• Cooling India
  • Sep 15, 2017

Wind Driven Ventilation: Holistic Approach of Cooling

Wind speed plays a major role in natural ventilation. From the view of human comfort wind speed should neither be very large nor very low. Best wind speed ranges for passive cooling is the wind speed for breeze which varies from 2-8 m/s. Now the question is the availability of this wind speed range in that particular location. This article focuses on wind meteorology, wind driven natural ventilation techniques and effects on natural ventilated building cooling and design in Indian scenario…

- Dr. J. Sarkar, Dr. A. Sarkar


 Wind driven ventilation arises from the different pressures created by wind around a building or structure, and openings being formed on the perimeter which then permit flow through the building. Due to rapid increase of environmental pollution it is very much necessary to maintain harmony with nature. Most of the technologies for cooling such as conventional refrigeration and air conditioning are not ecofriendly. On the other hand, India has a lot of urban and suburban areas in the coastal regions where wind speeds are very much prominent. In the wind zone map of India, basic wind speed of India varies from 33 m/s in Bangalore to 55 m/s in Darbangha. Based on the availability of wind speed, wind induced natural ventilation can be considered to be an appropriate technology for holistic cooling in India. However, for successful implementation of this technology, it is required to give attention on some key processing areas such as wind speed data measurement, non-stationary analysis of wind speed data, probabilistic modeling of wind speed data, wind direction analysis with wind rose, probabilistic modeling of wind direction, physics of wind flow and natural circulation leading to corridor setting. The knowledge of the urban climatology i.e. the wind around the buildings is crucial when evaluating the air quality and thermal comfort inside buildings as air and heat exchange depends on the wind pressure on facades.

  As naturally occurring wind blows across a building, the wind hits the windward wall causing a direct positive pressure. The wind moves around the building and leaves the leeward wall with a negative pressure, also known as a sucking effect. If there are any openings on the windward and leeward walls of the building, fresh air will rush in the windward wall opening and exit the leeward wall opening to balance and relieve the pressures on the windward and leeward walls. Capturing the wind and bring ventilation to the building depend on the building shape, building orientation and location, building form and dimensions, window typologies and operations, types, shape and size of openings, construction methods and detailing, external elements, urban planning consideration, etc.

Figure 1: A typical wind rose of Trivandrum

Wind Meteorology

  Wind speed plays a major role in natural ventilation. From the view of human comfort wind speed should neither be very large nor very low. Best wind speed ranges for passive cooling is the wind speed for breeze which varies from 2-8 m/s. Now the question is the availability of this wind speed range in that particular location. Wind speed is highly location specific. Even in a particular location wind speed greatly varies in different months. To determine availability of a particular wind speed range, an availability factor can be defined which can be equated to the probability of occurrence of wind speed in the particular range for that location. Wind speed data are generally measured either by cup and cone anemometer or dyne pressure tube anemograph. Care should be taken while measuring wind speeds as the current methods of measurements sometimes lead to sampling errors and the future prediction of wind speed from the supplied data by Meteorological Departments would be grossly affected. Suppose one anemometer pulse has been calibrated to 0.12 m/s depending on a particular instrument. Then two pulses would be equated to 0.24 m/s and three pulses would be equated to 0.36 m/s. Now if the Meteorological Department supplies the wind speed data to the end user considering a class width of 0.1 m/s which has been considered as the finest interval in many European countries, the data would be supplied as 0.1, 0.2 and 0.4 m/s. Hence, in this calibration 0.3 m/s would never appear in the wind speed data set which may not be true while considering natural wind speeds. Hence, the trend of the wind speeds would be biased which leads to the erroneous future prediction. Moreover, in India the class width for supplying wind speed data is 1 km/h which is too higher than the case cited in this example and this would cause more sampling error. Not only this, in India wind speeds are initially measured in knots and later converted to integer km/h by multiplying a factor of 1.852. Since knot and km/h has no common integer multiple, sampling error in wind speed data would obviously be induced due to this conversion. After removing the influence of sampling error by increasing the class width of the histogram, the correct trend of the same can be obtained.

