AA E+E Environmental & Energy Studies Programme Architectural Association School of Architecture Graduate School
MSc Sustainable Environmental Design Dissertation project 2011 -‐2012
the impact of window design on the environmental performance of work environments in são paulo
João Pinto de Oliveira Cotta September 2012
abstract The windows are the transparent part of the building envelope and the most visual elements on the façade. They also play a major role in the architectural composition, whether they are distributed as sequence of apertures on the opaque envelope or applied as a curtain wall, where the glazing becomes the wall itself. The window has a direct relation with the climate and also with the occupants since it can provide the opportunity for them to adjust the indoor conditions to suit their needs, potentially increasing their level of satisfaction. The topic of window design in this dissertation started to be studied without a clear idea regarding the broadness of the subject. Nevertheless, through the research process the window element revealed to have a direct relation with an endless range of aspects that can impact on the environmental performance and comfort provision in working environments. Thus, in pursuit of covering most of these aspects, a holistic analytic approach was adopted, attempting to analyse different comfort constituents separately and then synthesize the findings. At the end, the research revealed that a proper design can result in effective improvement in the qualitative environmental aspects while providing significant energy savings. The context chosen for this study was São Paulo, where fully glazed office buildings, completely dependent on air conditioning and artificial lighting have marked the most common typologies of commercial buildings. Although energy consumption cannot be associated with carbon emissions in Brazil since power is largely supplied from hydroelectricity, considering the existing scenario where growth pressures conflict with the limited capacity of energy production, buildings that rely less on the existing grid become more valuable. Buildings are responsible for 48,3% of total electrical energy consumption in Brazil, 23% of which commercial and public buildings are accounted for. The importance of this proportion becomes even more relevant in São Paulo which hosts the highest amount of commercial building in the country. Ultimately, this dissertation through the investigation of window design, aims to provide guidance for the applicability of bioclimatic techniques in order to achieve low energy and environmentally responsive work places.
i
ii
authorship declaration form AA E+E ENVIRONMENT & ENERGY STUDIES PROGRAMME ARCHITECTURAL ASSOCIATION SCHOOL OF ARCHITECTURE GRADUATE SCHOOL PROGRAMME MSc Sustainable Environmental Design SUBMISSION Dissertation Project 2011-12 TITLE The Impact of Window Design in the Environmental Performance of Work Environments in S達o Paulo. NUMBER OF WORDS (excluding footnotes and references) 16,952 STUDENT NAME Jo達o Pinto de Oliveira Cotta DECLARATION I certify that the contents of this document are entirely my own work and that any quotation or paraphrase from the published or unpublished work of others is duly acknowledged. SIGNATURE
DATE September 14th, 2012
iii
iv
table of contents 1 Introduction 1.1 Methodology
1 4
2 Theoretical Background 2.1 Window definition 2.2 Window configuration and daylighting in working environments 2.3 Building orientation and aperture components 2.4 Window design and natural ventilation 2.5 Conclusions
7 9 9 11 14 18
3 Precedents/Field Work 3.1 Introduction 3.2 Thermal comfort 3.3 Natural ventilation studies 3.4 Findings
21 23 24 26 28
4 S達o Paulo Context and Climate 4.1 S達o Paulo context 4.2 S達o Paulo climate 4.3 Jundiai City Hall field work
31 33 33 36
5 Analytic Work 5.1 Preparing the shoebox 5.1.1 The plan 5.1.2 The fa巽ades 5.2 Design Proposal: Shading devices 5.2.1 North 5.2.2 South 5.2.3 East 5.2.4 West 5.3 Design Verification: Daylight simulations 5.3.1 North-south orientation 5.3.2 East-west orientation 5.4 Design Verification: Thermal simulations 5.4.1 Without occupancy 5.4.2 With occupancy + Ventilation 5.4.3 Winter studies 5.5 CFD analysis and comfort assessment 5.5.1 North 5.5.2 South 5.5.3 East 5.5.4 West 5.5.5 Conclusions
47 47 47 48 49 50 51 52 53 54 54 56 57 58 62 66 67 68 72 76 80 83
6 Research Outcomes and Applicability 6.1 Research Outcomes 6.2 Applicability 6.2.1 Cost Analysis
85 86 88 89
7 Conclusions
97
Bibliography
99
Appendices
103
v
vi
acknowledgements I would like to show my sincere gratitude towards the entire staff of the Environment & Energy programme for their support throughout the course, especially to Professor Simos Yannas, for his essential guidance in this dissertation and for his patience and care while teaching me the principles of environmental architecture, which I will carry for my entire life. It was an honour to have worked with him. Special thanks are also to be offered to Joana Carla Soares Gonรงalves who always motivated me and demonstrated to be a reliable friend since the first day of this course, always giving me valuable advices. I would like to acknowledge my SED colleagues Sandra Mayumi Morikawa, Meital Ben Dayan, Nikhil Deotarase and Rodolfo Pedro Augspach, who were always so supportive, and especially Bilge Kobas, with whom I have had essential discussions regarding the window design topic. I am also grateful for the architect Araken Martinho, who provided the opportunity of visiting and conducting a field work in an important case study. Additionally, I would like to acknowledge Annabel Rootes from Rogers Stirk Habour + Partners, for her help during the field work and my colleagues who also participated in this study: Ronak Gawarwala, Tomas Swett, Izza Salim, Valli Chidambaram, Nikhil Deotarase and Laura Vasquez. Finally, I would like to express my profound gratitude to my family for their unconditional support and to Mariana Fagundes for her constant encouragement.
vii
viii
1 introduction
1
1 introduction The invention of air conditioning system by Willis Carries in 1928 combined with the improvements in the artificial lighting technologies had a great impact and influence on the design solutions for office buildings and working environments. In this context, a universal style of architecture emerged, where building design could be completely independent of the exterior climate, relying all year round on artificial heating, cooling and lighting to provide comfort. Since then, the technical aspects of building design for different climates and contexts have been neglected, and an architectural style driven by the intention of exploring the building form and materiality, has come into view. This architectural approach for commercial buildings is still prevalent in growing cities like São Paulo, and it is characterized by fully glazed façades with the absence of external shading devices and deep floor plate. The fully glazed transparent aesthetic, in combination with the tallness of commercial buildings, has a powerful effect in creating landmarks in the cities, thus defining a strong identity where the buildings are seen as symbols of technological advancement, prestige and prosperity. Nevertheless, in regards to the building performance, the benefits of good views and high daylight levels around the perimeter of the plan are offset by heat losses, increased glare and overheating due to excessive solar gains even in temperate climates. Hence, the dependence on air conditioning systems throughout the year is one of the main consequences of this trend which engenders a wide range of environmental impacts. In a global scale, this widespread disregard with the climate in the design process is leading to a significant increase in energy demand in the building sector, which is now responsible for 60% of the world’s total energy consumption (Barker et al, 2007). In the occupant’s perspective, problems concerning air quality and limited possibilities for personal control on temperature not only affect well-being but they are also the major causes of Sick Building Syndrome The most common symptoms found in sealed buildings with no operable windows are: dry eyes, dry throat, stuffy nose, itchy eyes, lethargy and headache (Hedge et al, 1989). Additionally, different fieldworks conducted in environmentally-controlled office buildings, found a rate of 50 percent of people slightly dissatisfied or very dissatisfied with their thermal comfort, well below the 20 percent PPD goal recommended by ASHRAE (Choi, J et al. 2010). Furthermore, a study of office buildings in Lisbon showed that even in hotter climates, occupants’ satisfaction in some air conditioned buildings may be significantly less than in some naturally ventilated ones (Baker, N 2009). This way, any effort towards avoiding the use of air conditioning, in pursuit of implanting passive cooling strategies, can be an advantage not only in terms of energy saving, but also in terms of having better environmental quality. In the context of naturally ventilated buildings, the presence of adaptive opportunities is a key factor that can increase occupants’ satisfaction even when the conventional comfort standards are not met. Studies have shown that the ability of the occupant to make changes in their personal conditions or in their environment, through the use of several attributes, can increase their toleration in regards to temperature excursions up to 5°C above the conventional upper temperature limits and around 3°C below conventional lower limits (Baker and Standeven, 1994). This statement is illustrated in the figure 1.1, which shows that much higher and lower temperatures are considered acceptable in free running buildings in comparison to air-conditioned ones.
3
Fig 1.1 Comfort temperatures as a function of outdoor temperature for buildings which are freerunning (A) and with heating and cooling (B). From the ASHREA database (de Dear and Brager 1998) (left) and from Humphreys (1978) Source: Nicol and Humphreys (2002)
Amongst the adaptive attributes, operable window and its components (e.g. shading devices) play an important role in enhancing the environmental quality of workplaces. There is currently growing evidence that the presence of daylight and views from windows are valued by the occupants and can increase the sense of well-being and productivity (Baker, N 2009). This fact was also confirmed by a recent survey conducted in office buildings where 80 per cent of the staff said that they wanted to sit by an operable window (CIBSE 1999). Moreover, a research found that natural ventilation and improved daylight design can enhance employees’ productivity up to 28 per cent (Loftness & Hartkopf, 2003). Finally, one can conclude that this direct relation with the buildings’ occupants increases the importance of the window element in this context of working environments even more. Thus, this dissertation will focus on the window design, having it as a central pillar for the investigation of daylight, natural ventilation and thermal performance of working environments in São Paulo.
1.1 Methodology The structure of this dissertation is as follows: First of all, all the potential issues related to the window design with reference to warm humid climates (São Paulo’s climate) were investigated based on a literature review. Secondly, a field work conducted in Rogers Stirk Harbour’s office in London during a summer day; provided additional useful information regarding the issues associated with window operations and its relation with the internal layouts. Thirdly, another field work was conducted in a natural ventilated office building located near São Paulo and which was useful to provide an identification of the specific climatic and operational issues that can affect design decisions regarding windows. Afterwards, all the aspects observed and research until thus far were carried to the analytic work chapter. In order to have a holistic investigation, the impact of the windows on the building’s thermal performance, daylight availability and ventilation conditions were examined. This thorough research helped to reveal how window design affects the space, by quantitative and qualitative means. Finally, a cost analysis was carried out to assess the applicability of the proposed façade scenario under realistic conditions, in comparison to a conventional fully glazed office building model. 4
5
6
2 theoretical background 2.1 Window definition 2.2 Window configuration and daylighting in working environments 2.3 Building orientation and aperture components 2.4 Window design and natural ventilation 2.5 Conclusions
7
8
2 theoretical background 2. 1 Window Definition The window is defined by Baker and Steemers (2002) as the transparent part of the building envelope, and the most visual element on the façade. It also plays a major role in the architectural composition, whether it is distributed as sequence of apertures on the opaque envelope or a curtain wall, where the glazing becomes the wall itself. The functions of windows are numerous in any climatic context, since they provide visual contact with outdoors, daylight, sun penetration in the winter and ventilation. This way it is possible to realize that the window is the most vital link with climate and its size, position and orientation are key determinant factors in the design of more environmentally responsive buildings.
2.2 Window configuration and daylighting in working environments As previously mentioned, one of the main issues that affects the environmental performance of office buildings nowadays, is the lack of criteria in the design process concerning the depth of the floor plan. Consequently, a considerable amount of buildings have their plans too deep to be daylit, therefore relying on artificial lighting, which is the second largest use of electricity in commercial buildings, after air conditioning (Gonçalves 2010). In addition, artificial lighting usually represents a considerable portion of offices’ internal heat gains, being another reason to be avoided especially in warm climates. Different authors mention that as a rule of thumb, the depth of the floor plan should follow a proportion of twice the distance of the floor to ceiling height in order to guarantee acceptable levels of daylight. Nevertheless, Baker and Steemers (2002) are more specific and relate this ratio to the arrangement of windows. As shown in the Figure 2.1, they suggest that the depth of useful levels of daylight should be approximately two times the distance from the floor to the top of the window opening. Hence, it is clear that tall windows can allow a better daylight penetration at the back of the room by increasing the range of sky component (view from the sky from one point in the room). However, other factors such as the obstruction angle, the type of glass and the presence of shading devices might considerably affect the daylight levels and in most cases a more accurate verification must be performed in order to define the daylit zone.
Fig 2.1 Daylight profile of side-lit room Source: Baker and Steemers (2002)
9
The concept of quality in regards to daylight in working environments is not about the quantity of daylight, but primarily concerning daylight distribution on the work plane (900 mm above the floor) and avoiding glare (Baker and Steemers, 2002). It is known that an uneven daylight distribution can increase the probability of the occupants to turn the lights on, even during daytime, since the accentuated contrast of luminance can create a perception that the darker areas in the room are darker than they actually are. In this context, windows’ proportion and location will have a significant impact on the energy performance of buildings and the quality of working spaces. Figure 2.2 shows the results of a computer simulation where different window heights were assessed in relation to daylight levels in a room which is 7.2 m deep. It is evident that the room is too deep for providing acceptable levels of daylight at the back, nevertheless the only scenario where the max/min ratio approaches 10, which is considered as an acceptable value of distribution, is the one with the highest window. This way, one can conclude that the distribution of daylight is improved by higher windows.
Fig 2.2 Comparison of distribution of high and low openings (SUPERLIT simulation) Source: Baker and Steemers (2002)
In parallel, especially in working environments, the availability of views should also be considered when deciding windows’ position and configuration. In this case, higher windows will not be an advantage if they exclude the possibility of having visual contact with outdoor. Wells (1996), who interviewed 2500 employees in different office buildings, reported that 89 per cent of the subjects felt that a view out was very important. In addition, Osterhaus (2001) after extensive research in nine daylit office buildings in the USA and Germany, described that visual contact with outdoor was so important to the occupants, that glare was less of an issue or was even ignored in the west and east-facing offices when pleasant views from the windows were available. Furthermore, Ne’eman and Hopkinson (1970) after an experiment conducted with 318 occupants of 3 buildings in the UK concluded that the type of outside view and the view content was a more significant factor in the choice of the preferred window type, than daylight availability and the sun position. In this case, views triggered the choice for wider windows, as they could provide higher occupant satisfaction. Similar results were also found by Keighley (1973), who reported that large horizontal windows that were 25 per cent of the wall area were the most appreciated by subjects in his research since they provide the best access to skyline and horizon views.
10
Hence, it is possible to conclude that a combination of wide horizontal windows in the level of the work plane, with clerestory windows adjacent to the ceiling, can be the most suitable in working environments located in sites which offer a possibility of having good external views. This way the full potential of the daylit zone depth can be also developed, as seen in the Figure 2.3. However, it is important to highlight that some aperture components such as fixed shading devices can be a disadvantage in terms of allowing external views, though they must be considered as part of the window systems, especially in warm climates since they have an important function of avoiding direct solar radiation from entering into a room.
