effects of building separation and podium on pedestrian-level wind ...
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
EFFECTS OF BUILDING SEPARATION AND PODIUM ON PEDESTRIAN-LEVEL WIND ENVIRONMENT C.W. Tsang1, K.C.S. Kwok2 and P.A. Hitchcock3 CLP Power Wind/Wave Tunnel Facility, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong S.A.R., P.R. China, [email protected] 2 School of Engineering, University of Western Sydney, Sydney, New South Wales, Australia, [email protected] 3 CLP Power Wind/Wave Tunnel Facility, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong S.A.R., P.R. China, [email protected] 1
ABSTRACT High-rise buildings affect the surrounding pedestrian-level wind environment. In recent years, awareness and concern have increased about the creation of low wind speed areas around buildings which may lead to poor outdoor air ventilation. Moreover, many modern building developments are not restricted to a single building but may comprise a group of buildings. Very few systematic studies have focused on the low wind speed areas around a group of buildings. In this research, a series of parametric wind tunnel studies was carried out to investigate the effects of building separation and podium on the pedestrian-level wind environment around a group of tall buildings. Mean wind speeds were used to determine the low wind speed areas where poor air ventilation may exist, and Gust Equivalent Mean (GEM) wind speeds were used to identify high wind speed areas with the potential to cause discomfort under strong wind conditions. KEYWORDS: PEDESTRIAN-LEVEL WIND ENVIRONMENT, BUILDING SEPARATION, PODIUM
Introduction In many densely populated cities, such as Hong Kong, urban renewal is an important aspect of sustainable development for the community to make the best use of available land and infrastructure. Under the renewal projects, modern high-rise buildings have been built inside the closely packed old districts. Due to the significant changes to the building forms, community concerns and awareness have been raised of the potential effects of newly built structures on the surrounding wind environment, particularly for tall and bulky buildings that, when closely packed together, form undesirable barriers which obstruct winds from penetrating into the downstream urban fabric and which result in poor natural air ventilation. However, community concerns and the corresponding solutions suggested by the designers and engineers are mostly based on personal and professional experience which may be subjective. There is a genuine lack of solid scientific literature on which to base arguments and design decisions. Wind flow patterns around buildings are very complicated and have been investigated for more than 30 years. A comprehensive review focusing on the flow features and pedestrian-level wind environment around buildings was conducted by Blocken and Carmeliet (2004). Since the 1960s, outdoor human comfort has been given a significant amount of attention. Wind environment studies have focused on pedestrian-level discomfort caused by strong winds near idealized buildings (Wiren 1975, Stathopoulos and Storms 1986, Stathopoulos and Wu 1995, To and Lam 1995, Blocken et al. 2008 and others). In those studies, the effects of building characteristics, such as dimensions, shape, building separation,
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
orientations of building groups and street canyons, have been investigated, from which knowledge-based expert systems were developed (Stathopoulos et al. 1992, Visser et al. 2000). These systems allow a preliminary and simplified evaluation of high wind speed areas around buildings. However, very few systematic studies have focused on the low wind speed areas around a group of buildings (Stathopoulos and Wu 1995, Chan et al. 2001, Kubota et al. 2008, Tsang et al. 2009). Tsang et al. (2009) have conducted sets of parametric studies to invesigate the effects of building dimensions and building separation on the low wind speed areas at pedestrian level. This paper presents the results of an extension of those studies for a single building and a pair of buildings to a row of four identical buildings with and without a podium. The pedestrian-level wind environments around these building configurations were investigated in a wind tunnel model study. The natural air ventilation and pedestrian comfort are both evaluated. Experimental Arrangement The experiments were carried out in the 3 m 2 m 29 m long high-speed test section of the CLP Power Wind/Wave Tunnel Facility (WWTF) at The Hong Kong University of Science and Technology. The mean wind speed profile of the approaching turbulent wind flow followed a power law exponent of 0.2, using a series of fences and roughness blocks, to simulate wind flow above a typical suburban terrain. All of the wind tunnel tests were conducted using a reference mean wind speed Ur of approximately 10 m/s at 150 m (in prototype scale) above ground. The building models were fabricated at a length scale of 1/200 and represented two building configurations, as shown in Figure 1. Case I comprised 12 building model configurations of a row of four square cross-section buildings with varying building height (h) and building separation (s). The building depth (d) and width (b) were fixed at 25 m at prototype scale. The height (h) was varied from 100 m (4b) to 150 m (6b), at increments of 1d, and the building separation (s) was varied from 6.25 m (0.25b) to 25 m (1b), at increments of 0.25b, to investigate the effects of h and s on the pedestrian-level wind environment. Case II comprised the 12 building configurations used for Case I, but with a podium located underneath the row of buildings to study the additional effects of the podium. The podium was 25 m (1b) tall, 187.5 m (7.5b) wide and 37.5 m (1.5b) deep. The distribution of wind speeds at pedestrian level was measured using 200 Irwin Sensors (Irwin, 1981) that were installed at a height equivalent to 2 m above ground in prototype scale (10 mm at model scale).