  Since wind speeds are highly uncertain, it is very much necessary to know the availability of a certain range of wind speed which would be required for wind induced natural ventilation. For this wind speed, data should be modeled by a suitable probability distribution. But for any probabilistic modeling the parent dataset needs to be random whereas wind speed data are generally highly non-stationary unlike any other stationary random data. Also, for the wind speed analysis of wind driven natural ventilation block maxima cannot be taken for making the wind speed stationary, which is frequently used for the specification of the design wind speed for civil structures. Hence, in this case the trend and seasonality should be removed from the data. It is also required to remove white noise from the data and the residuals need to be modeled by Auto Regressive Integrated Moving Average (ARIMA) or Auto Regressive Moving Average (ARMA). After making wind speed data set as independent and identically distributed (iid) data, probabilistic models can be applied. Since wind speed data are highly non-Gaussian, World Meteorological Organization (WMO) suggested Weibull model for probability distribution of wind speed data. But sometimes other probability distributions such as Gamma, Inverse Weibull and other mixture models have also been found to be suitable for wind speed distributions of different locations of India. Sometimes bimodal or multi-modal distributions have been found to be suitable.

  Wind direction also plays a crucial role for wind induced natural ventilation along with wind speed. Wind directions are generally measured by 16 point compass. After determining the available duration in the required wind speed, the most favorable wind direction should also be determined for a location which is required for the orientation of doors and windows as well as corridor setting. Wind rose can be used for this purpose. Wind rose is a pictorial representation of the fractions of wind coming from different directions. A typical wind rose for Trivandrum has been shown in Figure 1. Due to climate change, atmospheric boundary layer also changes which has an influence on wind direction. For the uncertainty of wind direction, it is also required to fit wind direction data into some suitable probability distributions. Many researchers concluded that 2 component and 4 component von-Mises distributions are suitable for this purpose. A modeling of wind direction data of Trivandrum by von-Mises distribution has been shown in Figure 2.

Figure 2: Wind direction distribution of Trivandrum

  Apart from the above discussions for natural ventilation, it is also equally important to understand the physics of wind flow. Due to adverse pressure gradient and subsequent flow separation stream line distortions take place and vortex eddies are formed leading to circulations. When wind flows over a bluff body, in the back side of the body there is always circulation. That is why, the persons who are sitting at the back of the auto rickshaw experience more wind rather than those sitting at front. Sometimes flow separation and large suction regions are formed at the sharp edge. The formation of a horse shoe vortex has also been found at the bottom front corner of a bluff body. This flow circulation can be used for natural ventilation especially for the settings of windows and corridors. Figure 3 shows some simulation results regarding wind driven natural circulation.

Figure 3: Numerical simulations of wind driven natural circulation in different configurations

  Based on the above discussions a case study can be carried out for determining the configurations of doors and windows at some coastal locations. Three coastal locations have been considered in this study namely Kolkata, Ahmedabad and Trivandrum. After fitting the wind direction data into von Mises distribution the most probable wind directions for Kolkata, Ahmadabad and Trivandrum have been found as south, south-west and north-west respectively. Hence, in Kolkata wind flows from south, in Ahmadabad wind comes from south-west and goes to north-east and in Trivandrum wind flows from north-west towards south-east. Accordingly, the preferred window configurations among the configurations shown in Figure 3, for these locations have been tabulated in Table 1.

Wind Driven Ventilation Techniques

  Wind driven ventilation can be classified as cross ventilation and single-sided ventilation. Wind driven ventilation depends on wind behavior, on the interactions with the building envelope and on openings or other air exchange devices such as inlets or chimneys. Depending on their operation and mode of engagement with the wind, the wind driven ventilation techniques can be grouped as (i) passive technique, (ii) directed passive technique and active technique (Table 2).

Limitations of Wind Driven Ventilation

• Unpredictability and difficulties in harnessing due to speed and direction variations
• The quality of air it introduces in buildings may be polluted for example due to proximity to an urban or industrial area
• May create a strong draught, discomfort.

Conclusions

  The passive cooling system through the natural ventilation of wind can be considered as a holistic solution in place of traditional cooling system. However, large uncertainties are involved in the availability of the desired wind speed range. Climate change can also influence both wind speed and direction. Accordingly, corridors should be oriented for passive cooling. Various traditional and modern wind driven ventilation techniques are discussed, which can be helpful to choose suitable ventilation technique with proper design strategies.


If you want to share thoughts or feedback then please leave a comment below.