Fig 2.3 Domino Haus, Reutlinger Germany. Clarestory lighting developing the full potential of the daylit zone. Source: Baker and Steemers (2002)
2.3 Building orientation and aperture components According to Givoni (1994), when considering the orientation of a building, the main issue is the orientation of the windows, since they are the most energy-transmissive element in the envelope and also can determine the air flow inside the building. It is known that direct solar radiation through large glass panes can elevate the indoor temperature high above the outdoor daytime level, even in buildings with low internal heat gains. By contrast, due to the amount of appliances, artificial lights and occupants’ density, the conventional office building is characterized by high internal heat gains. In this case and specifically in some climates, direct solar radiation can be undesirable even in the winter, enhancing the need for proper solar control. Yannas (2000) recommends that where there is choice over orientation, facing the glazed openings due north (for the north hemisphere), is more appropriate for buildings where solar gains are not desirable but which daylight is required and views are valuable. This way it is possible to considerably reduce or even eliminate (in some locations) the use of further means of solar control since this orientation receives lower incidence of solar radiation throughout the year. In addition, Yannas (2011) also mentions the importance of having solar control applied to all exposed orientations in hot climates, with the purpose of blocking unwanted solar gains. Hence, it is possible to understand that more important than the orientation of the building, is the level of exposure of the glazing areas.
11
The design of the proper solar control will have a great influence in the daylight performance of the building as it can contribute to redistributing the sunlight within the room and also to avoid problems such as glare and sunlight falling directly onto occupants, which could result in an increase in the physiological equivalent temperature of between 3°C and 7°C (Baker and Steemers, 2002). As shown in the Figure 2.4, one of the main consequences concerning the absence of an effective solar control in working environments is that during sunny days, even in cold climates the occupants will tend to put the blinds down and turn the lights on, significantly reducing the quality of the internal spaces and many times conflicting with the opening windows and ventilation.
Fig 2.4 Interior with blinds down and lights on Source: Baker and Steemers (2002)
One of the first questions that usually arise when adopting shading devices is whether to place it internally or externally. According to Yannas: “External shading is preferred. Obstruction of the incident radiation before it reaches building surfaces and is admitted indoors is a far more effective way of solar control than the use of internal blinds. Even when highly reflective the latter will let more radiation inside leading to a higher rise in indoor temperature. However, internal blinds or curtains may be sufficient for spaces of low solar control priority” (2000). This statement is illustrated in the Figure 2.5, which shows that a typical solar gain factor for external white louvers is 12 percent, while it is 46 percent for internal white louvers. In addition, Figure 2.6 shows an example of a tall office building that adopted external louvers as solar control strategy, which is not a very common solution nowadays, in this kind of typology.
12
Fig 2.5 Comparison between external and internal shading Source: Baker,N (2009)
Fig 2.6 New York Times Headquarters Source: www.aviewoncities. com
In regards to external shading devices Yannas (2000) also specifies that for south orientation overhangs and horizontal louvers will be the most effective solutions, on the other hand for the other orientations vertical ones are the most appropriate. However depending on the design of the shading device, especially on east and west facing orientations views might be compromised. It is known that the most efficient external shading devices in terms of providing a good quality of daylight for the interior spaces are the ones which besides reducing the amount of radiation transmitted through windows, also redistribute the direct sunlight, creating a more even distribution of light within the room. A lightshelf which can be applied to south orientations is a good example of this kind of solution that is based on the principle of reflection of light from surfaces. As seen in the Figure 2.7 the lightshelf reduces the illumination close to the window, though the daylight levels at the back of the room are kept within the acceptable levels.
Fig 2.7 The daylight profile of a side-lit room Source: Baker, N (2009)
13
Finally, one can conclude that, a due north or south orientation offer more possibilities to the designer to create solar control solutions, where daylight quality or views are not impaired. Hence, where there is a choice over the orientation of offices, facing the glazing areas due east or west should be avoided, since the solar angle is too low in those orientations and solar incident radiation in the vertical plane can have the highest summer values depending on the latitude, this way increasing the difficulty of designing an efficient solar control. According to Yannas (2000), if openings need to be placed on such orientations, the glazing elements must be placed with moderation.
2.4 Window design and natural ventilation Ventilation is a fundamental requirement for indoor air quality and health in buildings. According to Givoni (1994), in warm humid climates openings play an important role in determining occupants’ thermal comfort and their dimension, orientation and location will greatly affect the ventilation conditions of the building. Specifically to this kind of climate, two strategies of ventilative cooling stand out as being the most effective. The first one is ventilating the building at night when the external temperatures are lower, in order to cool the interior mass of the building which will remain cooler the next day, reducing the rate of indoor temperature rise. The second strategy is called comfort ventilation, which can be summarized as the action of opening the windows with the purpose of increasing the indoor air speeds, creating this way a physiological cooling effect. In regards to the first ventilation strategy, office buildings have an advantage in comparison to other building types, since they have defined occupancy patterns, which are usually unoccupied during most of the night time. This way it is possible to avoid some level of discomfort to the occupants that could feel too chilly with the cool air of the night. One of the most important building features that guarantees the effectiveness of this strategy is the presence of thermal mass, which acts as a heat sink, absorbing the undesirable heat gains during daytime and dissipating them during night time by convection. Another fundamental aspect is that it should be coupled to the occupied space (Baker,N 2009). In addition, as shown in the Figure 2.8, the most effective type of thermal mass is the one which has maximum surface area possible, being well distributed around the working spaces. These criteria will also have an important impact on the definition of the window size, since having smaller apertures piercing the opaque envelope can offer higher possibilities of having more surface of thermal mass than a curtain wall, where the façade is completely glazed. When considering the window typology for night time ventilation, it is also important to consider the issues of security and rain. Concerning the latter, top hung windows with moderate size can be a good solution (Figure 2.9).
Fig 2.8 The most effective thermal mass application. Source: Baker, N (2009)
14
Fig 2.9 Examples of top hung windows Source: CIBSE (1999)
As a rule of thumb, it is known that an air speed of 2m/s can produce a cooling effect for occupants, equivalent to about 3°C reduction in the air temperature (CIBSE, 1999). This way, different researches were conducted in order to assess occupants’ preference of air movement in working environments. Zhang et al (2011) analyzed a database of indoor air quality performed in over 200 buildings in the United States and extracted office workers’ responses concerning air motion. His main finding was that dissatisfaction with the amount of air motion is very common, with little air movement cited far more commonly than too much air movement. Another interesting conclusion is that complaints regarding high air speeds or draughts decrease when the occupants are feeling “neutral” or “hot”, and when people feel cold they usually prefer less air movement (see fig.2.10). These findings provide an indication regarding the need of having two different ventilation strategies for summer and winter. Thus, in the same façade, at least two different window types should be incorporated allowing for enough air movement in order to dissipate the excess heat and afford physiological cooling during the warm days and low ventilation rates in the winter so as to guarantee air quality. Figure 2.11 exemplify window systems where both winter and summer ventilations can be achieved. The most important aspect is that the lower windows allow the opportunity of increasing or reducing the effective opening area, increasing the occupant control regarding the ventilation rates.
Fig 2.10 Air movement preferences for those who complained about temperature being often too hot, often too cold, and both. Source: Zhang et al (2011)
15
Fig 2.11 Window system which allows for control of different ventilation rates Source: Best practice programme (1997)
In addition, Brager and De Dear (2001), in a survey that was carried out with 230 office workers of Berkeley Civic Centre, found that a significant percentage of the building occupants wanted more air movement even in the winter. The percentage of people wanting more air movement in the summer was 45%, against 28% in the cool season. One of the most important points that drew the researchers’ attention was that almost no negative sensation in regards to draughts was found in this building (only 3 percent of the occupants wanted less air movement). As a conclusion Bragger stated that the effectiveness of the operable windows in terms of air movement was modest, due to the upward airflow pattern created by the hopper window typology. Another important finding was that more than half of the occupants perceived air movement as beneficial in terms of enhancing productivity, as shown in the Figure2.12. Aditionally, Lamberts et. Al (2011) after extensive research in natural ventilated buildings located in warm humid climates (Brazil) stated that most of the occupants’ complaints, stemmed from their preference for ‘more air movement’. These considerations highlight the important role of the windows in defining the ventilation strategy and the air flow patterns inside the building.
Fig 2.12 Air movement and its impact on productivity. Source: Zhang et al (2011)
16
The first step to increase average indoor air velocities is to orient the windows towards the prevailing winds. Givoni (1994) mentions that in hot humid climates, if there is a conflict between orienting the windows according to solar exposure or prevailing wind direction, ventilation should be considered as the primary factor, since shading devices can minimize the overheating caused by direct solar radiation. Ghiabaklou (2010) mentions that in general, for the majority of window configurations tested in different researches, orientation of inlets at 90° to the wind provided the highest average indoor speed ratios; however it was also found that some inlet configurations could provide the same results or even better ventilative cooling in oblique winds (up to 45°). Similar considerations were also made by Givoni. In order to achieve effective natural ventilation, the depth of the room should follow a specific criterion for each kind of strategy. Baker (2009) recommends a maximum plan depth of 3 times the dimension of the floor to ceiling height for single sided ventilation and 6 times the floor to ceiling height for cross ventilation. It is also important to highlight that even following these recommendations, cross ventilation is by far a more effective way of promoting comfort ventilation compared to single sided ventilation. The latter will be more common in cellular offices and meeting rooms and in these cases the efficiency of the strategy will be increased if multiple ventilation openings are arranged at different heights on the façade taking advantage of the stack effect (Figure 2.13).
Fig 2.13 Single sided ventilation double opening After CIBSE (1999)
In the context of working environments, cross ventilation is a strategy more commonly found in open plan offices. The driving force which causes air flow in this case, is the difference of pressure generated by the wind at the inlets and outlets of a volume. Concerning the windows’ location, Givoni (1994) found that rooms with windows on adjacent walls ventilated better than traditional cross-ventilated rooms with windows on opposite walls, when the incident wind angle was perpendicular to the inlet. However at oblique wind incidences, traditional cross-ventilated rooms performed better than rooms with adjacent windows. Additionally, he also mentions that having the same area for the inlet and outlet creates the highest indoor air speeds. Moreover Sobin (1981) during the 1960s conducted wind tunnel studies investigating many window types and their effects on the indoor airflow. One of his most interesting findings is that horizontal windows (windows that are wider than their height) created greater wind speeds than vertical ones (windows that are higher than their width).
17
In parallel, even in some office buildings located in warm climates, high indoor air speeds can be a reason for occupant discomfort. As a part of this dissertation, a survey was held with the occupants of a natural ventilated office building located in the city of Jundiai – Brazil. One of the most important findings was that draught was the major cause of discomfort for the occupants sitting close to the window. It is important to highlight that the building was located on the top of a hill outside the urban fabric, which considerably increases the wind speed in the site. Furthermore, the window design didn’t allow the occupants to have a proper control of the air movement. In any event, the typical tasks performed in office environments require the use of papers and according to Holger Koch-Nielsen ( 2002) air speeds higher than 1 m/s at the working table in offices can create inconvenience by blowing papers. Ultimately, one can conclude that occupant control concerning air speed is one of the most important factors that influence comfort in natural ventilated offices. Thus, different typologies of apertures in the envelope should be provided, in order to allow a wider range of possibilities to the occupants, where they can alter indoor air velocities to suit their cooling needs.
2.5 Conclusions The effectiveness of the natural ventilation strategy in this context of working environments located in warm humid climates depends not only on the application of the basic concepts of passive design (e.g minimizing cooling needs by solar control), but also on having a natural ventilation strategy under strict control, through the proper design of windows. In such climates, where solar control must be applied to all orientations, shading device typologies should be further investigated in order to find better solutions where obstruction of views can be avoided and daylight conditions can be improved. In addition, analytical work must be done, to test the impact of shading devices in the air flow and indoor air speeds of buildings.
18
19
20
3 precedents/field work 3.1 Introduction 3.2 Thermal comfort 3.3 Natural ventilation studies 3.4 Findings
21
22
3 precedents/field work 3.1 Introduction Rogers Stirk Harbour’s office was selected for field work since it is one of the few office buildings in London which makes a good use of daylight and doesn’t rely on air conditioning to provide comfort to its employees, while having an attractive proposal in terms of working layout and building typology. The building is located at Fulham neighbourhood in London, by the River Thames. The working spaces are distributed in a five-storey building which is partially a refurbished brick warehouse and partially a lightweight structure which adds another 2 volumes to the original building in the ground floor level and in the top floors (see Fig 3.1). In total 160 employees work in the building. The site’s location offers a great advantage for the applicability of natural ventilation strategies, since it’s a quiet area protected from the road, thus avoiding problems concerning noise and pollution, which is a major determinant of the air quality inside and around buildings. The visit was done on May 25th, between 14:00 and 17:30. This time was chosen because the building has its main façade facing southwest, this way in the afternoon it is possible to observe how the movable solar control devices are operated in order to avoid glare and direct solar radiation and how these devices impact the environmental performance of the working environments (see Fig 3.2). Several spot measurements were conducted across the building, measuring lux levels, air temperature, air velocities and surface temperature. In addition, dataloggers were placed in each floor, and were kept there during a one week period in order to measure temperature and humidity. During the visit, informal interviews with employees provided useful information about how they interact with the building in order to adapt to the conditions to achieve comfort along the year. Later, in the second week (June 1st), another visit was carried out with the focus on obtaining a more accurate data concerning occupants’ satisfaction and comfort. 60 questionnaires were filled during this day and the result analysis provided good indication about the points to be enhanced in the building performance.
Fig 3.2 External view showing the movable shading devices
Fig 3.1 Diagram section showing the existing structure and the extensions
23
3.2 Thermal comfort As mentioned previously, the case study is located in London which lies in the southwest part of UK at 51.4°N latitude and 0.1°W longitude. The climate is characterised as temperate marine, which suggests relatively mild changes between seasons and rare extreme weather phenomena. In general, the winters are chilly and summers moderately warm, which can be beneficial in terms of applying natural ventilation strategies. Looking at the climate data from Meteonorm v. 6.1, London’s average temperature during winter is 6.2°C, and 17.9°C during summer. The highest temperature measured during a 10 year period (1996-2005) was 29.7°C, in August. The building visit was carried out in late May, and during the visit the outdoor temperature was 23°C, which is well above 13.5°C, May’s daily average (Meteonorm 6.1). This provided an opportunity to observe how the building performs in warmer days. The indoor spot measurements in comparison to the outdoor temperature provided a good indication of the effect of high internal gains on the working spaces. As seen in the Figure 3.3, the highest temperatures were found in the ground floor, which in the day of the visit had the highest number of employees (computers and some artificial lights that were on during daytime also contributed to the higher temperature, see Fig. 3.6). The average indoor temperature in the ground floor was around 5°K above the outdoor temperature. In the next step of the field work, four dataloggers were placed in different floors of the building, and the measurements were plotted against the comfort band deducted for May based on EN 15251. The dataloggers were placed in the level of the workplane (900 mm above the ground). As seen in the Figure 3.4, the highest temperatures in the ground floor were found during the weekend, as a result of the direct solar radiation on the southwest fully glazed façade, in which during those days had no solar control applied. In addition, since the building was not occupied during the weekend, also the windows weren’t opened, increasing the indoor temperature even more. By contrast the first floor is performing better, since during the same days the peak temperatures were much lower, showing that the heavyweight structure of the refurbished building is helping to reduce the rate of indoor temperature rise.