Figure 1: Building model configurations
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
Wind Speed analysis Mean wind speed (U) and Gust Equivalent Mean (GEM) wind speed (UGEM) were analyzed to evaluate the pedestrian-level wind environment. Mean wind speeds were used to determine the low wind speed areas where poor air ventilation may exist, while GEM wind speeds were used to indicate the high wind speed areas corresponding to conditions of potential discomfort. The GEM wind speed was defined by Lawson (1990) as the maximum of: (i) mean wind speeds and (ii) 3 seconds gust wind speed divided by a factor of 1.85. In the current study, 3 seconds gust wind speeds were determined by the „multiple extremes method‟, in which the maximum wind speeds from multiple, i.e. not less than five, one-hour samples were identified and the gust wind speed was calculated by averaging those identified maximum wind speeds (AWES Quality Assurance Manual, 2001). Test Results and Discussion Mean wind speed (U) and GEM wind speed (UGEM) were normalized by the reference mean wind speed (Ur) of the approaching wind flow at 150 m in prototype scale. At the beginning and at the end of the experiment, baseline studies, with no buildings installed, were conducted as a reference. The magnitudes of the corresponding normalized mean wind speed and normalized GEM wind speed with no buildings installed were around 0.5. Mean wind speed distribution - Effects of building separation Figure 2 shows the top view of the normalized pedestrian-level mean wind speed distribution (U/Ur), ranging from 0.0 to 1.0, on the left half of the buildings with no podium, assuming the distribution to be symmetrical about the centerline. Areas with U/Ur lower than 0.3 were designated as low wind speed zones, which correspond to mean wind speeds of around 1 to 2 m/s for an annual probability of exceedance of 50% in a wind environment such as Hong Kong. For the normalized mean wind speed distributions shown in Figure 2, the building height was fixed at 125 m (5b) and building separation (s) was varied from 0 to 25 m (1b). From the figure, four low wind speed zones were identified: i) upstream far-field low wind speed (UFLWS); ii) upstream near-field low wind speed (UNLWS); iii) downstream nearfield low wind speed (DNLWS) and iv) downstream far-field low wind speed (DFLWS). The low wind speed in the UFLWS zone is mainly due to the downwash effect at the windward face of the building that results in flow reversal at pedestrian level, approximately along the centerline upstream of the building. This creates a low wind speed zone where the approaching wind flow and opposing backflow meet. The UNLWS zone is located at the stagnant area in close proximity to the windward face of the building. In the downstream area, the DNLWS zone is due to the direct shielding effect of the building, while the DFLWS zone is due to the reattachment of the vertical recirculation behind the building, which is also affected by the strength of the horizontal recirculation. As the building separation was increased, the UFLWS zone was observed to be closer to the building and to have expanded. It is thought that the downwash created by the buildings was more dispersed and weakened by flow leakage through the building separations. As a result, the overall strength of the backflow, induced by the downwash, was weakened. In contrast, conditions within the DNLWS zone were improved due to the additional wind flowing between the buildings and penetrating into the wake region, thereby enhancing air movement and ventilation in the lee of the buildings. In the downstream far-field area, increasing the building separation to 6.25 m (0.25b) caused a detrimental effect, enlarging significantly the DFLWS zone. This is due to the flow passing through the building separations weakening the air movement created by the vertical recirculation behind the buildings. However, as the building separation was incrementally increased to 25 m (1b) the extent of the DFLWS zone was reduced. The cause of the changes
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
to the flow pattern was demonstrated by flow visualization, as shown in Figure 3. Figure 3(a) shows the flow pattern behind the buildings for s = 0 m, where the air movement was driven by the vertical recirculation and hence the direction of flow was opposite to that of the approaching wind. When a narrow separation was introduced between the buildings, as shown in Figure 3(b), no predominant wind direction was observed. This is attributed to the interaction between the two opposing flows, i.e. the backflow created by the vertical recirculation and the flow passing through the building separations. When the magnitudes of these two flows are similar, they counteract each other creating a large area of low mean wind speed and high turbulence. As the building separation was increased to 25 m (1b), as shown in Figure 3(c), the flow direction was largely the same as the direction of the approaching wind. This indicates that the flow passing through the building separations became much stronger than the backflow and dominated the air movement in the lee of the buildings, reducing the extent of the low mean wind speed area. From these measurements and observations, it is evident that the air movement at the downstream far-field area is mainly influenced by the two opposing flows: the backflow created by the vertical recirculation and the flow passing through the building separations. When one is sufficiently stronger than the other, noticeable air movements are induced which have the potential to provide reasonable air ventilation at that area. Conversely, large areas with poor ventilation potential were created for a row of buildings with small separation, s ≤ 12.5 m (0.5b).
