer, dominated by mechanical effects at the ground level and inside the canopy, up to 2 times the roughness elements heights.
2. The MOST layer where the Richardson number is inferior to a value of about 0.10 (Grachev et al. 2013. In this layer, the model of Yamada and Arritt (Hurley, 1997) is used with a modification in the mixing length formulation in order to take into account a limiting buoyancy length scale.
3. A transitional layer above the MOST layer, where the fluxes are either modeled by a “z-less scaling” or a local MOST theory where the fluxes decrease with height up to a critical Richardson number.
4. The free atmosphere, where the turbulent fluxes are not zero, but depend on large scale turbulence generated at the meso-scale level (Zilitinkevitch, 2010)
This model is implemented in the CFD software Meteodyn WT and discussed in comparison with large met tower measurements both in simple and complex terrain.
Another issue is that for wind resource assessment, CFD modeling has to consider wind statistics. Up to now these statistics are represented by a joint distribution of wind speed and direction. Considering the usual practice (for instance MCP methods do not consider stability classes), the present Meteodyn approach is to consider an “effective average stability class” for one site and one wind direction sector, i.e. to choose the stability class which fits at best with directional average wind profiles. The stability is then defined by a single Obukhov length value, linked to the gradient Richardson number, which is generally available through standard measurements.
In the new approach, the stability class is defined by the height HMOST of the MOST layer. The notion of “effective average stability” disappears as we have to consider wind speed/direction/stability statistics. For establishing these statistics, a method is proposed considering wind profiles, temperature profiles and fluxes measurements, if available.


The result of this study is a step forward for a better consideration of the atmospheric thermal stability in wind resource assessment. The new methodology concerns the aspects of on-site measurements treatment, Reanalysis data interpretation as well as models for numerical simulations of the wind flow through the following topics:
• The determination of realistic stabilities from Reanalysis and mesoscale datasets: The relationship between the stability given by mesoscale models and the stability deduced from mast measurements at several sites has been analyzed and conclusions are driven on the reliability of such data.
• The characterization of stability classes from measurements at met masts for an optimal application to the wind flow numerical simulations. It includes the analysis of several parameters: the gradient and the flux Richardson numbers , the Obukhov length, the height of the MOST layer and the height of the Stable Boundary layer. The height dependency of the Richardson number is also considered.
• A new approach, using a k-L model, for modeling the turbulent fluxes taking into account situations of strong stability beyond the applicability of the MOST theory. In the MOST layer the model of Yamada and Arritt (cited by Hurley, 1997) is modified for a better consideration of the buoyancy scales. In the transitional layers, modified Obukhov functions and a z-less approach are proposed. Moreover, the presence of turbulence above the Stable Boundary Layer has to be considered.
• An implementation in the commercial software Meteodyn WT and an assessment versus measurements at large heights on several sites (flat or complex terrains).
• A methodology for a statistical treatment of the joint distribution of wind speed/wind direction/thermal stability in an industry leading CFD approach

Learning objectives
Determination of stability information from various data sources.
Assessment of a new approach for turbulence modeling in CFD codes in order to improve wind extrapolation at large heights.
Improved method to include stability into wind resource assessment considering wind statistics.

Grachev, A., Andreas E., Fairall C., Guest P., Persson P. (2013) “The Critical Richardson Number and Limits of Applicability of Local Similarity Theory in the Stable Boundary Layer”, Boundary-Layer Meteorology, Vol. 147, Issue 1, pp 51-82
P.J. Hurley (1997) “An evaluation of several turbulence schemes for the prediction of mean and turbulent fields in complex terrain”. Boundary-Layer Meteorology, Vol. 83, Nr 1, pp. 43-73(31)
Zilitinkevich, S. S., Eleperin, T., Kleeorin, N., Rogachevskii, I., L’vov, V., Esau, I., Kouznetsov, R. (2010) “Turbulence closure for stably stratified flows: local and non-local formulations”, 8th European Conf. on Applied Climatology, EMS-2010, Zürich, Switzerland, 13-17 September 2010