Fig 3.4 Dataloggers’ results 24
Fig 3.3 Building section containing temperature spot measurements taken during the day of the visit
During the weekdays the temperatures in both floors presented different patterns, being in some days well above the outdoor peak temperature and in other days following the outdoor temperatures closely. These variations show how the building thermal performance is sensitive to the different occupancy patterns and operation of windows and shading devices. Ultimately, the indoor temperatures are most of the occupied hours within the limits defined by the comfort band, apart from one day when the outdoor temperatures were already exceeding the upper threshold. However, it is known that the provision of adaptive opportunities such as desk fans can help to minimize occupants’ discomfort in this kind of conditions. The analysis of the outdoor temperatures shows a good potential for applying night-time ventilation in this building since in some warm days the temperature swings between day and night can be higher than 10°K. A fundamental requirement for the effectiveness of this strategy is the presence of thermal mass coupled with the indoor environment, which will act as a heat sink, absorbing the undesirable heat gains during daytime and dissipating them during night time. In the first and second floor, the exposed concrete slab works as an effective thermal mass (high density material) and if applied night time ventilation it could lead to a reduction in the peak daytime temperatures by as much as 3 – 5°K (Baker, 2009), thus avoiding the indoor temperatures rising above the comfort zone. Nevertheless, the ground floor and the top floors are lightweight constructions, and in this case even if ventilated during night time, the building structure wouldn’t be able to absorb the daytime heat gains in an effective way, as shown in the Figure 3.5. It is worth mentioning that in terms of refurbishment, the application of Phase Change Materials (PCM) can be a good solution for increasing the thermal mass in the floors built with metallic structure. In parallel, improving the window design in order to embrace security issues might be an important measure to be adopted, since, as highlighted by the facility manager, this was the main reason for not having the windows opened after the working hours.
Fig 3.5 Effect of thermal mass and ventilation rate on peak indoor temperature Source: CIBSE, 2005
Fig 3.6 Internal view from the ground floor extension
25
3.3 Natural ventilation studies The analysis of the building typology shows that the open plan interior layout allows the application of cross ventilation in all floors and the plans’ depth respect the recommended maximum proportion of 5 times the floor to ceiling height (CIBSE AM10, 2005). The ground floor, which according to the spot measurements had the highest temperatures, has high top hung windows (1,80 m from the ground), and also some rooflights (see Fig. 3.7) which provide daylight for the working area and help to enhance the ventilation rate due to stack effect. In order to have a more accurate verification concerning the effectiveness of the ventilation strategy in this floor, CFD simulations were performed using the same conditions as the day of the visit. Figure 3.8 shows that the high top hung windows are too close to the roof, and in this case the air flow is being directed to the rooflights. Consequently, in the level of the work stations low air movement is found, explaining some of the occupants’ complaints regarding feeling stuffy during the warm days. In addition, higher temperatures are also found in that area since the air movement is not enough to dissipate the excess heat generated by people and appliances.
Fig 3.8 CFD simulations showing air speeds and temperature in the ground floor level, for the current scenario. Source: EDSL Ambiens
26
Fig 3.7 The rooflights
According to CIBSE (2005), if natural ventilation is to be adopted the system should be able to provide a wide range of controllable ventilation rate, such as 0.5 to 5 Air Changes/Hour or even more. In this case another CFD simulation was conducted, in order to assess the effect of adding lower windows in the level of the workplane and its effectiveness in terms of providing a more efficient air flow distribution within the environment and providing physiological cooling effect to the occupants. As seen in the Figure 3.9 the air flow in this case is removing the excess heat from the lower levels and consequently the overall temperature of the environment is lower and more uniform. Ventilation rates up to 1 m/s can be found in the level of the workplane, which according to the Figure 3.10, could provide a cooling reduction equivalent to 2.7째K in dry resultant temperature. These lower windows are meant to be part of the main ventilation system, providing the occupants a wider range of possibility to adjust the ventilation rate to their cooling needs. Additionally, this strategy should be used mainly in the summer and warm days, and it can also be useful during the afternoons when the external awnings are down, considerably blocking the air flow in the upper inlets. Furthermore, the detail of the window could help to direct the air flow upwards, avoiding any discomfort to the occupants during the windy days.
Fig 3.10 Effect of air speed on dry resultant temperature Source: CIBSE, 2005
Fig 3.9 CFD simulations showing the effect on air speeds and temperature in the ground floor level, when adding lower inlets and outlets. Source: EDSL Ambiens
27
The ventilation strategy in the top floors follows a different pattern in comparison to the other parts of the building since in the mezzanine level the southeast façade completely opaque. This way, the fresh air is delivered by the windows located on the façade of the floor under the mezzanine (3rd floor) and an exhaust duct placed on the roof functions as outlet. From the occupants’ point of view, this ventilation strategy works well in general, however, during windy days; the bottom hung windows (see Fig. 3.11) tend to direct the strong airflow to the mezzanine, creating turbulence and draught in the adjacent workstations. Figure 3.13 shows a CFD simulation which illustrates and confirms the occupants’ statement. Consequently, the solution usually adopted by the employees is closing the windows, this way reducing the overall fresh air provision of the space, causing feeling of stuffiness, especially during the warm days.
Fig 3.11 Bottom hung window typology in the floor under the mezzanine
3.4 Findings The questionnaires revealed, that the overall occupants’ perception regarding comfort in this office building is quite high (85%), staying above 80%, which is the minimum recommended by ASHRAE(2004). This information shows that even though there was a relatively high rate of votes stating that the indoor temperatures were cold in the winter and sometimes too hot in the summer, the provision of daylighting, views and adaptive opportunities (e.g. operable windows, desk fans, task lighting, movable blinds and awnings) were fundamental to increase occupants’ satisfaction. In addition, the building has a robust design, thus a small improvements in the design of windows could easily enhance the ventilation conditions especially in the ground floor extension and in the top floors.
Fig 3.12 Key plan showing the cross section line in the mezzanine level.
Fig 3.13 CFD simulation showing strong air flow and potential draught in some of the mezzanine’s workstations Source: Autodesk CFD 28
29
30
4 s達o paulo context and climate 4.1 S達o Paulo context 4.2 S達o Paulo climate 4.3 Jundiai City Hall field work
31
32
4 são paulo context and climate 4.1 São Paulo context It is known that energy consumption alone does not determine the environmental performance of buildings. It is also necessary to assess the origin of the primary energy source in order to know if it is linked to carbon emission or other environmental issues. In Brazil, power generation is largely from hydroelectricity, which accounts for approximately 91 percent of the total (Lamberts et al. 2011). This way, electricity consumption in most of the cases cannot be directly related to carbon emissions; however energy efficiency of buildings is necessary and can be an important step to more sustainable development, taking the existing scenario into account where growth pressures conflict with the limited capacity of energy production. In 2001 Brazil faced a major energy crisis, due to inadequate infrastructure investments and the extreme climatic events, such as lack of rain to drive hydroelectricity generation. Consequently, the entire country was compelled to achieve a consumption reduction of 20 percent. Since then the Brazilian government has been promoting energy-conservation programmes and building guidelines focusing on a more sustainable approach. The government initiatives however, were not enough to create awareness amongst engineers, architects and developers concerning the principles of environmentallyresponsive architecture. This way, fully glazed office buildings, completely dependent on air conditioning and artificial lighting have marked the most common typologies of commercial buildings that have been built in growing cities like São Paulo. Buildings are responsible for 48,3% of total electrical energy consumption in Brazil; 23% of which commercial and public buildings are accounted for, while 22% is dedicated to residential sector (Ministerio das Minas e Energia,2007). The importance of this proportion becomes even more relevant in São Paulo which hosts the highest amount of commercial building in the country. In addition to this, considering that the climate is mild in comparison to northern regions of Brazil (closer to Ecuador line), where more extreme conditions can be found; São Paulo presents a good opportunity to create more impact in energy savings with relatively easier interventions.
4.2 São Paulo climate São Paulo is located in the southeast region of Brazil, at the latitude 23°3’ south and longitude 46°4’ west of Greenwich. It is positioned approximately 85 km from the coast and 800 m above the sea level. São Paulo’s population is over 11 million, distributed in area of 1,509 km² (Prefeitura de São Paulo, 2009). According to the Brazilian Institute of Geography and Statistics (IBGE), São Paulo’s climate is classified as mild moderate. As seen in the Figure 4.1 the temperatures can go up to 33°C in the summer and as low as 8°C in the winter, however most of the year they stay between 15°C and 25°C. The high range of diurnal fluctuation, which is around 10°K, indicates a potential for applying passive cooling solutions through thermal inertia.
Hourly maximum temp. Mean outdoor temp. Hourly minimum temp.
°C 40 30 20 10 0 -10
Jan
Feb
Mar
Apr
May
Jun
Fig 4. 1 Monthly diurnal avarege temperatures for São paulo Source: Ecotect Weather Tool (data: Meteonorm 6.1)
Jul
Aug
Sep
Oct
Nov
Dec
33
Lamberts (2007) analyzed São Paulo’s psychrometric chart as being 27,1% of the hours within the comfort zone however for the remaining 72,9% discomfort conditions are found, of which 59,4% are a due to low temperatures and only 13,5% are due to high temperatures. This statement confirms that in general, for most of the cases, the summer temperatures are not the biggest challenge to be overcome in the design process if the undesirable solar gains are blocked through the effective use of solar control. The most commonly found sky condition in São Paulo is overcast (7-8 octas), accounting for 60% of the year; while clear sky (0-3 octas) is only 21% and intermediate (4-6 octas) is 19%. The high frequency of overcast sky conditions increases the amount of diffuse radiation in comparison to the direct radiation. As seen in Figure 4.2, the amount of diffuse and direct radiation is almost the same throughout the year.
Wh/m ² Fig 4. 2 Mean irradiance of direct and diffuse horizontal radiation. After Meteonorm 6.1
The Global Horizontal Illuminance levels for São paulo has an yearly avarage value of 30 Klux at 9:00, which for working environments is relevant since usually this time marks the beginning of the working hours. Moreover, these values can be as high as 60 Klux in some months of the year, between 11:00 and 13:00 ( see Fig. 4.5). Another important aspect that has great impact on the overall comfort conditions is the high humidity rates. As seen in Figure 4.3, an average relative humidity of 80% is found throughout the year. It is known that high humidity restricts evaporation, increasing skin wetness, which might lead to discomfort. In this case ventilation and air motion play an import role in enhancing comfort conditions, since they enable convective cooling of the skin. 100% 80% 60% 40% 20% 0%
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Fig 4. 3 Relative humidity throughout the year. Source: Ecotect Weather Tool (data: Meteonorm 6.1) 34
Oct
Nov
Dec
According to Lamberts et al (2011) air movement is the great determinant whether higher summer operative temperatures will be acceptable or not. In addition they also mentioned that Brazilian occupants accept temperatures swings during the day and year in naturally ventilated buildings and usually prefer higher air speeds if control is provided. Figure 4.4 specifies the limits of air movement acceptability in a range of operative temperatures, as a result of field experiments carried out in non-residential buildings in different climate zones in Brazil, where over 5000 questionnaires were applied in both summer and winter. It is possible to realize that more than 80% of acceptability can be found within a range of air velocities between 0.2 and 0.9 m/s for temperatures varying from 18°C to 32°C. Also, It is worth mentioning that even though 0.8 and 0.9 m/s was found to be widely acceptable at high operative temperatures, it does not mean that higher air velocities would result in draught perception. Arens et al (2006) stated that while draught might be a relevant concern in cold climates, it is less important in warm environments. Additionally, the new version of ASHRAE standard 55: 2010 defines the limit of indoor air movement acceptability in 1,2 m/s when control is provided. São Paulo’s prevailing wind direction is from south-east and east as seen in the Figure 4.6. The average wind speeds vary from 2 m/s to 4 m/s for most of the year. Higher wind speeds between 7 m/s and 10 m/s can be found, however only for a few hours in the year.
Fig 4. 4 Air movement acceptability in response to different operative temperatures. Source: Lamberts et al (2011) Hours
Fig 4. 5 Global Horizontal Illuminance for São Paulo, monthly mean of hourly values (klx). After Meteonorm 6.1
Fig 4. 6 São Paulo’s yearly wind rose Source: Ecotect Weather Tool (data: Meteonorm 6.1) 35
4.3 Jundiai City Hall field work The fieldwork was carried out in a building located in Jundiai, which is a city of 370,000 habitants, situated 50 km away from São Paulo. The building chosen was the city hall, designed by the architect Araken Martinho, who won the first prize in a national competition held in 1987. The building design addresses issues related to natural ventilation and daylighting in the work environments and transitional spaces. The concern regarding achieving occupant comfort by passive means, led the architect to consult a specialist in environmental comfort, which wasn’t a very common measure by that time. The outcome of this thorough design process is a building which has been working as free running for the past 30 years in a climate very similar to São Paulo, where almost 100% of the new office buildings are designed to be air conditioned. This investigation provided a good opportunity to understand the potential issues related to window design and operation in the context of naturally ventilated office buildings in terms of building performance and occupant comfort.
Fig 4. 7 External View of the North Facade
4.3.1 The building The city hall is a nine stories building located outside the city centre lying on the top of a hill ( see Fig. 4.7). The site is surrounded by a park, and the adjacent streets are exclusive for the building access, resulting to a very quiet urban environment. Its main façades are oriented towards north and south, and the building mass is divided into two blocks, which are connected by an atrium. The atrium provides good internal views between different building sectors, and also helps to enhance the ventilation strategy, while allowing daylight penetration (see Fig.4.8). The ground floor where the main access of the building is located, also hosts a covered square which works as a transitional space. The building was entirely built with reinforced concrete and tinted glass was applied in all of the windows apart from the atrium skylight, where clear laminated glass was used. The internal layout in the working environments is predominantly open plan, with few meeting rooms and cellular offices distributed along the floors.
36
12m
Fig 4. 8 Internal view of the atrium
12m
12m
Fig 4. 9 Typical floor plan. Source: Araken martinho
4.3.2 Building visit The building visit was done on April 11th between 12:00 and 17:00. During this period the external temperatures varied between 27째C and 30째C which can be considered as quite high values for this season (spring), providing a good opportunity to learn how the building would perform in summer days and how the occupants respond to these conditions. The visit focused on two floors, one facing north and the other one facing south. As a part of the investigation, spot measurements were taken and informal interviews and questionnaires were conducted. The questionnaires were applied to 30 employees, which represents more than 50% of the staff of each investigated floor. Additionally, in order to have a more accurate data regarding the building performance during a one week period, 4 dataloggers were placed in different locations (an external balcony, ground floor transitional space, north and south facing offices, see Figure 4.10). This way, it was also possible to understand how the building performs without the effect of internal heat gains, when it is unoccupied during the weekends.
Figure 4.10. Cross section showing the data loggers position and the analysed floors. After Araken martinho 37
4.3.3 Daylight Spot measurements were carried out in a typical open plan office area facing south in order to assess daylight levels in the work environment. The measurements started at 14:00, and during that time the cloud cover prevailing in the sky was 1 Octa. It is important to highlight that the artificial lights were turned off just before starting the spot measurements. As seen in the Figure 4.11, this work environment presents poor daylight conditions, with acceptable levels in the workstations close to the window; however an abrupt drop was identified in the next few meters from the perimeter, reaching down to 0 Lux in the middle of the room. The corridor facing the atrium had the highest values of daylight, however the adjacent workstations don’t benefit from that, since there is a high partition separating the working area from public circulation.
Fig 4. 11 Cross section of the 4TH floor facing north, showing the spot measurements values.