Figure 2: Effects of building separation on the normalized mean wind speed distribution around buildings
Mean wind speed distribution - Effects of a podium The effects of a podium on the low wind speed zones are compared in Figure 4 for a fixed building height of 125 m and for building separations of 12.5 m (0.5b) and 25 m (1b). The inclusion of a podium underneath the row of buildings created a significant blockage effect which dominated the pedestrian level flow near the buildings. As a result, the extents of the UNLWS zone and DNLWS zone were increased significantly. For the downstream area, the size of the DFLWS zone in the configurations with a podium was similar to that without a podium when the building separation was small (s ≤ 12.5 m (0.5b)). For building configurations with building separation greater than 0.5b, the extent of the DFLWS zone with a podium was approximately three times larger than that without a podium. For the buildings
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
without a podium, the flow passing through the building separation was strong enough to dominate the air movement on DFLWS zone which provided higher magnitudes of U/Ur. However, for the buildings with a podium, a portion of the flow passing through the separation was blocked at the lower level. As a result, the flow was not strong enough to enhance air movement but was weakened by the backflow created by the vertical recirculation behind the buildings. Therefore, the magnitudes of U/Ur were relatively low and similar to the narrow building separation configurations. In general, a large podium is likely to adversely affect the natural air ventilation in the nearby surrounding environment.
Figure 3: Flow visualization downstream of a row of buildings: (a) s = 0 m, (b) s = 6.25 m, (c) s = 25 m
Figure 4: Effects of a podium on the normalized mean wind speed distribution around buildings
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
GEM wind speed distribution - Effects of building separation The general features of the normalized GEM wind speed distribution (UGEM/Ur) on the left half of the buildings are shown in Figure 5 to indicate zones of potential pedestrian discomfort caused by strong winds. For this study, a threshold pedestrian discomfort wind speed was set at 10 m/s for an annual probability of exceedance of 5%, which corresponds to the maximum allowable GEM wind speed for comfortable business walking in Lawson‟s comfort criteria (1975). The corresponding UGEM/Ur for Hong Kong is around 0.8 and areas where UGEM/Ur were higher than 0.8 were designated as high wind speed (HWS) zones. Figure 5 shows the normalized GEM wind speed distribution on the building configurations with different building separation (s). The building height was fixed at 125 m (6b) and the building separation was varied from 0 to 25 m (1b). For s = 0 m, a Lateral High Wind Speed (LHWS) zone can be observed and the location of maximum GEM wind speed was at the upstream corner of the building. This LHWS zone is caused by the high speed horse-shoe vortices created by the downwash from the windward face of the building and the accelerated flow around the building. The introduction of a separation between the buildings created additional high wind speed zones, identified as the Middle High Wind Speed (MHWS) zone. As the building separation was increased, the LHWS zone diminished and the MHWS zones expanded, reducing the total high wind speed area from 13.3% to 4.9% and the corresponding normalized maximum GEM wind speeds were around 1.06 to 1.15. These results indicate that a large building separation is desirable to provide overall better wind comfort conditions around buildings by reducing the total areas of the high wind speed zones. However, the usage of the spaces at the separations between buildings should be limited unless mitigation measures are implemented to ensure pedestrian comfort in those areas.
Figure 5: Effects of building separation on the normalized GEM wind speed distribution around buildings
GEM wind speed distribution - Effects of a podium In terms of pedestrian comfort, it is generally believed that a podium can protect pedestrians at ground level from strong winds that are typically created by downwash effects. Therefore, it is valuable to quantify the effects of a podium by comparing the results with
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
those for the buildings without a podium, as shown in Figure 6. Firstly, it was found that there was a very large high wind speed zone on the lateral side of the building structure and its extent was too large to be captured by the lateral coverage of the sensors. Secondly, maximum wind speeds appeared at two locations: one was near the upstream corner, similar to that for a single building configuration; the other was near the downstream corner of the podium. The magnitudes of the wind speeds at these two locations were within 3%. Compared with the buildings without the podium, the magnitudes of the maximum wind speeds around the buildings were similar. However, the podium resulted in uncomfortable winds over a larger area, which has the potential to affect more people in the surrounding environment.
