Wind analysis | Wind nuisance

Insight in air flows around buildings is important for several reasons. There are several themes that could be evaluated; wind nuisance; comfort-cooling options in summer; air quality due to exhaust of chimneys and traffic; air quality around parking garages; options for energy for wind-turbines; options to support ventilation systems in buildings (inlets and/or outlets) and options for natural ventilation.

In this page you can find tutorials that will serve to give you insight on wind nuisance in the built environment with a general introduction into air flows around buildings. Wind nuisance can be assessed by reference cases, hand calculations (Verhoeven 1982), wind tunnel tests and air flow calculations (CFD). The focus of this page is on air flow simulations (CFD).

Special attention needs to be paid to the provision and careful consideration of outdoor spaces, at different heights:

• at +c. 30m

• between 75 and 90 meters

• near the top of the building

Content

> Computational domain

> Assessment of wind nuisance

> Recommended literature

> Wind Simulation Tutorials

> Computational Domain

One of the main starting points is the determination of the size of the computational domain. The smaller the domain the faster the calculation can be executed. For a simulation at least one row of buildings (if there is) around the evaluated site will give acceptable results (Yoshie et al 2005).

Fig. 1: At least one row of buildings around the evaluated block (b) is necessary to obtain sufficient accurate (uncertainty below 10 %) results (Yoshie et al 2005)

In the following picture an example of this approach is presented, a wind study for a hospital in Delft where the buildings are generally lower than 15 - 30 m. The size of the domain is 600 x 300 x 60 m, with 136.400 cells. The NEN 8100 advises to include the surrounding buildings, generally within a minimum radius of 250 m.

Fig. 2: CFD-model of the Reinier de Graaf-hospital. Only the direct surrounding buildings are modelled (EGM et al 2011). The numbers A, B, C and D show the most relevant locations where wind-comfort is evaluated.

In order to be very accurate a very large size of the calculation domain would necessary: up to 15 times the maximum size of and around the evaluated building (figure 2). For high rise buildings this assumption has much effect, because the size of the domain for a building of 200 m would become 6 x 6 km. This is due the large size of a wake at the leeward side. In reality, because the influence of the size of up and downward air flows in very rough surroundings is limited, often a domain size of around 500 x 500 x 300 m will do (Yoshie et al 2005). In this case a grid with 1.300.000 cells was sufficient. For a detailed assessment of comfort in city areas with buildings of a moderate size (generally below 30 m) a domain-size of, for instance, 1700 x 700 x 200 m is an option with around 18.000.000 cells (Peutz 2018). It depends as well on the available computational speed and time for a first assessment of the location what kind of approach is feasible.

Fig. 3: Maximum domain size which takes into account the effect of upward and downward air flows (wake and eddies)

In order to have a quick start it is recommended to start with a model with a very coarse grid of less than 100.000 = 200.000 cells (Lee et al, 2011), and gradually increasing the grid refinement, especially near the buildings. Of course, this is also dependent of the size of the buildings and location (site).
For the computer-simulations it is important to know what the terrain roughness is. This is related to the wind profile (Verhoeven 1982). This can be a power law function:

Altitude of the divide between the Atmospheric Boundary Layer (ABL) and the outer flow, measured from ground level (m). This altitude is usually called gradient altitude and depends on terrain roughness.

Mean wind velocity at gradient altitude (m/s)

Coefficient, which depends on terrain roughness (-)

Fig. 4: Image and values for a power law approach

or a logarithmic function:

measure for terrain roughness (m)

known reference altitude (m)

known mean wind velocity at reference altitude (m/s)

Table 1: Relation between the terrain roughness and the power law profile of the wind

(α: Values for a power law approach)

Fig. 5: Example of a terrain roughness map around the Reinier de Graaf hospital (radius 6 km) from the NPR 6097 database. For instance, the red colour represents a roughness of 1.6 m (EGM et al 2011).

However, not only the average wind velocity is important. Wind is in general not stable occurs in gusts. These gusts can cause dangerous situations. It is possible to estimate the maximum wind velocity due to gusts according to Verhoeven (1982) as follows.

In a free field the following equation of the variation of wind can be used (Verhoeven, 1982):

The maximum wind velocity due to gusts is then:

Comparing the simulated average wind speed with the simulated wind speed at the corner of the building, the wind amplification factor (c) of this building can be calculated. The wind amplification value is the simulated wind speed at a certain position divided by the average wind speed at that height, see Verhoeven (1982).

> Assessment of wind nuisance

The procedure to assess wind nuisance is as follows:

1. The wind pattern of the location at a height of 60 is determined, following the calculation program from the NPR 6097.
2. In a CFD program or wind tunnel the velocity at a height of 1.75 m above the ground for 12 wind directions and critical locations are determined. This velocity is related to a reference air velocity at a height of 60 m.
3. The amount of hours per year that a wind velocity of 5 m/s (wind comfort) or 15 m/s (wind danger) at the location is exceeded is assessed:

Table 2: Assessment criteria for wind nuisance according the NEN 8100

Fig. 6: Assessment according the NEN 8100 with Phoenics related to all the relevant wind directions for the activity “strolling”. Generally spoken the outdoor wind comfort is good. Only near some corners wind comfort Is reduced. There is no wind danger.

For buildings lower than 15 m the NEN 8100 does not expect wind comfort-problems. For buildings between 15 and 30 m and especially higher than 30 m an additional assessment of a wind expert is necessary. It is possible to get weather- and wind-information from the recent past and in Rotterdam via the following link of the KNMI: http://projects.knmi.nl/klimatologie/daggegevens/selectie.cgi.

Wind-velocity and wind-direction has been measured at a height of 10 m. This height-value of 10 m is important as a reference in the CFD-simulations, when values at a height of 60 are not available.

Important note: the hourly average wind velocity values of the KNMI-website are not the peak-values combined with the effect of wind-gusts. Those values are used by construction engineers and can be derived from standards.

Fig. 7: Example of graphical and numerical wind statistics for a location in Delft at a height of 60 m from the NPR 6097. The South-West wind is the dominant direction (Peutz 2018).

> Recommended Literature

1. EGM Deerns Corsmit VOF. Nieuwbouw Reinier de Graafgroep Delft. Onderzoek windhinder en windgevaar. 2011.
2. Lee S, Song D. Evaluation of wind environment around the building in the early design stages using BIM-based CFD-simulation. International Journal of air-Conditioning and Refrigeration. Vol. 19, no. 4 (2011) 263-272.
3. NNI. NEN 8100. Wind comfort and wind danger in the built environment. 2006.
4. NNI. NPR 6097. Application of the statistics of the mean wind speed for the Netherlands. 2006.
5. Clara Sanchez_Pijpers-Van Esch . Wind climate in urban areas. Lecture wind nuisance at the Mega course. 2018_2019. (on Brightspace)
6. Peutz. Plan Nieuw Delft. Windklimaatonderzoek met behulp van CFD. 15 mei 2018. (downloaded via internet).
7. Verhoeven AC. Wind nuisance. TU-Delft, 1982. (English translation of the 1982 document)
8. Yoshie R, Mochida A, Tominaga Y, Kataoka H, Harimoto K, Nozu T, Shirasawa T. Cooperative project for CFD prediction of pedestrian wind environment in the Architectural Institute of Japan. Journal of Wind Engineering and industrial Aerodynamics 95. 2007.