It is known that the tinted glass has low light transmittance. This can be a possible explanation for such poor daylight conditions even with such shallow plans (12 m) and no external obstructions. With the intention of verifying this assumption, daylight simulations using Radiance were done for the same floor where the spot measurements were carried out, using the same sky conditions identified during the field work. The Figure 4.12 shows the base case simulation where the same glass type as the existing building was applied. It is possible to observe that the tinted glass is indeed reducing the overall daylight levels in the working area, even close to the windows. The illuminance levels achieved in this case are similar to the ones measured in the visiting, which could be considered as a calibration of the model. Afterwards, another simulation was done changing the glazing type of the windows to clear glass, which offers the highest light transmittance. As seen in the Figures 4.13 and 4.14, in this case the daylight conditions are much more adequate for a work environment, given the fact that the lowest daylight levels found in the room was 400 lux which is high above the minimum level (200 lux) recommended by CIBSE (Guide A). In this case it can be concluded that, changing the glazing type to clear glazing would have a positive impact in the daylight conditions of the room, thus the amount of hours using artificial lighting could be reduced considerably, which would consequently reduce the internal heat gains.
38
In regards to solar control strategy, for most of the year, the building offers a good protection from direct sunlight. The concrete overhangs in the façade block the sun in the north orientation during the cooling season; however in the spring and winter when the sun is lower in the sky, sunlight enters into the room and creates visual discomfort to the occupants sitting by the windows and also physiological discomfort when the direct sun light falls onto them. The building has internal curtains which are currently not working as result of the lack of maintenance, so in this case the occupants don’t have the option of having manual control. In addition, similar to the north façade, the south orientation is also mostly protected however during the summer; the façade is exposed to the sun during the first hours of the morning. Vertical shading devices would be necessary in that orientation in order to have a more efficient solar control.
Lux
Fig 4. 12 Base case (tinted glass). Illuminance levels for the 4th floor at 15:00 on April 11th. Sky condions: Overcast Source: Radiance
Fig 4. 13 Clear glass scenario. Illuminance levels for the 4th floor at 15:00 on April 11th. Sky condions: Overcast Source: Radiance 39
Lux
Fig 4. 14 Cross section. Clear glass scenario. Illuminance levels at 15:00 on April 11th. Sky conditions: Overcast Source: Radiance
4.3.4 Thermal Comfort The extensive use of concrete in this building, combined with the perforated ceiling type which allows the thermal mass to be coupled with the working environments, has a positive effect in reducing the diurnal temperature fluctuations by absorbing the heat gains during day time and releasing them during night time. Figure 4.15 shows the temperatures recorded by the dataloggers between 11th and 19th of April. It is possible to see that, as a result of the thermal inertia, the difference between average indoor temperatures for day and night is usually less than 4째K, while the outdoor is around 8째K. It is also worthy of mentioning that this strategy could be enhanced if night time ventilation was applied, however nowadays due to rain and security issues, the windows are closed at the end of the working hours. Figure 4.15 also shows that due to the internal gains the indoor temperatures are usually equal or slightly higher than the outdoor peak temperatures, even though the indoor environment is completely coupled with outdoors. The interviews and questionnaires provided good indication regarding occupant satisfaction throughout the year. During warm summer days, they claimed to be very dissatisfied with the thermal conditions in the building. However during the time when the questionnaires were conducted, most of the occupants were feeling slightly warm but still comfortable, despite the fact that the indoor temperatures were varying between 29째C and 30째C. Hence, it can be assumed that during warm summer days, the indoor temperatures could easily exceed these values. On the other hand, during the winter, most of the occupants interviewed stated that they felt comfortable, apart from the coldest days, when high infiltration through the old window frames allows undesirable air movement.
40
Fig 4. 15 Data logger measurements showing internal and external temperatures
4.3.5 Natural ventilation As explained previously in this chapter, achieving thermal comfort in warm humid climates depends highly on air movement in order to offset the skin evaporation effect caused by the high humidity rates. However, the data logger’s measurements, along the week, showed that the indoor humidity levels were below the limit set by CIBSE for working environments (70%) during all the working hours, having an average value of 65%. During the visit, indoor air velocities were measured over a one hour period and they varied in a range between 0.5 to 1.0 m/s with few peaks reaching up to 2.8 m/s, and even in that conditions no draught complaint was found. In addition 80% of the occupants who answered the questionnaires stated that during that day, they would even prefer to have more air movement. On the other hand, in the informal interviews it became clear that draught in the workstations close to the windows is usually one of the main factors that increase dissatisfaction. The top hung window typology (Figure 4.16), does not allow the occupants to have a proper control of the air movement. The city hall is located on the top of a hill, having no obstructions for the wind, which considerably increases the air velocities in and around the building. The employees sitting by the windows complain about excessive air velocities while the ones occupying the workstations in the middle of the space, usually feel pleasant with the amount of air movement. The employees sitting by the corridor complain about having almost no air movement. Figure 4.17 shows a CFD simulation that illustrates the occupants’ statement, proving that air velocities are actually higher closer to the windows and very low close to the atrium.
Fig 4. 16 Internal view showing the open plan layout in the 4th floor. 41
Fig 4. 17 CFD simulation showing draught in the work stations adjacent to the windows Source: Autodesk CFD
Control over the windows and blinds in open plan offices is also a major source of occupant dissatisfaction. Usually the occupants try to find an agreement amongst themselves on how to operate the adaptive elements. The main issue is that, even during some winter days and autumn days, the occupants feel the necessity of opening the windows to have a feeling of fresh air, so the employees who work by the windows fell very uncomfortable during most of the time. One employee stated that she had sinusitis 3 times in a six months period, due to the cold draughts. Other major issue, is that some occupants (especially women), have a higher degree of sensitiveness regarding air flow. So two people among 30, always complain about opening the windows even during the summer, because they feel too cold regarding the air velocities. For this problem a good solution is, having a more flexible layout, where each one of the employees can choose the location which better suits their preferences.
42
43
44
5 analytic work 5.1 Preparing the shoebox 5.2 Design Proposal: Shading devices 5.3 Design Verification: Daylight simulations 5.4 Design Verification: Thermal simulations 5.5 CFD analysis and comfort assessment
45
46
5 analytic work The first step to study the window design topic analytically was to define a shoebox with dimensions that could represent a wide range of layout configurations in work environments. Secondly, the most effective shading device typology for each orientation was assessed, in terms of blocking direct solar radiation and providing adequacy of the remaining (diffuse) radiation for daylighting purposes. In parallel, a reduced window to wall ratio was also tested in each orientation, and its impact in the overall daylight conditions was verified using the software Radiance. The purpose of these simulations was to ensure that the offices would be able to achieve a maximum independence from artificial lighting during the working hours. The following simulations were carried out in EDSL TAS (Thermal Analysis Simulation Software), aiming to understand how effective the application of solar control could be in terms of reducing the absorbed and transmitted solar gains, thus reducing the temperature fluctuations during warm days. Finally, a window design proposal was suggested for each orientation, and its efficacy in terms of providing controllability for a wide range of air speeds and avoiding draught was tested using CFD simulations (Autodesk Simulation CFD 2013). Ultimately, comfort conditions in the work environments were assessed using PMV calculations, which considered the effect of different factors that can affect comfort (air speed, air temperature, mean radiant temperature, clo value, and metabolic rate).
5.1 Preparing the shoebox 5.1.1 The plan Changing the internal layout in office buildings is a quite common practice as their needs and working methods are constantly adapting to the new trends. This necessity for flexibility requires a façade design that can respond to a wide range of layout configurations. Therefore for the analytical studies, a shoebox that can represent the typical office plan typologies is suggested. The dimensions of the plan also came from the idea having maximum flexibility; where an open plan office can be easily divided into cellular rooms/meeting rooms. As previously mentioned in Chapter 2, the recommended plan depth to benefit from daylight is twice the height from the floor to the top of the windows, which in this case is 3 m, giving a plan depth of 6 m. While the plan depth was calculated according to passive zone principle, the width of the plan was an outcome of architectural concerns. To see the differences between different groups of occupants, at least two sets of desks were placed in the plan, in order to achieve a required density of 6 m2 per occupant*. Also, for internal layout flexibility a virtual modular system was suggested, consisting of modules of 4 m, which would, again, satisfy the density requirements if a room of 6 x 4 m was to be created. The shoebox, in the end, had a plan of 6 x 8 m (two modules). Figure 5.1 shows the possible layout options.
* The occupancy density was also a primary concern. After analysing three case studies (see Appendix I) in São Paulo, it was concluded that a typical density for open plan office buildings is between 5.5 and 6.5 m2 per occupant. This finding was also confirmed by ABNT (Associação Brasileira de Normas Técnicas, Brazilian Association for Technical Standards) as the typical density being 6 m2 per occupant. 47
6.00 m
Corridor
8.00 m
4.00 m
16.00 m Fig 5.1 The shoebox layouts for open plan and cellular office scenarios
The layouts presented here also suggest some strategies: The tables are placed perpendicular to the façade as CIBSE recommends (CIBSE Lighting Guide 7), to reduce glare problems on computer screens. The workstations are detached from the façade, creating a corridor between the tables and the windows. This strategy, as seen in the field work conducted in Rogers Stirk Harbour + Partners office (chapter 3), enables a better and easier operation of windows. Also, getting further from the façade decreases the chance of facing extreme conditions in terms of ventilation and illumination.
5.1.2 The façades As mentioned earlier in theoretical background and elaborated in the field work that was conducted in Brazil, the reflective glass that is used in fully glazed office buildings reduces daylight levels dramatically, resulting in higher needs for artificial light. Therefore in the simulations clear glazing will be used in all cases due to its high light transmittance properties. A fully glazed façade was selected as the basecase since it is a common façade typology for commercial buildings. As explained earlier in the climate analysis, São Paulo has almost equal amount of direct and diffuse radiation. This means in addition to providing solar control devices to avoid direct radiation, reducing the amount of diffuse radiation is also important to avoid glare problems and overheating. This strategy was used in the other façade typologies that will be tested; the window to wall ratio (WWR) was reduced to 50% in both cases, one with vertical and the other with a horizontal composition. In the vertical composition, the opaque parts between windows are left intentionally for a potential investigation of having opaque vents on those parts. In the horizontal composition, the transparent element of the façade is divided into two parts regarding the conclusions driven from the theoretical background (see Fig. 5.2). Besides the sustainable concern, diminishing environmental problems by passive means could be also cost effective. These strategies will not only decrease the operational costs but the initial costs of suggested façade typologies can by default be lower since in general most of the opaque construction materials are cheaper than transparent elements in Brazil, and this difference will compensate for shading device costs. 48
4.00 m
Fully Glazed WWR: 100%
Vertical WWR: 50%
Horizontal WWR: 50% Fig 5.2 Different WWR configurations
5.2 Solar Control Proposal/Daylight Analysis After defining the faรงade compositions, shading devices were designed for each orientation. The design process included testing of different shading device typologies, assessing their efficacy on blocking direct radiation through stereographic diagrams where the sun path is projected and their effect on daylight levels using Radiance. The daylight levels were checked in overcast sky at 9.00 am in the morning (start of the working hours). This criterion was based on the assumption that if the occupants have sufficient daylight when they reach their offices, they would not turn the artificial lights on, which, in case of the office buildings with only manual controls, would prevent the lights staying on throughout the day. Another criterion that was adopted was to avoid direct radiation inside the space after 9.00 am throughout the year, since as mentioned in the previous chapters, this can lead the occupants to pull the blinds down and turn on the artificial lights. Additionally, the sunpatches falling onto the occupants can create discomfort. The comfort conditions concerning daylight levels were accepted as 200 lx for minimum and 2000 lx for maximum (CIBSE Guide A). The effect of reducing the glazing ratio on the amount of absorbed and transmitted diffuse radiation in the building will be examined in further steps. 49
5.2.1 North For the north orientation horizontal shading devices are tested, with the addition of a lightshelf for the horizontal window typology. Since the same type of shading device is tested for all 3 cases, only one sunpath diagram is presented (see Fig. 5.3). In all cases, horizontal louvers seem to be effective to prevent direct radiation after 9.00 am throughout the year. As for daylight levels, although fully glazed faรงade seems to have a good amount of daylight penetration, glare might be a possible problem especially in the parts of the place closer to the faรงade since the amount starts getting excessive. On the other hand, the shoebox with horizontal windows seems to provide adequate daylight with a good uniformity throughout the space, especially with the addition of the lightshelf. Although the light shelves are known to be working more effectively in clear sky conditions, it is apparent in the simulations that it does help daylight reaching further in the room, even with overcast sky. This deep penetration is not to be found in the faรงade with vertical windows, in addition to having a poor evenness in front of the faรงade (See Fig. 5.4)
100% WWR
The horizontal composition has also another benefit in means of views since it offers the most uninterrupted visual connection to the outside. This feature is highly appreciated by the occupants as previously mentioned in Chapter 2.
50% WWR
50% WWR Fig 5.3 The sunpath diagram 50
Fig 5.4 Illuminance levels at 9:00 am for an overcast sky (30 Klux). Source: Radiance
5.2.2 South For the south orientation the vertical shading devices were examined with same dimensions, apart from the faรงade with vertical windows due to the faรงade composition. In that case the depth of the shading devices was increased while the number of vertical fins was decreased. The daylight levels for the fully glazed faรงade and vertical windows seem to be almost the same, and the horizontal windows seem to have slightly lower daylight levels in comparison to these two. But overall, there seems to be high risk of over illumination for the first two cases, therefore the faรงade typology with the horizontal windows was assumed to be more efficient.
100% WWR
50% WWR
50% WWR Fig 5.5 The sunpath diagram
Fig 5.6 Illuminance levels at 9:00 am for an overcast sky (30 Klux). Source: Radiance 51
5.2.3 East For this orientation, three different types of shading devices were tested; vertical, triangular and horizontal. Three sunpath diagrams are shown respectively. While the horizontal and vertical ones work effectively, triangular shading device shows inconsistencies throughout the year. All the options presented were tested with the maximum glazing ratio (100%) in order to maximize daylight penetration since it is known that for east and west orientations the shading device can have a much bigger impact on daylight levels. Comparing the vertical and horizontal shading devices, it is evident that horizontal ones provide a more even daylight distribution, while the vertical ones have poor daylight levels especially at the back of the room (160 lx), in addition to blocking the views.
100% WWR
100% WWR
100% WWR Fig 5.7 The sunpath diagrams 52
Fig 5.8 Illuminance levels at 9:00 am for an overcast sky (30 Klux). Source: Radiance
5.2.4 West The west orientation presents similar problems as the east: the low solar angles require vertical shading devices in order to block direct sun as much a possible. In the east orientation the occupancy hours helped with this problem since when the sun angles were very low in the early morning the office was not occupied either way and the solar radiation levels were not as high. However, in the west orientation occupancy schedule does not compensate for the problem, therefore horizontal shading devices were excluded for this case. The remaining typologies were vertical and mesh. The sunpath diagram and Figure 5.11 illustrate that, in the case of mesh, solar radiation is not being blocked effectively in comparison to the vertical fins. The vertical shading devices, as seen in the sunpath diagrams, can block the direct sun in the summer; however for a fully effective solar control throughout the year it is imperative to have movable vertical fins that would change seasonally. Therefore the vertical shading devices were chosen over the mesh, despite the fact that mesh provided better daylight conditions. On the other hand, even though the lowest daylight level with the vertical fins (160 lx), was below the adopted comfort band, it can still be acceptable according to Baker (2002), when he remarks that a minimum level of illumination required for tasks carried out on computer screens is 150 lx.