Figure 6: Effects of podium on the normalized GEM wind speed distribution around buildings
Conclusions The effects of building separation and podium on natural air ventilation around a row of buildings was evaluated by analyzing the mean wind speed distribution at pedestrian level in a 1/200 scale wind tunnel model study. It was found that the air movement behind the buildings is governed by two opposing flows – the backflow created by the vertical recirculation and the flow passing through the building separations. When the backflow is much stronger, such as for zero building separation, the local wind direction is opposite to the approaching wind direction and reasonable air movement is created at the downstream farfield area. When the flow passing through the building separations is stronger, such as when the building separation is greater than half of the building width, the wind direction is the same as the approaching flow which creates reasonable air movement at both near-field and far-field areas. However, when these two opposing flows are similar in magnitude, such as for building separation that are less than half of the building width, the flow becomes highly turbulent and results in a large low wind speed zone which worsens the natural ventilation. The inclusion of a podium was found to generally adversely affect the air movement in both the downstream near-field and far-field areas by blocking the approaching wind at the pedestrian level. As a result, podia are not recommended for regions where favourable natural air ventilation conditions are required at pedestrian level.
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
In terms of pedestrian comfort, high wind speed zones occurred at the lateral sides of the row of buildings and between the buildings. It was found that a wider building separation configuration improved the pedestrian comfort conditions by reducing the total area of the high wind speed zones. The results with and without a podium indicated that the maximum wind speeds around the buildings were maintained at similar magitudes. However, the podium resulted in uncomfortable winds over a larger area, which has the potential to affect more people in the nearby surrounding environment. Acknowledgements This research project is funded by the Research Grants Council of Hong Kong Special Administrative Region, China (Project HKUST6301/04E). References Blocken B. and Carmeliet J., 2004, “Pedestrian wind Environment around buildings: Literature Review and Practical Examples”. Journal of Thermal ENV. And BLDG. SCI., Vol 28 Wiren, B.G., 1975, “A Wind Tunnel Study of Wind Velocities in Passages between and through Buildings”, In: Proceedings of the 4th International Conference on Wind Effects on Buildings and Structures, Cambridge University Press, Heathrow, pp. 465–475. Stathopoulos, T. and Storms, R, 1986, “Wind Environmental Conditions in Passages between Buildings”, Journal of Wind Engineering and Industrial Aerodynamics, 24: 19–31. Stathopoulos, T., Wu, H., 1995, “Generic models for pedestrian-level winds in built-up regions”, Journal of Wind Engineering and Industrial Aerodynamics, 54-55: 515-525. To, A. P. and Lam, K.M., 1995, “Evaluation of pedestrian-level wind environment around a row of tall buildings using a quartile-level wind speed descriptor”, Journal of Wind Engineering and Industrial Aerodynamics, 5455: 527-541. Kubota T., Miura M., Tominaga Y. and Mochida A., 2008, “Wind tunnel tests on the relationship between building density and pedestrian-level wind velocity: Development of guidelines for realizing acceptable wind environment in residential neighborhoods”. Building and Environment 40, pp1699-1708. Blocken B., Stathopoulos T., ASCE F. and Carmeliet J., 2008, “Wind Environment Conditions in Passages between Two Long Narrow Perpendicular Buildings”, Journal of Aerospace Engineering, May, 280-287. Stathopoulos, T., Wu, H. and Be´dard, C., 1992, “Wind Environment Around Buildings: A Knowledge-Based Approach”, Journal of Wind Engineering and Industrial Aerodynamics, 41–44: 2377–2388. Visser, G.T., Folkers, C.J. and Weenk, A., 2000, “KnoWind: a Database-Oriented Approach to Determine the Pedestrian Level Wind Environment Around Buildings”, Journal of Wind Engineering and Industrial Aerodynamics, 87: 287–299. Chan, A.T., So, E.S.P. and Samad, S.C., 2001, “Strategic Guidelines for Street Canyon Geometry to Achieve Sustainable Street Air Quality”, Atmospheric Environment, 35(24): 4089–4098. Tsang, C.W., Kwok, K.C.S. and Hitchcock, P.A., 2009, “Effects of building dimensions and building separations on pedestrian-level wind environment”, Proceedings of the 5th European and African Conference on Wind Enigneerin, Florence Italy, July 19-23, 2009. Irwin, P.A., 1981, “A simple omnidirectional probe for the measurement of pedestrian level winds”, Journal of Wind Engineering and Industrial Aerodynamics, vol7, pp219-239. Lawson, T.V., 1990,” The Determination of the wind environment of a building complex before construction”, Department of Aerospace Engineering, University of Bristol, Report Number TVL 9025. Australasian Wind Engineering Society, Quality Assurance Manual for Cladding and Environmental Wind Studies, 2001. Lawson, T.V., 1975, “The determination of the wind environment of a building complex before construction”, Bristol University, Department of Aeronautical Engineering.
