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.

Arno Motté Vrije Universiteit Brussel, Belgium
Jochem Vermeir (1) F Arno motté (1) P Diego Dominguez (1) Mark Runacres (1) Tim De Troyer (1)
(1) Vrije Universiteit Brussel, Anderlecht, Belgium

Printer friendly version: printer.gif Print




Can small wind energy systems contribute to urban sustainable development in the Brussels Capital Region? One approach to answer this question is by performing wind measurements on potentially interesting sites chosen by their location, surrounding building heights and topographical information. Using these measurements we can estimate the annual energy production on these locations. Our wind measurements also serve as a validation tool for a wind map of Brussels, created with the methods of Landberg et al. (2003) and Millward-Hopkins et al. (2013). In this way they can help to predict the wind potential throughout Brussels.


The main goal of our work is to determine the wind potential of the Brussels Capital Region. Therefore we conducted a measurement campaign on four different locations in the city. Each of the locations is identified based on several criteria: topographical data, location of the site (to have measurements at different regions in Brussels) and the height of the building where the measurement equipment will be installed (if installed on a rooftop). Preliminary results show wind speeds that are comparable with wind speeds measured at low heights (~10 m) close to the North Sea. This means that the wind potential at these Brussels’ sites is large.

To have a more global idea about the wind potential in different regions in Brussels, we use two different analytical methods to estimate the above-roof mean wind speed for other locations in Brussels. These methodologies use detailed geometric data describing buildings and city topography. The geometrical data can afterwards be used to predict the aerodynamic characteristics of the urban environments. The approach of this work is to repeat this process for different locations in Brussels and so to create a wind map with the above-roof mean wind speed. Therefore we apply two different methodologies (one simplified method proposed by the UK Met Office and one more complex model by Millward-Hopkins et al. (2013)) and compare their results. The data gained by the measurement campaigns will be used to validate the accuracy of the methods.

Both methods are based on what is known as the ‘wind atlas methodology’ (Landberg, et al. 2003). They scale the wind speed data (from a free-field meteorological station) at low height up to the geostrophic wind (height at which the frictional effect of the surface is negligible). Next, the geostrophic wind is scaled back down accounting for the effect of the surface roughness upon the wind profile.

In the simplest method (referred as the Met Office model) aerodynamic parameters are calculated on a regional scale during the downscaling process. The more advanced Millward-Hopkins method includes the influence of wind direction and uses more detailed data to determine the aerodynamic parameters.

Main body of abstract

It is possible to estimate mean wind speeds analytically over an urban area by applying a ‘Wind Atlas Methodology’. This method requires information on both the regional wind climate and the roughness characteristics of the surface, and enables to generate what is known as urban wind maps. Different methodologies could be followed depending on the resolution and the level of detail of the available data. Here we use two of them: a classical method only requiring a general description and regional data of the locations (the Met Office method), and a more sophisticated one proposed by Millward-Hopkins et al. (2013) employing highly detailed urban maps and incorporating the influence of changing wind direction.

The procedure to follow in order to predict wind values at measurement sites is now described. First a mean wind speed value from a meteorological station located in open terrain is scaled up to the top of the urban boundary layer (zUBL) using the standard logarithmic wind profile and the open terrain roughness length. After that, this velocity at the top of the urban boundary layer is scaled down through the urban boundary layer to the blending height, making use of aerodynamic parameters, like the roughness (z0-fetch) and the displacement heights (dfetch), calculated on a regional scale. Finally, the velocity at blending height is scaled down again to the measurement height, but now using roughness and displacement heights (z0-local dlocal) adapted to the local area in the surrounding 100-200 m. In the Met Office model, the aerodynamic parameters depend on the mean building height hm of the area of interest. This means that, if we want to create a wind map of a greater area, it has to be divided into a grid (as showed by Drew et al. (2013) and by Bottema et al. (1998)). For each cell the mean building height has to be determined.