100% WWR
75% WWR
Fig 5.9 The sunpath diagrams
Fig 5.11 Avarge daily solar radiation on the facade Source:Ecotect
Fig 5.10 Illuminance levels at 9:00 am for an overcast sky (30 Klux). Source: Radiance
Wh 53
5.3 Design Verification: Daylight simulations After choosing the most effective solar control strategy for each orientation, a more accurate verification regarding daylight availability in different sky conditions was performed using false-colour pictures (generated in Radiance), which illustrate and quantify the illuminance levels in various zones of the office space. In this case, in order to conduct the simulations in a more realistic open plan layout scenario, the shoebox analyzed had its plan depth increased from 6 m to 12 m, which resulted in a work environment with two exposed faรงades. Two hours of the day were chosen to represent the extreme daylight conditions that can be found in the work environments. The first hour is 12:00, when during some months of the year the global horizontal illuminance has an average value of 60 Klux (see chapter 4) and this hour was simulated under clear sky conditions. The second hour is 9:00, which will represent the conditions found early in the morning and late in the afternoon and in this case the simulations were carried out using overcast sky with a global horizontal illuminance of 30 Klux.
5.3.1 North-south orientation Figure 5.12 shows a daylight simulation conducted at 12:00, in a scenario where the exposed faรงades (north and south) were shaded; therefore the WWR was kept as the basecase (100%). The purpose of this simulation was to create a comparative measure to quantify how effective the reduction of the glazing ratio was in terms of minimizing over-illumination issues. It is evident that in this case the high glazing ratio is increasing the daylight levels excessively, especially in the work stations facing south faรงade where the range of sky component is higher due to the shading configuration. In this case, illumination levels up to 3600 lx on the work plane were found which could create discomfort especially if the occupants are performing tasks using computer screens. Consequently, the blinds are likely to be pulled down. On the other hand, when the glazing ratio is reduced by 50% (Figure 5.13), the daylight levels are within the limits of the defined comfort band in most of the office space, apart from few areas on the workstations closer to the windows, where the illuminance levels reach 2050 lux. However, these levels are still considered acceptable since they are exceeding the threshold by less than 5% and additionally the clear sky conditions only happen during a small percentage of the year (21%).
North faรงade
South faรงade
100% WWR
Fig 5.12 False colour image showing daylight conditions in January at 12:00 - Clear Sky ( 60 Klux). Source: Radiance 54
100% WWR
As seen the Figure 5.14, the false colour picture is confirming the result of the previous simulations, showing that minimum levels of 300 lux were found at 9:00, which demonstrates that the strategy of reducing the WWR was successful not only in terms of reducing over illumination issues, but also in terms of providing acceptable conditions of daylight early in the morning and late in the afternoon. However, it is worth mentioning that, after 17:00 the availability of global horizontal illuminance in the sky reduces considerably, which is likely to result in the use of artificial lights to compensate for low internal daylight levels. Nevertheless, this is still acceptable since at this time, even during the warm summer days, the external temperatures are already dropping after reaching its peak at around 13:00, so the internal gains generated by the artificial lighting will not have a negative impact in the amount of overheating hours. In terms of energy saving, this strategy is already reducing the amount of hours using the artificial light by 80% in comparison to a building with tinted fully glazed façade: the artificial lighting hours drop from 10 hours to 2 hours: 09:00 – 19:00 to 17:00 – 19:00. North façade
South façade
50% WWR
50% WWR
Fig 5.13 False colour image showing daylight conditions in January at 12:00 - Clear Sky ( 60000 Lux). Source: Radiance
Fig 5.14 False colour image showing daylight conditions in July at 09:00 - Overcast Sky ( 30 Klux). Source: Radiance 55
5.3.2 East-west orientation For the east – west orientation the WWR considered was 75%, where the only difference from the simulations conducted in the previous subchapter (WWR 100%) was the replacement of the transparent elements in the sill for opaque ones. As it is known, this modification will have no negative impact in the daylight performance since for this particular case, relevant illuminance levels are assessed above the sill level (on the work plane level). Figure 5.15 shows that at 12:00 the overall daylight distribution in the office space is more even in comparison to the north – south orientation. In addition, the maximum daylight levels that were found are lower than those of the previous case, and all within the defined comfort band. On the other hand, in the beginning of the occupation hours (09:00), lower daylight levels are found in the workstations close to the west façade as expected (see Fig. 5.16). However, they are still higher than the simulations conducted in the single sided shoe box. A possible explanation for this fact, is that the better daylight availability that the east façade provides is helping to enhance the overall illuminance levels. West façade
East façade
75% WWR
75% WWR
Fig 5.15 False colour image showing daylight conditions in January at 12:00 - Clear Sky ( 60000 Lux). Source: Radiance
Fig 5.16 False colour image showing daylight conditions in July at 09:00 - Overcast Sky ( 30000 Lux). Source: Radiance 56
5.4 Design Verification: Thermal simulations While the sun path diagrams were previously used to assess the effectiveness of the shading devices in terms of blocking direct radiation, in this step of the analytic work, the same shoebox will be analyzed using EDSL Tas. The purpose of these simulations is to quantify the amount of diffuse solar radiation that is being admitted into the room in aforementioned WWR configurations. For this set of simulations, the office space was considered as unoccupied, in conditions similar to what would be found during the weekends in a real office building. The windows are assumed to be closed; therefore there is no ventilation for space cooling, but infiltration. Excluding the effect of ventilation and occupant related heat gains provides a more clear understanding regarding how the solar gains are affecting the indoor temperatures. In addition, it is possible to quantify how the solar control strategy is contributing in the reduction of indoor temperature fluctuations. The simulations were conducted in a typical week of January which is the hottest month (chapter 4). The solar gain rates were also an outcome of the thermal simulations, and the numbers present the daily average values for January. As seen in the Tas inputs, no insulation layers are applied to the external walls and the glass type is single glazing, both of which correspond to the typical constructions used in most of the building typologies that have been built in São Paulo.
TAS INPUTS Wall Concrete block U value: 2.69 W/m²K Internal Floor/ Ceiling Concrete Slab + Raised Floor U value: 1.45 W/m²K Glass Single Glazing - Clear U value: 5.55 W/m²K Frame U value: 5.35 W/m²K Infiltration Rate 0.35 ACH – According to ABNT (NBR 6401)
57
5.4.1 Without occupancy 5.4.1.1 North
kWh/Day
As seen in the Figure 5.17, while applying shading devices to a fully glazed façade drops the average daily solar gains by 40%, decreasing the WWR to 50% results in a reduction of 30%. The combination of these two strategies leads to a reduction of 70% in total. The fact that reducing the WWR creates almost the same impact as shading devices proves the initial hypothesis that this strategy would have a great effect in the solar control strategy, blocking the undesirable diffuse radiation. Figure 5.18 shows that with this solar control strategy it is possible to reduce the resultant temperature by around 7°C. Additionally, it is possible to observe that through blocking most of the solar radiation, not having internal heat gains, together with the exposed thermal mass; it is possible to have resultant temperatures up to 4.5°K lower than the external dry bulb temperature. This finding confirms the statement that was done in the climate analysis, where it was concluded that the summer temperatures in most of the cases are not too big of a challenge to overcome in this climate if a proper solar control strategy is applied.
WWR: 100%
Fig 5.18 Typical summer week thermal simulation (Source: EDSL Tas) 58
100% 60% 30%
Fig 5.17 Solar gains (absorbed + transmitted) reduction (Source: EDSL Tas)
WWR: 100% + shading
WWR: 50% + shading
5.4.1.2 South
kWh/Day
In the south orientation, as seen in the Figure 5.19, the shading devices are reducing the solar gains only by 25%, which is 15% less than in the north orientation. The reason behind this is that the vertical shading devices are exposing bigger portions of the room to diffuse radiation since the sky component in this case is higher when compared to the façade with horizontal shading devices. It is worth reminding that the direct radiation is blocked in both cases. The same conclusion is also noticeable in Figure 5.12 in the daylight analyses, in which the higher illuminance levels are evident closer to the south façade. On the other hand, when the WWR is decreased to 50%, the solar gains drop by 65% in comparison to the basecase, showing that the impact of diminishing the WWR as a solar control strategy is even more effective and necessary in this orientation. As seen in the Figure 5.20, the total reduction in the solar gains absorbed and transmitted into the room is decreasing the resultant temperatures considerably in the office space, however in average, they are still 2.5°K higher than as in the north facing shoebox. A possible explanation for this could be the higher amount of solar incident radiation on the south façade and/or the shading devices on the north façade being more effective.
WWR: 100%
100% 75% 35%
Fig 5.19 Solar gains (absorbed + transmitted) reduction (Source: EDSL Tas)
WWR: 100% + shading
WWR: 50% + shading
Fig 5.20 Typical summer week thermal simulation (Source: EDSL Tas) 59
5.4.1.3 East and west Figures 5.22 and 5.23 show that as a result of the high incident solar radiation, the east and west orientations present the highest values of solar gains in the basecases in comparison to other orientations (45kWh/day in comparison to 30 and 28 kWh/day for the south and north orientation respectively). Additionally, the contribution of the shading devices in terms of blocking both direct and diffuse solar radiation is much more effective in this orientation, reducing the solar gains by 65%. In this case the reduction in the WWR is only helping to decrease the solar gains by 10%. Note that the WWR considered for these two orientations is 75%, not 50% as in south and north, as previously explained in the daylight analysis part. Figures 5.21 and 5.24 are showing that the 75% reduction in the solar gains is dropping the resultant temperature by up to 12°K in the east façade and 16,5°K in the west. The reason of the difference between two orientations is because the west façade receives higher amounts of incident radiation. Ultimately, one can conclude that the strategy of reducing the WWR is more important in the south and north orientation than in the east and west.
WWR: 100%
WWR: 100% + shading
Fig 5.21 East orientation: Typical summer week thermal simulation (Source: EDSL Tas) 60
WWR: 75% + shading
kWh/Day
kWh/Day
100%
100%
35%
35% 25%
Fig 5.22 East facade: Solar gains (absorbed + transmitted) reduction (Source: EDSL Tas)
WWR: 100%
25%
Fig 5.23 West facade: Solar gains (absorbed + transmitted) reduction (Source: EDSL Tas)
WWR: 100% + shading
WWR: 75% + shading
Fig 5.24 West orientation: Typical summer week thermal simulation (Source: EDSL Tas) 61
5.4.2 With occupancy + Ventilation After assigning proper WWR’s and shading devices to each orientation, the next step of the analytical work was to verify how such an office space would thermally perform with a more realistic scenario, where occupancy related gains and ventilation were considered. As seen in Figure 5.25, in this case the shoebox had again its plan depth increased from 6 m to 12 m, which meant that it now represents a module of a typical rectangular shape office building typology. Two types of working environments were defined, cellular and open plan offices, which allowed for assessing and comparing the effect of cross and single sided ventilation. The density considered in the open plan office was the same as previously mentioned (6 m² per occupant). It is known that in most of the offices, the workstations are not occupied full time. However, there is a new trend that is becoming more common nowadays, in which more than one employee share the same workstation. This approach is based on the premise of increasing space efficiency, thus achieving higher energy savings. This way, for the internal heat gains definition, full occupancy was considered. Based on the findings from the daylight analyses, it was assumed that the artificial lights were on for only 2 hours per day after 17:00. For all the simulations a weather file was generated in Meteonorm 6.1, which has measurements for a 30 year period (from 1960 to 1990). The comfort band used in this research is based on the EN 15251 equation, which defines Thermal Neutrality as: TN =18.8 + 0.33Trm, where Trm= weighted running mean of the daily external temperature. In addition, the comfort band was defined by adding + and -3°K to the calculated TN, which would result in a PPD of 10%. As previously mentioned, exposed thermal mass was used in order to reduce diurnal temperature swings. For this section of the analytical work, the effect of the placement and distribution of the thermal mass was examined. Two scenarios were compared; one with the
Fig 5.25 TAS Model
density Open Plan: 6 m²/Person Cellular: 20 m²/Person appliances Open Plan: 22.5 W/m² (CIBSE Guide A) Cellular: 10 W/m² (CIBSE Guide A) Lights: 12 W/m² (CIBSE Guide A) 62
Fresh Air: 30 m³/h - Person 2 hours of artificial lights (17:00 - 19:00) u – values Wall: Concrete Block + Render– 2.64 W/m².°C Glass: Single Glazing – 5.50 W/m².°C Floor: Concrete Slab + Raised Floor - 1.43 W/m².°C Internal Partitions: Single Glazing 5.50 W/m².°C
exposed thermal mass only on the ceiling, and the other one with the thermal mass distributed throughout the walls, floor and the ceiling. The results revealed no significant difference between the two (less then 1째K); thus it was concluded that the ceiling is the most important area to have the exposed thermal mass, since the warm air tends to go up. Therefore the first scenario was kept for the following simulations. This also allowed the floor to be considered as raised floor, which is a common application in office buildings. (see Appendix II) The following set of simulations investigates the effect of ventilation in a typical summer week. The basecase has ventilation rates defined in order to comply with minimum fresh air requirements (1,6 ac/h) during the occupied hours; while the second model had an additional night time cooling of 10 ac/h. As expected, the temperatures in the second case are lower. However, it is also evident that this strategy alone is not enough to reduce daytime peak temperatures caused by high internal gains. Figure 5.26 shows that the resultant temperatures go up to 5째K above the comfort band.
Fig 5.26 Typical summer week thermal simulation (Source: EDSL Tas)
The next step was to increase the night time ventilation rate to 20 ac/h and the day time to 10 ac/h. Figure 5.27 shows that in this case the resultant temperatures remained within the comfort band during most of the occupancy hours, apart from some peaks in the hottest days.
Fig 5.27 Typical summer week thermal simulation (Source: EDSL Tas) 63
After that, as seen in Figure 5.28 the daytime ventilation rate was increased to 30 ac/h. This strategy helped to decrease the peak temperatures only by 1°K. Based on this finding, it is possible to assume that with ventilation rates above 20 ac/h (for day and night), no significant temperature reduction will be achieved.
Fig 5.28 Typical summer week thermal simulation (Source: EDSL Tas)
After having a profound indication of the proper ventilation rate required for space cooling, new simulations were carried out with assuming the windows opened, instead of having a set value for the number of air changes per hour. The upper window (clerestory) was considered to be open 24 hours per day in summer, which would simplify issues related to its operation since it is located at 2,5 m above the floor level. Additionally, the lower window pane was assumed to be opened only during the occupancy hours. It is worth mentioning that in both windows an aperture factor of 0.3 was applied, which meant to represent the effective opening area of a top hung or bottom hung window, so that the rain issues can be avoided. Figure 5.29a is showing that although lower ventilation rates are found in the cellular office in comparison to the open plan (cross ventilated), in most of the hours the resultant temperatures in the open plan space are slightly higher due to the higher density. Additionally, it can also be observed that in this case the temperatures in both spaces are within the comfort band for the great majority of the time, apart from few hours in one day, where they only exceed the threshold by 1°K. Also it is important to highlight that, in a one year period the resultant temperatures are only exceeding the threshold of 29.8°C (EN 15251) in 1.5% of the occupied hours (2610 h) in the open plan office and 1% in the cellular office. These higher temperatures will occur during 10 days of the year, in the summer season, between 12:00 and 15:00 and most of the hours they are between 30 and 30.5 °C, reaching up to 31°C for few hours. On the other hand, when the same shoebox is facing east and west, slightly higher temperatures are found as seen in the Figure 5.29b, which resulted in increasing the amount of overheating hours to 5.2%, with temperatures reaching up to 32°C for few hours. The number of overheating hours in the case of cellular office is 3.2%.