EFFECTS OF BUILDING SEPARATION AND PODIUM ON PEDESTRIAN-LEVEL WIND ENVIRONMENT C.W. Tsang1, K.C.S. Kwok2 and P.A. Hitchcock3 CLP Power Wind/Wave Tunnel Facility, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong S.A.R., P.R. China, [email protected] 2 School of Engineering, University of Western Sydney, Sydney, New South Wales, Australia, [email protected] 3 CLP Power Wind/Wave Tunnel Facility, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong S.A.R., P.R. China, [email protected] 1
ABSTRACT High-rise buildings affect the surrounding pedestrian-level wind environment. In recent years, awareness and concern have increased about the creation of low wind speed areas around buildings which may lead to poor outdoor air ventilation. Moreover, many modern building developments are not restricted to a single building but may comprise a group of buildings. Very few systematic studies have focused on the low wind speed areas around a group of buildings. In this research, a series of parametric wind tunnel studies was carried out to investigate the effects of building separation and podium on the pedestrian-level wind environment around a group of tall buildings. Mean wind speeds were used to determine the low wind speed areas where poor air ventilation may exist, and Gust Equivalent Mean (GEM) wind speeds were used to identify high wind speed areas with the potential to cause discomfort under strong wind conditions. KEYWORDS: PEDESTRIAN-LEVEL WIND ENVIRONMENT, BUILDING SEPARATION, PODIUM
Introduction In many densely populated cities, such as Hong Kong, urban renewal is an important aspect of sustainable development for the community to make the best use of available land and infrastructure. Under the renewal projects, modern high-rise buildings have been built inside the closely packed old districts. Due to the significant changes to the building forms, community concerns and awareness have been raised of the potential effects of newly built structures on the surrounding wind environment, particularly for tall and bulky buildings that, when closely packed together, form undesirable barriers which obstruct winds from penetrating into the downstream urban fabric and which result in poor natural air ventilation. However, community concerns and the corresponding solutions suggested by the designers and engineers are mostly based on personal and professional experience which may be subjective. There is a genuine lack of solid scientific literature on which to base arguments and design decisions. Wind flow patterns around buildings are very complicated and have been investigated for more than 30 years. A comprehensive review focusing on the flow features and pedestrian-level wind environment around buildings was conducted by Blocken and Carmeliet (2004). Since the 1960s, outdoor human comfort has been given a significant amount of attention. Wind environment studies have focused on pedestrian-level discomfort caused by strong winds near idealized buildings (Wiren 1975, Stathopoulos and Storms 1986, Stathopoulos and Wu 1995, To and Lam 1995, Blocken et al. 2008 and others). In those studies, the effects of building characteristics, such as dimensions, shape, building separation,
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
orientations of building groups and street canyons, have been investigated, from which knowledge-based expert systems were developed (Stathopoulos et al. 1992, Visser et al. 2000). These systems allow a preliminary and simplified evaluation of high wind speed areas around buildings. However, very few systematic studies have focused on the low wind speed areas around a group of buildings (Stathopoulos and Wu 1995, Chan et al. 2001, Kubota et al. 2008, Tsang et al. 2009). Tsang et al. (2009) have conducted sets of parametric studies to invesigate the effects of building dimensions and building separation on the low wind speed areas at pedestrian level. This paper presents the results of an extension of those studies for a single building and a pair of buildings to a row of four identical buildings with and without a podium. The pedestrian-level wind environments around these building configurations were investigated in a wind tunnel model study. The natural air ventilation and pedestrian comfort are both evaluated. Experimental Arrangement The experiments were carried out in the 3 m 2 m 29 m long high-speed test section of the CLP Power Wind/Wave Tunnel Facility (WWTF) at The Hong Kong University of Science and Technology. The mean wind speed profile of the approaching turbulent wind flow followed a power law exponent of 0.2, using a series of fences and roughness blocks, to simulate wind flow above a typical suburban terrain. All of the wind tunnel tests were conducted using a reference mean wind speed Ur of approximately 10 m/s at 150 m (in prototype scale) above ground. The building models were fabricated at a length scale of 1/200 and represented two building configurations, as shown in Figure 1. Case I comprised 12 building model configurations of a row of four square cross-section buildings with varying building height (h) and building separation (s). The building depth (d) and width (b) were fixed at 25 m at prototype scale. The height (h) was varied from 100 m (4b) to 150 m (6b), at increments of 1d, and the building separation (s) was varied from 6.25 m (0.25b) to 25 m (1b), at increments of 0.25b, to investigate the effects of h and s on the pedestrian-level wind environment. Case II comprised the 12 building configurations used for Case I, but with a podium located underneath the row of buildings to study the additional effects of the podium. The podium was 25 m (1b) tall, 187.5 m (7.5b) wide and 37.5 m (1.5b) deep. The distribution of wind speeds at pedestrian level was measured using 200 Irwin Sensors (Irwin, 1981) that were installed at a height equivalent to 2 m above ground in prototype scale (10 mm at model scale).