When using the Millward-Hopkins approach, slight modifications are made to the method:
• The influence of changing wind direction is taken into account: The height of the urban boundary layer as a function of the distance to the edge of the city and z0-fetch and dfetch vary for different wind directions.
• The roughness lengths and displacement height are calculated in a more accurate way making use of highly detailed grids.
The aerodynamic parameters now also depend on the building/ground ratio and width/height ratio of the building in each grid. Therefore these areas need to be modeled more accurately. An example of such a detailed cell of the grid for Brussels is given in figure 1.

Figure 1: Detailed 3D model of one cell for Brussels

The results from the Met Office model have been validated by wind measurements performed on four locations in Brussels. The measurement campaign has started in February 2013. An overview of the different sites is shown in table 1.

Table 1: Description of the measurement sites

To validate the accuracy of the predictions, the estimated wind speed is compared with the measured mean wind speed at the site. This comparison is repeated for all validation sites. In table 2, the wind speeds and the estimation error are listed.

Table 2: Validation of the Met Office model

The table shows that the accuracy of this model is in the order of ±0.5 m/s at best. The estimation of Site 4 is much worse. We conclude that, with simple models, we have errors around 10-20 %, with outliers up to 60 %.

Although the results presented above are reasonable given the simplicity of the adopted procedure, preliminary results using the Millward-Hopkins approach show a significant improvement of the accuracy of the predictions. These results will be included in the full version of this paper.


In this paper, we present our work in assessing the wind potential in Brussels. We perform wind measurements at four different locations, and we use two different analytical methods to estimate the mean wind speed at other locations. Using these methodologies, we intend to create a wind map of the city of Brussels. For the Met Office model, validation showed that the accuracy of the estimation is between 10 and 20%. One site showed a larger deviation between the measured and the estimated mean wind speed. This is probably caused by two reasons:

• The first reason is the complexity of the location. For this site, the measurement equipment is installed on a bridge that crosses the canal. Both sides of the canal are built up with tall buildings. These buildings channel the flow and locally increase the wind speed, making this location very interesting for small wind turbines. The channeled flow is clearly visible when the frequency of the wind is plotted per wind direction for this site. The wind rose in figure 2 clearly shows the southwest to northeast orientation, which is also the orientation of the canal. Because of these rather complex wind conditions it is difficult to predict the wind speed for this location with too simple a model.

Figure 2: Wind rose site 4

• The second reason for the low accuracy is that the height at which the measurements were taken (12 m) is below the local mean building height or the ‘canopy height’. Below the canopy height the flow patterns are too complex to be adequately captured by this method.

Given that this simplified model only uses non-detailed building data, more accurate predictions are unachievable. Preliminary results show that the accuracy is improved with the Millward-Hopkins approach. The results of this approach, including the validation of the estimations, will be further discussed in detail in the full paper.

Learning objectives
Can small wind energy systems contribute to urban sustainable development in the Brussels Capital Region? To answer this question, we performed wind speed measurements in the urban environment on carefully chosen locations. To have a more global idea about the wind potential in different regions in Brussels, we use two different analytical methods to estimate the above-roof mean wind speed for other locations in Brussels

Bottema M., Mestayer P.G. “Urban roughness mapping - validation techniques and some first results.” Journal of Wind Engineering and Industrial Aerodynamics 74 (1998): 163-173.

Drew D.R., Barlow J.F., Cockerill T.T. “Estimating the potential yield of small wind turbines in urban areas: A case study for Greater London, UK.” Journal of Wind Engineering and Industrial Aerodynamics (Elsevier) 115 (2013): 104-111.

Landberg, L., L. Myllerup, O. Rathmann, E. L. Petersen, B. H. Jorgensen, and J. Badger. “Wind resource estimation- an overview.” Wind Energy, no. 6 (2003): 261-271.

Millward-Hopkins, J. T., A. S. Tomlin, L. Ma, D. B. Ingham, and M. Pourkashanian. “Mapping the wind resource over UK cities.” Renewable Energy, no. 55 (2013): 202-211.

Wieringa, J. “Updating the Davenport roughness classification .” Journal of Wind Engineering and Industrial Aerodynamics 41 (1992): 357-368.