64
Upper Aperture: Op. factor 0.3 – 24hs Lower Aperture: Op. factor: 0.3 – Occupancy Hs
Fig 5.29a North/South Orientation: Typical summer week thermal simulation (Source: EDSL Tas)
Fig 5.29b East/West Orientation: Typical summer week thermal simulation (Source: EDSL Tas)
Although there are a certain number of overheating hours, the cooling loads can easily be dismissed using adaptive opportunities. As Baker (2007) states, desk/ceiling fans or a free dress code can each give an additional 2.5°K tolerance regarding temperature excursions. Additionally, a similar conclusion was stated in ASHRAE Standard 55 (2010), which acknowledges the effect of air movement in increasing the acceptable operative temperature limits. It is mentioned that an indoor air speed of 1.2 m/s can increase the acceptability for 2.2°K, which in this case would be enough to offset the residual overheating hours.
65
5.4.3 Winter studies So far the strategies were focused on cooling, whereas it was not checked upon whether blocking direct solar radiation and lowering the WWR would create heating loads. Therefore the final set of simulations investigates this aspect in a typical winter week. For this set of simulations, only minimum fresh air requirements were fulfilled and no further ventilation was considered. As seen in Figures 5.30 and 5.31, the temperatures are within the comfort band during the occupied hours for both orientations, due to high internal heat gains. Also, 0 heating loads with a thermostat setpoint of 19.5째C guarantees that this conclusion applies to the entire winter. This finding reveals that double glazing and insulated walls are not required in this particular case.
Fig 5.30 North/South Orientation: Typical winter week thermal simulation (Source: EDSL Tas) Thermostat set on 19.5째C: 0 heating loads
Fig 5.31 East/West Orientation: Typical winter week thermal simulation (Source: EDSL Tas) Thermostat set on 19.5째C: 0 heating loads 66
5.5 CFD analysis and comfort assessment In this step of the analytic work, a more detailed window design was proposed for each orientation, where three different window typologies were applied to: Bottom hung, top hung and horizontal pivot (see Fig. 5.32). Their distribution throughout the envelope followed a set of criteria; in which providing controllability for a wide range of air speeds was the priority, in addition to achieving adaptability to different types of office layout and preventing rain issues. After that, a wind rose for the summer months was created based on São Paulo’s weather data extracted from Meteonorm 6.1, where the average air speeds and the frequency of occurrence were identified in each orientation (Figure 5.33). The wind rose provided the basic inputs for the CFD simulations where the impact of airflow distribution in the work environment was assessed. It is worth mentioning that for all the orientations assessed (N, S, E, W), the prevailing wind direction was always oblique to the façade, and three scenarios were tested for each case: Low wind speed (1.2 m/s), average wind speed (varies according to orientation) and high wind speed (7,5 m/s). Since this study focused on wind driven ventilation, most of the simulations were carried out in an open plan office layout, which allows cross ventilation. Few simulations regarding single sided ventilation were conducted (see Appendix III), and similarly to the thermal simulation findings, the ventilation rates were much lower in this case in comparison to cross ventilated spaces. An upper threshold of 1,2 m/s (ASHRAE -55) was considered as draught limit throughout the space, whereas on the workplane level, the adopted strategy was to keep the air speed lower than 1 m/s in order to avoid inconvenience by blowing papers. The physiological cooling effect of air movement was assessed using the predicted mean vote (PMV), which uses seven-point thermal sensation scale to define comfort, where 3=hot, 2=warm, 1=slightly warm, 0=neutral, -1=slightly cool, -2=cool, and -3=cold. As described by Nicol et.al (2012), the votes outside the central 3 points of the scale (-1, 0, +1) are counted as dissatisfied, while the other are considered as acceptable. It is known that different researches concluded that the PMV model present inconsistencies in terms of assessing occupant satisfaction in a real building since it doesn’t take into consideration the psychological aspect of comfort such as availability of occupant control and adaptive opportunities. However it can provide a good indication regarding occupants’ thermal sensation since it is one of the few methods that take into account all physical and physiological variables that have impact on comfort: air temperature, mean radiant temperature, humidity, air movement, clothing value and activity rate.
Horizontal Pivot
Top Hung
Bottom Hung Fig 5.32 Window typologies (Source: CIBSE AM 10)
Fig 5.33 Wind rose for the summer months in São Paulo (After Meteonorm 6.1)
pmv inputs Air temperature: 29.8°C Mean radiant temperature: 26.9°C Humidity: 51% Clo value: 0.7 Activity rate: 1.0 Met Air speed: According to the scenario The Votes were calculated using the ISO 7730 – 1993 comfort calculator. The inputs for the calculation were extracted from EDSL Tas, from a typical peak hour of a hot summer day. 67
5.5.1 North
Fig 5.34 Internal view, north facade
Figure 5.34 confirms that in the north orientation, thanks to the horizontal configuration of the transparent elements, visual connection with outdoors is quite unobstructed. Thus, in order to enhance this strong visual connection the number of frames was reduced to a minimum. On the work plane level, bottom hung and horizontal pivot windows were applied: While fully glazed bottom hung windows offer a better view and provide the ventilation rate required for space cooling, the pivot windows allow a higher degree of air flow control. On the sill level a lower opaque vent was used for two reasons: First of all, the vent can be used to dissipate the excess heat generated by occupants and equipments in the lower level, as the simulations carried out for Rogers Stirk Harbour + Partners field work revealed. Secondly, they give flexibility to the faรงade; in case the room layout is changed to cellular, this small vent could contribute to enhance ventilation due to stack effect, since the distance between this window and the top one is bigger than 1,5 m (see Chapter 2). It is worth mentioning that the lower vent strategy will be applied to all orientations. The prevailing wind direction considered in the simulations is NE, and average wind speed is 4.59 m/s. Figure 5.35 shows the operation of windows Fig 5.35 Window operations for the tested scenarios (low, average and high in detail, for the three wind speed scenarios that were tested. wind speeds; respectively) 68
5.5.1.1 Low Wind Speed Scenario (1.2 m/s)
m/s
N
Figure 5.36a. Plan cutting on the workplane level.
0.7 pmv 0.7 ppd 15.3%
pmv 0.7 ppd 15.3% 0.8
0.7
pmv 0.7 ppd 15.3%
0.7
0.8
0.7
0.6
pmv 0.8 ppd 18.5%
pmv 0.8 pmv 0.8 ppd 18.5% ppd 18.5%
pmv 0.8 ppd 18.5%
0.7
0.7
pmv 0.6 ppd 12.5% 0.8
pmv 0.8 pmv 0.7 ppd 18.5% ppd 15.3%
An average air speed of 0,5 m/s is found resulting in an average PMV of 0.7 and PPD of 15.3%. It can be seen in the figure that the workstations close to the windows benefit from higher air velocities, which increase the occupant satisfaction to a PMV of 0.6. On the other hand, it can also be seen that the bottom hung windows create low air speed zones.
Figure 5.36b. Cross section showing that the air turbulence adjacent to the faรงade is being avoided thanks to the layout.
Figure 5.36c. Detail section showing that the windows can be fully open due to low wind speed.
69
5.5.1.2 Average Wind Speed Scenario (4.59 m/s)
m/s
Figure 5.37a. Plan cutting on the workplane level. An even air distribution in the space with an average air speed of 0.8 m/s results in a PMV of 0.6 in most of the areas. Note that this is slightly higher in comparison to low wind speed scenario. Although some parts have 1.2 m/s of air speed, which is slightly above the accepted workplane level threshold, it increases the PMV to 0.5, and PPD drops to 10.2%, which is not found in the previous scenario.
Figure 5.37b. section.
Detail
Note that the lower pivot windows are closed to prevent potential draughts on the workplane level.
70
0.6
0.6 pmv 0.5 ppd 10.2%
0.6
0.6
0.6
0.6
pmv 0.7 ppd 15.3% 0.6
0.6 pmv 0.6 ppd 12.5%
0.6
pmv 0.6 ppd 12.5% pmv 0.5 pmv 0.7 ppd 10.2% ppd 15.3%
Cross
Although there are air speeds higher than comfort limits, it is still acceptable because thanks to the window typology, the air stream is being redirected upwards.
Figure 5.37c. section.
pmv 0.6 ppd 12.5%
pmv 0.6 ppd 12.5%
pmv 0.6 ppd 12.5%
pmv 0.7 ppd 15.3%
pmv 0.6 ppd 12.5%
N
5.5.1.3 High Wind Speed Scenario (7.50 m/s)
m/s
N
Figure 5.38a. Plan cutting on the workplane level.
pmv 0.6 ppd 12.5% 0.6
0.6
0.6
0.6 pmv 0.5 ppd 10.2%
pmv 0.6 ppd 12.5% 0.5
Although the air velocities are generally higher, similar results as previous scenario are found, with an average PMV of 0.6.
Figure 5.38b. section. pmv 0.5 ppd 10.2% 0.6
0.5 pmv 0.6 ppd 12.5%
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
pmv 0.6 pmv 0.7 ppd 12.5% ppd 15.3% 0.6
0.6
pmv 0.7 pmv 0.6 ppd 15.3% ppd 12.5%
Cross
For this scenario, the lower pivot panes and vents were closed and the effective opening areas of the upper pivot panes were kept to a minimum. The ventilation is regulated mainly through bottom hung and clerestory windows. Any larger opening, even in the upper pivots, could easily result in excessive air speeds; since even with the minimum aperture factor some zones with high air velocities can be observed in the plan.
Figure 5.38c. Detail section. As previously explained in the thermal simulations, the clerestory window is open 24 hours throughout the summer, making the operation easier for the occupants. Additionally, the top hung window typology offers good protection from the rain, so that the occupants will not be forced to close the windows every time it rains.
71
5.5.2 South
Fig 5.39 Internal view, south facade
In the south faรงade, the window typologies are the same as the ones applied in the north orientation; however in this case, the vertical shading devices create some obstruction to the views as seen in Figure 5.39. The similarity between the two orientations (north and south) allows for comparisons in terms of air flow distribution, where the effect of the vertical fins can be clearly observed. The wind direction considered in these simulations is southeast, which according to the wind rose is the most frequent wind direction (28% of the time). The average air speed is 2,83 m/s. Figure 5.40 shows the operation of windows in detail, for the three wind speed scenarios that were tested.
Figure 5.40. Window operations for the tested scenarios (low, average and high wind speeds; respectively) 72
5.5.2.1 Low Wind Speed Scenario (1.2 m/s)
m/s
N
Figure 5.41a. Plan cutting on the workplane level.
0.7 0.7
pmv 0.8 ppd 18.5% 0.6
0.7 pmv 0.7 ppd 15.3%
0.8 pmv 0.6 ppd 12.5%
pmv 0.8 ppd 18.5% 0.7
pmv 0.8 ppd 18.5% 0.7
0.8 pmv 0.7 ppd 15.3%
0.8 0.7
The wind direction is also 45° to the façade as it was in the north façade simulations; however it can be observed that the interior air flow angle is different. The reason behind this is that the vertical fins are redirecting the wind. Additionally, the average air speed is lower than the one in the north scenario, since the shading devices are decreasing the air intake blocking the openings on the sides of windows.
Figure 5.41b. Cross section showing a variation in the PMV range between 0.7 and 0.8. pmv 0.7 pmv 0.7 ppd 15.3% ppd 15.3%
pmv 0.8 pmv 0.8 ppd 18.5% ppd 18.5%
Figure 5.41c. Detail section showing that the horizontal pivot windows can achieve a high effective opening area and still provide rain protection. However it is worth mentioning that under heavy tropical rains the effective opening area would need to be reduced and in that case the ventilation would be regulated through the lower opaque vent, bottom hung window and clerestory window (top hung). 73
5.5.2.2 Average Wind Speed Scenario (2.83 m/s)
m/s
Figure 5. 42a. Plan cutting on the workplane level. In the average wind speed scenario, the wind speed is much lower (2.83 m/s) than the one in north (4.59 m/s). Together with the effect of shading devices, this results in much lower air velocities and higher PMV’s in the space.
0.7 pmv 0.8 ppd 18.5%
0.6
0.6
pmv 0.7 ppd 15.3% 0.8
0.7 pmv 0.8 ppd 18.5%
pmv 0.6 ppd 12.5%
0.6
pmv 0.7 ppd 15.3%
pmv 0.6 ppd 12.5%
pmv 0.7 ppd 15.3% 0.8
pmv 0.7 ppd 15.3% 0.6
Figure 5.42b. Cross section.
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
Figure 5.42c. Detail section showing the benefit of dividing the pivot windows into two panes instead of having a larger one. Under high speed conditions the lower pane can be kept closed to avoid excessive air speeds on the workplane level, while the upper one still provides enough air movement for occupant comfort. Additionally, a bigger pane can obstruct the corridor between the workstations and façade when it’s open. 74
pmv 0.7 pmv 0.6 ppd 12.5% ppd 15.3%
N
5.5.2.3 High Wind Speed Scenario (7.50 m/s)
m/s
N
Figure 5.43a. Plan cutting on the workplane level.
0.6 0.6
0.5
0.6
pmv 0.6 ppd 12.5% 0.6
pmv 0.5 ppd 10.2% 0.6
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
0.6
pmv 0.7 ppd 15.3%
pmv 0.5 ppd 10.2%
0.6
0.7
pmv 0.6 ppd 12.5%
0.5
pmv 0.7 ppd 15.3% pmv 0.6 ppd 12.5%
pmv 0.6 ppd 12.5%
An even air flow distribution is observed, which is an improvement in comparison to the low wind speed scenario. Note that in that scenario, the bottom hung windows create zones with lower air speed. However in this case, higher wind speed compensates for that and air distribution becomes more homogeneous.
Figure 5.43b. Although the window operation is the same as north faรงade in high wind speed scenario, the opaque vent on the sill level is open in the case of south in order to compensate for the reduction in the effective opening area caused by the vertical fins. Even though the air speed is high, the bottom hung vents can redirect the wind upwards, preventing draughts.
Figure 5.43c. Detail section. 75
5.5.3 East
Fig 5.44 Internal view, east facade
The shading device typology that is expected to be used on an east facing faรงade is usually vertical fins, which offer a lower quality of views. However, in this case the use of horizontal louvers allow for good skyline views. Moreover, as will be explained in the following set of simulations, they also enhance the average indoor air speed and air flow distribution, especially when the wind speed is low (see Fig. 5.44). The shading devices also defined the arrangement of windows on the faรงade. Since the transparent elements were already divided in three panes by the overhangs, the strategy that was used to minimize the obstruction that the frames would cause was to use only wider horizontal pivot windows. The wind direction considered in these simulations is also southeast. Figure 5.45 shows how the windows can be operated in different wind speed scenarios.