Figure 1: Building model configurations
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
Wind Speed analysis Mean wind speed (U) and Gust Equivalent Mean (GEM) wind speed (UGEM) were analyzed to evaluate the pedestrian-level wind environment. Mean wind speeds were used to determine the low wind speed areas where poor air ventilation may exist, while GEM wind speeds were used to indicate the high wind speed areas corresponding to conditions of potential discomfort. The GEM wind speed was defined by Lawson (1990) as the maximum of: (i) mean wind speeds and (ii) 3 seconds gust wind speed divided by a factor of 1.85. In the current study, 3 seconds gust wind speeds were determined by the „multiple extremes method‟, in which the maximum wind speeds from multiple, i.e. not less than five, one-hour samples were identified and the gust wind speed was calculated by averaging those identified maximum wind speeds (AWES Quality Assurance Manual, 2001). Test Results and Discussion Mean wind speed (U) and GEM wind speed (UGEM) were normalized by the reference mean wind speed (Ur) of the approaching wind flow at 150 m in prototype scale. At the beginning and at the end of the experiment, baseline studies, with no buildings installed, were conducted as a reference. The magnitudes of the corresponding normalized mean wind speed and normalized GEM wind speed with no buildings installed were around 0.5. Mean wind speed distribution - Effects of building separation Figure 2 shows the top view of the normalized pedestrian-level mean wind speed distribution (U/Ur), ranging from 0.0 to 1.0, on the left half of the buildings with no podium, assuming the distribution to be symmetrical about the centerline. Areas with U/Ur lower than 0.3 were designated as low wind speed zones, which correspond to mean wind speeds of around 1 to 2 m/s for an annual probability of exceedance of 50% in a wind environment such as Hong Kong. For the normalized mean wind speed distributions shown in Figure 2, the building height was fixed at 125 m (5b) and building separation (s) was varied from 0 to 25 m (1b). From the figure, four low wind speed zones were identified: i) upstream far-field low wind speed (UFLWS); ii) upstream near-field low wind speed (UNLWS); iii) downstream nearfield low wind speed (DNLWS) and iv) downstream far-field low wind speed (DFLWS). The low wind speed in the UFLWS zone is mainly due to the downwash effect at the windward face of the building that results in flow reversal at pedestrian level, approximately along the centerline upstream of the building. This creates a low wind speed zone where the approaching wind flow and opposing backflow meet. The UNLWS zone is located at the stagnant area in close proximity to the windward face of the building. In the downstream area, the DNLWS zone is due to the direct shielding effect of the building, while the DFLWS zone is due to the reattachment of the vertical recirculation behind the building, which is also affected by the strength of the horizontal recirculation. As the building separation was increased, the UFLWS zone was observed to be closer to the building and to have expanded. It is thought that the downwash created by the buildings was more dispersed and weakened by flow leakage through the building separations. As a result, the overall strength of the backflow, induced by the downwash, was weakened. In contrast, conditions within the DNLWS zone were improved due to the additional wind flowing between the buildings and penetrating into the wake region, thereby enhancing air movement and ventilation in the lee of the buildings. In the downstream far-field area, increasing the building separation to 6.25 m (0.25b) caused a detrimental effect, enlarging significantly the DFLWS zone. This is due to the flow passing through the building separations weakening the air movement created by the vertical recirculation behind the buildings. However, as the building separation was incrementally increased to 25 m (1b) the extent of the DFLWS zone was reduced. The cause of the changes
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
to the flow pattern was demonstrated by flow visualization, as shown in Figure 3. Figure 3(a) shows the flow pattern behind the buildings for s = 0 m, where the air movement was driven by the vertical recirculation and hence the direction of flow was opposite to that of the approaching wind. When a narrow separation was introduced between the buildings, as shown in Figure 3(b), no predominant wind direction was observed. This is attributed to the interaction between the two opposing flows, i.e. the backflow created by the vertical recirculation and the flow passing through the building separations. When the magnitudes of these two flows are similar, they counteract each other creating a large area of low mean wind speed and high turbulence. As the building separation was increased to 25 m (1b), as shown in Figure 3(c), the flow direction was largely the same as the direction of the approaching wind. This indicates that the flow passing through the building separations became much stronger than the backflow and dominated the air movement in the lee of the buildings, reducing the extent of the low mean wind speed area. From these measurements and observations, it is evident that the air movement at the downstream far-field area is mainly influenced by the two opposing flows: the backflow created by the vertical recirculation and the flow passing through the building separations. When one is sufficiently stronger than the other, noticeable air movements are induced which have the potential to provide reasonable air ventilation at that area. Conversely, large areas with poor ventilation potential were created for a row of buildings with small separation, s ≤ 12.5 m (0.5b).