Figure 5.45. Window operations for the tested scenarios (low, average and high wind speeds; respectively) 76
5.5.3.1 Low Wind Speed Scenario (1.2 m/s)
m/s
N
Fig 5.46a. Plan cutting on the workplane level.
pmv 0.6 ppd 12.5% pmv 0.7 ppd 15.3%
pmv 0.7 ppd 15.3%
0.7
pmv 0.7 ppd 15.3%
pmv 0.7 pmv 0.6 ppd 15.3% ppd 12.5%
0.7
0.7
0.7 0.7
pmv 0.7 ppd 15.3%
pmv 0.7 ppd 15.3%
0.7
0.7
pmv 0.7 ppd 15.3%
0.7 0.7
0.7
pmv 0.7 ppd 15.3%
The best air flow distribution is found in this scenario, since all of the windows can be fully open. Normally, as explained for the south faรงade, this could create a conflict due to tropical rains, but in this particular case, the horizontal louvers already provide protection, allowing the use of maximum opening area.
Fig 5.46b. Cross section. It is noticeable that the solar control elements are not obstructing the air stream.
Fig 5.46c. Detail section. It is possible to observe that the pivot windows allow for a larger effective opening area.
77
5.5.3.2 Average Wind Speed Scenario (2.83 m/s)
m/s
Fig 5.47a. Plan cutting on the workplane level. Similar to the previous scenario, a high uniformity in air distribution is achieved, with PMVs of 0.6 throughout the space. This conclusion is important since this scenario represents the most frequent wind direction and speed.
0.6
pmv 0.6 ppd 12.5%
0.6
0.6
0.6
pmv 0.6 ppd 12.5%
0.6
0.6
pmv 0.6 ppd 12.5% 0.6
pmv 0.6 ppd 12.5% 0.6
0.6 0.6
0.6 pmv 0.7 ppd 15.3%
Fig 5.47b. Cross section.
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
Fig 5.47c. Detail section. In comparison to the previous scenario, the only difference in the operation of windows is the closed lower pane on the level of the workplane.
78
pmv 0.6 pmv 0.7 ppd 12.5% ppd 15.3%
N
5.5.3.3 High Wind Speed Scenario (7.50 m/s)
m/s
N
Fig 5.48a. Plan cutting on the workplane level.
pmv 0.6 ppd 12.5% 0.6
pmv 0.6 ppd 12.5% 0.6
0.6 0.6
0.6 0.6
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
0.7
0.6
pmv 0.7 pmv 0.6 ppd 15.3% ppd 12.5%
Unlike the previous scenarios, the air distribution is not as even. However the PMVs do not vary that much and they stay at 0.6 for most of the occupants.
pmv 0.8 pmv 0.7 ppd 18.5% ppd 15.3% pmv 0.6 0.6 ppd 12.5%
pmv 0.7 pmv 0.7 ppd 15.3%ppd 15.3%
Fig 5.48b. Although the effective opening areas were kept to a minimum in order to avoid draughts, due to the high wind speed it is still possible to achieve adequate ventilation rates.
Fig 5.48c. Detail section. This is the only case so far that requires a change in the position of the top windows. However, this is not a scenario that is likely to happen frequently.
79
5.5.4 West
Fig 5.49 Internal view, west facade
As it was expected the west facing faรงade is providing poor views availability, due to its vertical shading devices configuration (see Figure 5.49). In this case, again the bottom hung windows with higher glass panes were used with the intention of minimizing the amount of frames in the transparent elements. In addition, pivot windows were also applied in order to increase the effective opening area and also providing possibilities of redirecting the air streams according to the occupants cooling needs. The prevailing wind direction in this case is southwest, and the average wind speed is 1,37 m/s. Since the average wind is speed is already low, only two scenarios were simulated: low wind speed and high wind speed. Figure 5.50 is showing the operation of windows in the two wind speed scenario simulated.
Figure 5.50. Window operations for the tested scenarios (low, average and high wind speeds; respectively)
80
5.5.4.1 Low/Average Wind Speed Scenario (1.37 m/s)
m/s
N
Figure 5.51a. Plan cutting on the workplane level.
0.7
pmv 0.8 ppd 18.5%
pmv 0.7 ppd 15.3%
0.8
0.8
0.8
pmv 0.7 ppd 15.3%
0.8
pmv 0.8 ppd 18.5%
0.8
0.8
0.7 0.8
pmv 0.8 ppd 18.5%
pmv 0.8 ppd 18.5%
0.8
Because the south west wind is perpendicular to the shading devices, the majority of air intake is blocked, resulting in the lowest air speeds found so far.
Figure 5.51b. section.
pmv 0.7 pmv 0.8 ppd 15.3% ppd 18.5%
pmv 0.8 pmv 0.8 ppd 18.5% ppd 18.5%
Cross
As a consequence of poor ventilation rates, the highest PMVs of mostly 0.8 are found throughout the space.
Figure 5.51c. section.
Detail
The effect of shading devices is highlighting the importance of the lower vents in this case, since they are the only unobstructed air intake source.
81
5.5.4.2 High Wind Speed Scenario (7.50 m/s)
m/s
Figure 5.52a. Plan cutting on the workplane level. Although the effect of the shading devices is being mitigated with high wind speed, according to the climate data this scenario is not very common.
0.8
pmv 0.8 ppd 18.5%
0.8
0.8
pmv 0.6 ppd 12.5%
0.6
0.6
pmv 0.6 ppd 12.5%
pmv 0.6 ppd 12.5%
0.6
0.8
pmv 0.8 ppd 18.5%
0.7
0.6
pmv 0.7 ppd 15.3%
0.6
Figure 5.52b. Cross section. Similar to the east faรงade, the shading devices protect the interiors from the rain, thus allowing the windows to be opened fully.
Figure 5.52c. Detail section. Unlike the east faรงade, in the high wind speed scenario a reduction in the opening area of the top windows is not required since the shading devices are considerably reducing the wind speed. This is beneficial in terms of operation since it eliminates the necessity of interaction.
82
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
pmv 0.6 pmv 0.6 ppd 12.5% ppd 12.5%
N
5.5.5 Conclusions From the simulations it is possible to conclude that the bottom hung window typology (which was applied to the north, south and west façades) reduces the ventilation rates and creates low air speed zones especially when the external wind speeds are lower. However, its applicability can still be justified through the fact that its inward opening geometry directs the air flow upwards, thus bigger glass panes can be used without worrying about draughts. Reducing the amount of frames obstructing the views in the transparent element is particularly relevant in this case since these new façade proposals are aiming to supersede the traditional fully glazed façade. Additionally, it can also be stated that this window typology will reduce the need for occupant operation since its geometry already deals with a range of issues by default: protection from the rain, provision of adequate amount of ventilation rates for space cooling (see thermal simulation analyses) and draught prevention. Finally, it is possible to conclude that even though the low air speed zones are found during less windy days, this scenario is not typical according to the wind rose and even in the worst case (wind coming from the west façade) PMV levels are still lower than 1 and PPD is lower than 20%, which is the threshold recommended by ASHRAE. It is also worth mentioning that in terms of physiological cooling, the use of desk fans can still compensate for lower indoor speeds. Regarding occupant comfort, it is important to highlight that even though the average relative humidity in São Paulo is 80% according to the weather data, Tas simulations showed that during the warmest hours, due to the high internal gains, the humidity in the work environment drops considerably reaching around 50% when the air temperature is close to 30°C. This reduction has a significant contribution in enhancing occupant comfort. Assessing the overall potential for achieving comfort ventilation under high summer temperatures, one can conclude that the east facing façade is the one which offers the best air distribution and ventilation rates. This finding becomes even more valuable considering that east receives the most frequent incident winds, which account for a total of 60% (SE 28%, NE 18%, E 14%). Additionally, this will also have a bigger impact since the number of overheating hours is higher in the east-west oriented shoebox in comparison to south-north. Although it was shown in the daylight analyses that illuminance levels are within the defined comfort range, it is known that reflection from external surfaces with high albedo could cause glare. For these occasions internal blinds would be required. Since inward opening geometry (for the pivot and bottom hung windows) would conflict with the typical internal blinds, a internal blind that is fixed to the window frames is proposed as recommended in CIBSE Guide AM 10.
83
84
6 research outcomes & applicability 6.1 Research Outcomes 6.2 Applicability 6.2.1 Cost Analysis
85
6 research outcomes & applicability 6.1 Research Outcomes Regarding daylight availability, the reduction in the WWR proved to be effective in terms of reducing over illumination issues. Additionally, it was found that the minimum acceptable levels of daylight can be achieved even under overcast sky conditions with low sky illuminance levels, thus the offices are able to achieve independence from artificial lighting for 80% of the occupied hours. The thermal simulations carried out testing solar control strategies reveals that in São Paulo’s climate, applying shading devices is not a sufficient measure in terms of minimizing solar gains, since a considerable portion of them are a consequence of diffuse radiation. It is worth remembering, that this statement is based on the application of clear glazing, since it is known that the solar control glasses reduce the light transmittance, increasing the use of artificial lights which have a significant impact in the internal heat gains and energy consumption. Additionally from the simulations it was possible to identify that the importance of reducing the WWR is more significant in north and south orientations, where it was possible to drop the solar gains by 30% and 40% respectively, in comparison to a fully glazed façade with shading devices. On the other hand, for east and west façades, this reduction is less relevant, since the shading devices are already decreasing the solar gains by 65% and the reduction in the WWR is only contributing to increase the efficacy in 10% (see Figure 6.1). It is important to highlight that despite the fact that a higher solar gains decrease was achieved in east and west orientations ( 75% ), the north and south orientations have the lower amount of solar gains transmitted and absorbed into the space, since these two orientations receive less solar radiation throughout the summer. The second step of the thermal performance analysis (including occupancy related gains) demonstrated that the offices can perform as free running throughout the year. In the winter even in the cellular offices which have lower density, 0 heating loads were found showing that the mild winter temperatures combined to high internal gains are allowing the work environments to not rely on solar gains or construction techniques (insulation, double glazing) to achieve independence of space heating. On the other hand, in the summer the internal heat gains revealed to be a bigger challenge than the climate itself, since in some warm days, even using thermal inertia and high ventilation rates, a small amount of overheating hours can be still be found. It is important to highlight that orienting the building due north/south will result in a smaller number of overheating hours in comparison to the east/ west orientation (1, 5% against 5, 2%). As mentioned previously such small amount of overheating hours does not justify the use of air conditioning since adaptive opportunities like a desk fan would be able to enhance occupants comfort and offset the overheating (discomfort) conditions. The CFD simulations combined with the comfort assessment (PMV) confirmed the finds from the climate analysis and thermal simulations showing that comfort conditions can be found (PMV lower than 1) even with air temperatures higher than 29 °C , if there is availability of air movement and lower mean radiant temperatures. The low humidity rate which is a consequence of the high internal gains also plays an important role in enhancing the PMV levels.
86
In regards to the impact of the window design on air flow, it became clear that the east faรงade is allowing the best air flow distribution throughout the space in comparison to the other orientations, while the west is the worst case due to its shading devices configurations which besides obstructing the air stream, it also block the views. The north facing faรงade is the one which allows for the most uninterrupted visual connections with outdoor.
65% 75%
75%
70%
The CFD simulations also proved that the pivot window is the one which can provide the best air flow controllability, where the occupants can redirect the flow and also regulated the effective aperture area according to their needs. The bottom hung windows demonstrated a good potential for having bigger panes (allowing for better views) and avoiding draughts, thus providing adequate ventilations rates for space cooling with low needs of occupants operation. Regarding the lower opaque vent, it can be concluded that it is more relevant in the west facing faรงade since in that case is the only unobstructed inlet, however its applicability is more important when the open offices are converted into cellular ones. In the latter, the lower vent will enhance the room ventilation by stack effect.
Figure 6.1. Solar gains reduction for each orientation 87
6.2 Applicability As seen in the previous chapter the analysed shoebox revealed to be working efficiently in cellular and open plan office layouts. In the open plan scenario higher ventilation rates are achieved (through cross ventilation) which is fundamental to dissipate the high internal gains. On the other hand, even though in the cellular layout lower ac/h were found (single sided ventilation), this configuration presented lower resultant temperatures in the summer due to its lower density. In regards to the building shape, the cellular office shoebox could be applicable to any plan configuration that respects the principle of the defined passive zone, since it doesn’t rely on cross ventilation to achieve a good thermal performance. On the contrary, the open plan layout configuration which already had its efficacy tested in a typical rectangular building shape (see previous chapter) could have some issues in building shapes with central cores, such as the square shape typology. This issue will even grow bigger with tall buildings since the higher the building is, the bigger the core area becomes. So in that case, core could create a significant obstruction for the air stream, resulting in larger low air movement zones. In parallel, an interesting statement was made by Marcondes (2010) when she observed that in all the orientations the rectangular shape office building would perform better than a square one, in terms of number of overheating hours. Her explanation for that is not directly related to ventilation rates, but is based on the fact that the square shape has more exposed façades. Therefore one can conclude that ventilation would be even more important in the latter case to enhance occupant comfort. Ultimately, it is most likely that the square shape typology would perform better in low to mid-rise buildings, where the core dimensions can be moderate. Figure 6.2 shows two CFD simulations for a 7 storey-high square shape building . The simulations are showing that in this case just a small portion of the work environment is being obstructed by the core; however it is possible to achieve a better performance if the core allowed the air stream to cross through it. m/s
6.0 m
10.0 m
Figure 6.2. CFD simulations for the square shape typology. Wind direction: south east Wind speed 2.83 m/s Net lettable floor area: 412 m2 Source: Autodesk CFD 88
N
6.2.1 Cost Analysis After defining the most common building forms* where the shoebox could be applied (rectangular and square), a cost-effectiveness analysis was carried out in order to compare the operational and initial costs that those buildings would have in a fully glazed façade scenario and in another scenario which incorporates the façade design proposed in the previous chapter. Table 6.1 shows the materials and installation costs for different façade elements. Note that regarding the transparent elements, two types of glazing will be used (clear and reflective), while in the opaque elements three different finishing materials are proposed: render texture, aluminium panel and translucent glass, showing the different alternatives that are usually applied to office buildings**. Table 6.1. Costs for different façade elements
Figure 6.3. Different façade elements
* The square shape typology together with rectangular one, are the most common office building typologies in São Paulo. ** The cost of each façade element was obtained via direct contact with the manufacturer and the installation costs were provided by the developers BRA Engenharia. 89
6.2.1.1 Square Building Form
Nevertheless, the model proposed has proved to be able to perform as free running throughout the year. This will avoid the initial costs of the air conditioning equipment, as opposed to the fully glazed façade scenario, which in this case will have an overall initial cost much higher than the proposed model. Other construction costs such as structure, internal finishing, etc. were not taken into account in these calculations since they were considered to be the same in both cases.
22.50 m
6.00 m
Table 6.2 is showing that when the render texture is applied as finishing in the opaque elements, the initial costs of the proposed façade design is lower than fully glazed one. However, if aluminium or glass is applied in the opaque elements the costs will increase considerably and in both cases will be more expensive than the fully glazed. There are two aspects affecting this: First of all, clear glass with shading devices costs almost the double of reflective glass, and additionally, although relatively cheaper render finishing compensates for the expensive shading devices, when aluminium and glass are applied, the overall costs increase even more.