Figure 2: Effects of building separation on the normalized mean wind speed distribution around buildings
Mean wind speed distribution - Effects of a podium The effects of a podium on the low wind speed zones are compared in Figure 4 for a fixed building height of 125 m and for building separations of 12.5 m (0.5b) and 25 m (1b). The inclusion of a podium underneath the row of buildings created a significant blockage effect which dominated the pedestrian level flow near the buildings. As a result, the extents of the UNLWS zone and DNLWS zone were increased significantly. For the downstream area, the size of the DFLWS zone in the configurations with a podium was similar to that without a podium when the building separation was small (s ≤ 12.5 m (0.5b)). For building configurations with building separation greater than 0.5b, the extent of the DFLWS zone with a podium was approximately three times larger than that without a podium. For the buildings
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
without a podium, the flow passing through the building separation was strong enough to dominate the air movement on DFLWS zone which provided higher magnitudes of U/Ur. However, for the buildings with a podium, a portion of the flow passing through the separation was blocked at the lower level. As a result, the flow was not strong enough to enhance air movement but was weakened by the backflow created by the vertical recirculation behind the buildings. Therefore, the magnitudes of U/Ur were relatively low and similar to the narrow building separation configurations. In general, a large podium is likely to adversely affect the natural air ventilation in the nearby surrounding environment.
Figure 3: Flow visualization downstream of a row of buildings: (a) s = 0 m, (b) s = 6.25 m, (c) s = 25 m
Figure 4: Effects of a podium on the normalized mean wind speed distribution around buildings
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
GEM wind speed distribution - Effects of building separation The general features of the normalized GEM wind speed distribution (UGEM/Ur) on the left half of the buildings are shown in Figure 5 to indicate zones of potential pedestrian discomfort caused by strong winds. For this study, a threshold pedestrian discomfort wind speed was set at 10 m/s for an annual probability of exceedance of 5%, which corresponds to the maximum allowable GEM wind speed for comfortable business walking in Lawson‟s comfort criteria (1975). The corresponding UGEM/Ur for Hong Kong is around 0.8 and areas where UGEM/Ur were higher than 0.8 were designated as high wind speed (HWS) zones. Figure 5 shows the normalized GEM wind speed distribution on the building configurations with different building separation (s). The building height was fixed at 125 m (6b) and the building separation was varied from 0 to 25 m (1b). For s = 0 m, a Lateral High Wind Speed (LHWS) zone can be observed and the location of maximum GEM wind speed was at the upstream corner of the building. This LHWS zone is caused by the high speed horse-shoe vortices created by the downwash from the windward face of the building and the accelerated flow around the building. The introduction of a separation between the buildings created additional high wind speed zones, identified as the Middle High Wind Speed (MHWS) zone. As the building separation was increased, the LHWS zone diminished and the MHWS zones expanded, reducing the total high wind speed area from 13.3% to 4.9% and the corresponding normalized maximum GEM wind speeds were around 1.06 to 1.15. These results indicate that a large building separation is desirable to provide overall better wind comfort conditions around buildings by reducing the total areas of the high wind speed zones. However, the usage of the spaces at the separations between buildings should be limited unless mitigation measures are implemented to ensure pedestrian comfort in those areas.
Figure 5: Effects of building separation on the normalized GEM wind speed distribution around buildings
GEM wind speed distribution - Effects of a podium In terms of pedestrian comfort, it is generally believed that a podium can protect pedestrians at ground level from strong winds that are typically created by downwash effects. Therefore, it is valuable to quantify the effects of a podium by comparing the results with
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
those for the buildings without a podium, as shown in Figure 6. Firstly, it was found that there was a very large high wind speed zone on the lateral side of the building structure and its extent was too large to be captured by the lateral coverage of the sensors. Secondly, maximum wind speeds appeared at two locations: one was near the upstream corner, similar to that for a single building configuration; the other was near the downstream corner of the podium. The magnitudes of the wind speeds at these two locations were within 3%. Compared with the buildings without the podium, the magnitudes of the maximum wind speeds around the buildings were similar. However, the podium resulted in uncomfortable winds over a larger area, which has the potential to affect more people in the surrounding environment.