22.50 m Figure 6.4. Plan, 7 storey square shape building Net lettable floor area: 412 m2 Overall lettable floor area: 2,884 m2
Regarding the operational costs the table is showing that by applying the proposed strategies it was possible to achieve a saving of 80% in lighting energy consumption, and the energy saving achieved by not using air conditioning is even more impactful, since it resulted in an overall annual save of US$ 96,037.52. It is worth mentioning that even if the client decides to use air conditioning in the proposed scenario and opts for the most expensive façade; just the savings achieved by minimizing the use of artificial lighting would pay the extra façade cost (in comparison to the fully glazed) in 6.5 years.
OPERATIONAL COSTS
INITIAL COSTS
FACADE AREAS
WWR
Table 6.2. Initial costs vs. operational costs
90
100%
50% (North - South) + 75% (East - West)
Fully Glazed
Render Texture
Aluminium Panel
Translucent Glass
1,877.50 m²
1,173 m²
1,173 m²
1,173 m²
–
704.50 m²
704.50 m²
704.50 m²
420 m²
420 m²
420 m²
420 m²
Transparent Elements Type + Cost/m²
Reflective Glass US$ 380.00
Clear Glass + Shading US$ 617.50
Clear Glass + Shading US$ 617.50
Clear Glass + Shading US$ 617.50
Opaque Elements Type (Wall) + Cost/m²
–
C. Block + Render US$ 92.50
C. Block + Aluminium US$ 256.50
C. Block + T. Glass US$ 284.50
Op. Elements Finishing (Structure) + Cost/m²
Reflective Glass US$ 380.00
Render Texture US$ 38.00
Aluminium Panel US$ 202.50
Translucent Glass US$ 230.00
Total Facade Costs
US$ 873,050.00
US$ 805,453.75
Air Conditioning Equipment Cost
US$ 650.000,00
–
Total
US$ 1,531,600.00
US$ 805,453.75
US$ 986,436.75
US$ 1,021,357.75
Annual Lighting Energy Consumption + Costs
55 Kwh/m² US$ 28,551.60
11 Kwh/m² US$ 5,710.00
11 Kwh/m² US$ 5,710.00
11 Kwh/m² US$ 5,710.00
Annual Cooling Loads + Costs
141 Kwh/m² US$ 73,195.92
–
–
–
Total
US$ 101,747.52
US$ 5,710.00
US$ 5,710.00
US$ 5,710.00
Transparent Elements Area Opaques Elements Area (Walls) Opaque Elements Area (Exposed Structure)
US$ 986,436.75
US$ 1,021,357.75
–
–
Figure 6.5a. External perspective of south and east faรงades Transluscent glass is used as finishing on the opaque elements in this proposal.
Figure 6.5b. West faรงade partial view 91
For the rectangular building form it is possible to have a cheaper façade if the building is oriented toward north and south, since the overall WWR is smaller in this case, which would result also in a reduced amount of shading devices. The calculations presented in Table 6.3 do not take into account the cores of the building, which in these cases are usually independent from main building body. The east-west orientation was chosen to be studied in the cost analysis since the WWR is bigger (75%) and therefore this can be considered as the worst case scenario for this specific shape. The initial façade costs in this case can be even cheaper than in the square shape if render texture is applied, since the narrow sides of the building can be completely opaque without interfering in the daylight and natural ventilation performance. The operational costs will be slightly smaller than in the square shape building since the cooling loads in this case are also smaller (less exposed façades). The annual savings achieved in this case is US$ 87,212.48, which is still a very significant amount since it represents around 10% of the entire façade costs.
OPERATIONAL COSTS
INITIAL COSTS
FACADE AREAS
WWR
Table 6.3. Initial costs vs. operational costs
100%
34.30 m
6.2.1.2 Rectangular Building Form
12 m
Figure 6.6. Plan, 7 storey rectangular shape building Net lettable floor area: 412 m2
75% (East - West)
Fully Glazed
Render Texture
Aluminium Panel
Translucent Glass
1,932.40 m²
1,071.40 m²
1,071.40 m²
1,071.40 m²
–
861 m²
861 m²
861 m²
450.40 m²
450.40 m²
450.40 m²
Transparent Elements Type + Cost/m²
Reflective Glass US$ 380.00
Clear Glass + Shading US$ 617.50
Opaque Elements Type (Wall) + Cost/m²
–
C. Block + Render US$ 92.50
C. Block + Aluminium US$ 256.50
C. Block + Glass US$ 284.50
Op. Elements Finishing (Structure) + Cost/m²
Reflective Glass US$ 380.00
Render Texture US$ 38.00
Aluminium Panel US$ 202.50
Opaque Glass US$ 230.00
Transparent Elements Area Opaques Elements Area (Walls) Opaques Elements Area (Exposed Structure)
Clear Glass + Shading Clear Glass + Shading US$ 617.50 US$ 617.50
Total Facade Costs
US$ 905,464.00
US$ 758,347.23
US$ 973,642.00
Air Conditioning Equipment Cost*
US$ 650.000,00
–
–
Total
US$ 1,555,464.00
US$ 758,347.23
US$ 973,642.00
Annual Lighting Energy Consumption** + Costs
55 Kwh/m² US$ 28,551.60
11 Kwh/m² US$ 5,710.00
Annual Cooling Loads + Costs
124 Kwh/m² US$ 64,370.88
–
Total
US$ 92,922.48
US$ 5,710.00
450.40 m²
11 Kwh/m² US$ 5,710.00
US$ 1,010,136.00
US$ 1,010,136.00 11 Kwh/m² US$ 5,710.00
–
–
US$ 5,710.00
US$ 5,710.00
* The costs of the air conditioning equipment for the fully glazed building was specified by the company Contractors Engenharia (from Sao Paulo), based in the cooling loads provided by the author of this dissertation (obtained from EDLS Tas). ** The annual lighting energy consumption for the fully glazed office building (55 kWh/m²) was specified based on the benchmark indicated in the Energy Consumption Guide 19 (2003). UK Energy Efficiency Best Practice programme. The costs for the energy consumption in Sao Paulo ( kWh) was obtained from Aneel (National Agency for Electrical Energy) web site. And cross checked with an energy bill where the taxes were included. (US$ 0.18 / kWh). 92
Figure 6.7a. External perspective showing the north faรงade. Translucent coloured glass is used as a finishing on the opaque elements, while clear glass is used in the transparent ones.
Figure 6.7b. Perspective of the east faรงade.
93
94
7 conclusions
95
96
7 conclusions The application of a proper window design can have a significant impact in office buildings in terms of enhancing environmental quality and occupants’ satisfaction. The window is seen today by most of the architects as a universal element that could be applied to all contexts by only changing the glass type. However, throughout this dissertation, it was possible to understand that the window is the most vital link with the climate, therefore in order to achieve a proper design it must be specific to its context and a wide range of aspects must be considered in the design process including the internal layout, building shape, climatic constraints, building function, etc. Many times conflicts amongst these aspects are likely to be found, however the resolution is complex and in such cases it is necessary to prioritise. Additionally, apart from the qualitative benefits, it was found that the proposed façade typology for São Paulo can have a much lower initial cost when compared to the conventional fully glazed office building (if the cost of the systems are also considered), while it is still possible to achieve significant energy savings. It is worth mentioning that the efficacy of this strategy especially when it is related to superseding air conditioning, towards the adoption of natural ventilation strategies, will depend on a higher acknowledgment of the adaptive standard of thermal comfort by architects, engineers and developers. The government can also play an important role in this scenario, by increasing the electricity costs, which would make the savings achieved even more relevant and the option of environmentally controlled building less viable. Ultimately, although this dissertation aimed to tackle a wide range of aspects related to the topic of window design in work environments, it is important to highlight that some specific factors related to the urban context that were not considered in this scope such as, external air quality (pollution), noise and obstructions (overshadowing) can also have a significant impact in the window design and in the applicability of the passive strategies proposed (daylight and natural ventilation). Additionally, the topic of window operation and the limitations of manual and automated control could still be explored in depth. These remaining aspects open possibilities for furthers studies in this field, where the window element can still be the central pillar for the investigation towards bioclimatic techniques and sustainable design.
97
98
bibliography
99
Arens, E., Z. Hui, et al. (2006). Partial- and whole-body thermal sensation and comfort Part I: Uniform environmental conditions. Journal of Thermal Biology 31: 53-59. Arens, E; Turner, S; Zhang, H; Paliaga, G. (2009). Moving air for Comfort. In: ASHRAE Journal, May 2009. ASHRAE (2010) ANSI/ASHRAE Standard 55R – Thermal Environmental Conditions for Human Occupancy, Atlanta: American Society of Heating, Refrigerating and Air‐Conditioning Engineers, Inc. Baker, N., & Standeven, M. (1994). Thermal comfort for free-running buildings. En PLEA, Energy and Buildings (Vol.23, págs. 175 - 182). Dead Sea, Israel: Passive Low Energy Architecture. Baker,N and K. Steemers (2002). Daylight Design of Buildings. James & James Science Publisher. Baker, N.V (2009). A Handbook of Sustainable Refurbishment: Non-Domestic Buildings. Earthscan Boerstra et al.(2011). Impact of Available and Perceived Control on Comfort and Health in European Office Buildings. Proceedings of 7th Windsor Conference: The changing context of comfort in an unpredictable world. Windsor, UK. Bordas,W (1999). Best practice programme. General information report 56.Energy efficiency. BRE. Brager, Gail & de Dear, Richard (2001).Operable windows, personal control and occupant comfort. Center for Environmental Design Research.UC Berkeley CIBSE (1999). Daylighting and Window Design. Lighting Guide LG10. Chartered Institution of Building Services Engineers, London. CIBSE lighting guide 7 (2005). Office lighting.Chartered Institution of Building Services Engineers, London. CIBSE (2005). Natural ventilation in Non-Domestic Buildings. Application Manual AM10. Chartered Institution of Building Services Engineers, London. CIBSE (2006). Environmental design. Guide A, 7th Edition. Chartered Institution of Building Services Engineers, London. Chappels, H. and E. Stove (2004). Comfort: a review of philosophies and paradigms. Future Comforts Project, Uk ESRC programme. Choi,J et al. (2010) Post-occupancy Evaluation in 20 Office Buildings towards Future IEQ Standards and Guidelines. Cohen, J (2002). World population in 2050: Assessing the projections. Columbia University. Givoni,,B (1994). Passive and Low Energy Cooling of Buildings. Van Nostrand Reinhold Ghiabaklou, Zahra (2010). Natural Ventilation as a Design Strategy for Energy Saving. World Academy of Science, Engineering and Technology 71 2010 Gonçalves, J.C (2010). The environmental performance of tall buildings. Earthscan 100
Hedge A., Burge P.S., Robertson A.S., Wilson S. and Harris-Bass J., 1989, Work-related Illness in Offices: A proposed model of the ‘Sick Building Syndrome’, Environment International, Vol. 15, pp. 143-158. Holger Koch-Nielsen ( 2002). Stay Cool. A design guide for the built environment in hot climates. James & James Ltd. Keighley, E.C. (1973). Visual requirements and reduced fenestration in offices: a study of window shape, Building Science 8 311–320. Khalil,N et al.(2009) Post Occupancy Evaluation towards Indoor Environment Improvement in Malaysia’s Office Buildings. University Technology of MARA Perak, Malaysia. Lamberts,R. Dutra,L. & Pereira.F.O.R (1997). Eficiencia Energetica na Arquitetura.”Energy efficiency in architecture”. Translation. PW Editores. Loftness, V., & Hartkopf, V. G. (2003). Linking Energy to Health and Productivity in the Built Environment. Pittsburgh,PA: School of Architecture, Carnagie Mellon. Lamberts et. al (2011). Towards a Brazilian standard for naturally ventilated buildings: guidelines for thermal and air movement acceptability. In Adaptive comfort. Routledge. Marcondes, M (2010). Soluções Projetuais de Fachadas para Edifícios de Escritórios com Ventilação Natural em São Paulo. Phd Dissertation. University of São Paulo. Ne’eman, E. & Hopkinson, R.G. (1970). Critical minimum acceptable window size:a study of window design and provision of view, Lighting Research and Technology. pp17–27. Nicol,F and S. Roaf (2007). Adaptive Thermal Comfort and Passive Architecture. In Advances in Passive cooling. Earthscan. Nicol, J.F. (Ed. 2011). Adaptive Comfort. Special issue of Building Research Information Journal, Vol. 39, No2. Routledge. Nicol, F.; Humphreys, M; Roaf, S. (2012) Adaptive Thermal Comfort: Principles and Practice. Osterhaus, W.K.(2001). Discomfort glare from daylight in computer offices:What do we really know? in: Proceedings of the 9th European Lighting Conference (Lux Europa), Reykjavik, Iceland Sobin, H. J. 1980. Window Design for Passive Ventilative Cooling: Na Experimental Study. Passive Cooling Applications Handbook, Prepared for the Passive Cooling Workshop Amherst, Massachusetts. Wells, B.W.P (1996). Subjective responses to the lighting installations in a modern office building and their design implications, Building and Environment page 57–68. Yannas, S. (Ed. 2000). Designing for Summer Comfort. EC Altener Programme. Environmental & Energy Studies Programme, AA Graduate School, London. Yannas, S. ( 2008). Challenging the supremacy of Airconditioning. 2A Architecture and Art, Issue 7, pp20-43, Dubai. Zhang et al.(2011). Air movement preferences observed in office buildings. Windsor: workings of the CBE Thermal Comfort Workshop.
101
102
appendices Appendix I: Office density analyses Appendix II: Thermal mass analyses and Envelope Exposure Studies Appendix III: CFD simulations - single sided office layout
103
appendix I: Office density analyses
Figure I.a. Floor Area: 290 m² (excluding cores) Work Stations: 53 Density: 5.47 m ² / Employee Source: Oliveira Cotta Architects
Figure I.b. Floor Area: 1738 m² (excluding cores) Work Stations: 314 Density: 5.50 m ² / Employee Source: Oliveira Cotta Architects
Figure I.c. Floor Area: 1625 m² (excluding cores) Work Stations: 245 Density: 6.60 m ² / Employee Source: Oliveira Cotta Architects
104
appendix II: Thermal mass analyses and Envelope Exposure Studies
Figure II.a. Thermal simulation for a typical summer week showing that there is no significant difference between the two scenarios tested (Thermal mass). Source: Edsl Tas
Figure II.b. Thermal simulation for a typical winter week showing that the offices placed in the top floor would have higher heat losses due to its higher envelope exposure and this would reduce the resultant temperature by around 2 K in comparison to an office located in a middle floor. Source: Edsl Tas
105
appendix III: CFD simulations - single sided office layout
m/s
Figure III.a. Cross section. CFD simulations for a cellular office facing north. It is possible to observe that the single sided ventilation is not allowing the air stream to reach the back of the room. Source: EDSL Ambiens
Figure III.b. Cross section.CFD simulations for a cellular office facing north in typical summer hour. It is possible to observe that the ventilation rate is not enough to dissipate the internal heat gains. Source: EDSL Ambiens
106