Figure 6: Effects of podium on the normalized GEM wind speed distribution around buildings
Conclusions The effects of building separation and podium on natural air ventilation around a row of buildings was evaluated by analyzing the mean wind speed distribution at pedestrian level in a 1/200 scale wind tunnel model study. It was found that the air movement behind the buildings is governed by two opposing flows – the backflow created by the vertical recirculation and the flow passing through the building separations. When the backflow is much stronger, such as for zero building separation, the local wind direction is opposite to the approaching wind direction and reasonable air movement is created at the downstream farfield area. When the flow passing through the building separations is stronger, such as when the building separation is greater than half of the building width, the wind direction is the same as the approaching flow which creates reasonable air movement at both near-field and far-field areas. However, when these two opposing flows are similar in magnitude, such as for building separation that are less than half of the building width, the flow becomes highly turbulent and results in a large low wind speed zone which worsens the natural ventilation. The inclusion of a podium was found to generally adversely affect the air movement in both the downstream near-field and far-field areas by blocking the approaching wind at the pedestrian level. As a result, podia are not recommended for regions where favourable natural air ventilation conditions are required at pedestrian level.
The Seventh Asia-Pacific Conference on Wind Engineering, November 8-12, 2009, Taipei, Taiwan
In terms of pedestrian comfort, high wind speed zones occurred at the lateral sides of the row of buildings and between the buildings. It was found that a wider building separation configuration improved the pedestrian comfort conditions by reducing the total area of the high wind speed zones. The results with and without a podium indicated that the maximum wind speeds around the buildings were maintained at similar magitudes. However, the podium resulted in uncomfortable winds over a larger area, which has the potential to affect more people in the nearby surrounding environment. Acknowledgements This research project is funded by the Research Grants Council of Hong Kong Special Administrative Region, China (Project HKUST6301/04E). References Blocken B. and Carmeliet J., 2004, “Pedestrian wind Environment around buildings: Literature Review and Practical Examples”. Journal of Thermal ENV. And BLDG. SCI., Vol 28 Wiren, B.G., 1975, “A Wind Tunnel Study of Wind Velocities in Passages between and through Buildings”, In: Proceedings of the 4th International Conference on Wind Effects on Buildings and Structures, Cambridge University Press, Heathrow, pp. 465–475. Stathopoulos, T. and Storms, R, 1986, “Wind Environmental Conditions in Passages between Buildings”, Journal of Wind Engineering and Industrial Aerodynamics, 24: 19–31. Stathopoulos, T., Wu, H., 1995, “Generic models for pedestrian-level winds in built-up regions”, Journal of Wind Engineering and Industrial Aerodynamics, 54-55: 515-525. To, A. P. and Lam, K.M., 1995, “Evaluation of pedestrian-level wind environment around a row of tall buildings using a quartile-level wind speed descriptor”, Journal of Wind Engineering and Industrial Aerodynamics, 5455: 527-541. Kubota T., Miura M., Tominaga Y. and Mochida A., 2008, “Wind tunnel tests on the relationship between building density and pedestrian-level wind velocity: Development of guidelines for realizing acceptable wind environment in residential neighborhoods”. Building and Environment 40, pp1699-1708. Blocken B., Stathopoulos T., ASCE F. and Carmeliet J., 2008, “Wind Environment Conditions in Passages between Two Long Narrow Perpendicular Buildings”, Journal of Aerospace Engineering, May, 280-287. Stathopoulos, T., Wu, H. and Be´dard, C., 1992, “Wind Environment Around Buildings: A Knowledge-Based Approach”, Journal of Wind Engineering and Industrial Aerodynamics, 41–44: 2377–2388. Visser, G.T., Folkers, C.J. and Weenk, A., 2000, “KnoWind: a Database-Oriented Approach to Determine the Pedestrian Level Wind Environment Around Buildings”, Journal of Wind Engineering and Industrial Aerodynamics, 87: 287–299. Chan, A.T., So, E.S.P. and Samad, S.C., 2001, “Strategic Guidelines for Street Canyon Geometry to Achieve Sustainable Street Air Quality”, Atmospheric Environment, 35(24): 4089–4098. Tsang, C.W., Kwok, K.C.S. and Hitchcock, P.A., 2009, “Effects of building dimensions and building separations on pedestrian-level wind environment”, Proceedings of the 5th European and African Conference on Wind Enigneerin, Florence Italy, July 19-23, 2009. Irwin, P.A., 1981, “A simple omnidirectional probe for the measurement of pedestrian level winds”, Journal of Wind Engineering and Industrial Aerodynamics, vol7, pp219-239. Lawson, T.V., 1990,” The Determination of the wind environment of a building complex before construction”, Department of Aerospace Engineering, University of Bristol, Report Number TVL 9025. Australasian Wind Engineering Society, Quality Assurance Manual for Cladding and Environmental Wind Studies, 2001. Lawson, T.V., 1975, “The determination of the wind environment of a building complex before construction”, Bristol University, Department of Aeronautical Engineering.
