Clothing insulation as a behavioural adaptation for thermal comfort in Indian office buildings

Proceedings of 9 th Windsor Conference: Making Comfort Relevant Cumberland Lodge, Windsor, UK, 7-10 April 2016. Network for Comfort and Energy Use in Buildings, http://nceub.org.uk Clothing insulation as a behavioural adaptation for thermal comfort in Indian office buildings Rajan Rawal 1, Sanyogita Manu 1, Yash Shukla 1, Leena E. Thomas 2, Richard de Dear 3 1 Centre for Advanced Research in Building Science and Energy, CEPT University, Ahmedabad, India 2 School of Architecture, Faculty of Design Architecture and Building, University of Technology Sydney 3 Faculty of Architecture, Design and Planning, The University of Sydney, Sydney Abstract Regulating clothing is one of the most obvious behavioural responses to changing thermal conditions. The extent of clothing, in turn, affects thermal sensation and acceptability. A lack of extensive thermal comfort field studies in India has meant that there has been very limited data on clothing related occupant behaviour in Indian offices until now. This paper aims to understand clothing norms and practices in Indian offices using data gathered via an extensive field study of thermal comfort in India. It uses the office occupants response to thermal sensation, acceptability and preference questions as experienced right here, right now from more than 6000 surveys together with simultaneous measurement of environmental conditions, clothing and metabolic activity. These surveys are administered in five climate zones across three seasons in air-conditioned, naturally ventilated and mixed mode buildings. The paper analyses clothing insulation as a behavioural response to changes in the environment. The variation in clothing insulation with observed indoor and outdoor temperature is analysed for different seasons, building types and cities. The study also examines the extent of behavioural regulation in clothing between the male and female office workers. The results suggest that women tend to wear lower clothing insulation on an average in summer compared to men. In naturally ventilated and mixed mode buildings, variability in clothing insulation was higher compared to air conditioned buildings, emphasizing the role of clothing as an adaptive measure. Keywords: Clothing, Behavioural adaptation, Thermal comfort, Indian offices, Office users 1 Introduction India s electricity demand is expected to rise from 775 TWh in 2012 to 2499 TWh by 2030. This along with pledge by India to reduce emission intensity of India s GDP by 33-35% by 2030 from the 2005 level India s as part of Intended Nationally Determined Contributions (INDC) building energy efficiency becomes important mitigation tool to achieve intended goals (Government of India 2015). Estimates by National Institution for Transforming India (NITI Aayog) indicates that the mitigation activities for moderate low carbon development would cost India around USD 834 billion till 2030 at 2011 prices. An adaptive model of thermal comfort recognises that thermal comfort requirements of people depend on their past and present context and that these vary with the outdoor environmental conditions of their location. This concept can play a major role in reducing energy use whilst maintaining the comfort, productivity and well-being of occupants. Thermal neutrality can be achieved for a wide range of outdoor conditions by harnessing measures Windsor Conference 2016 - Making Comfort Relevant - Proceedings 403 of 1332

such as change in clothing insulation level (clo) and activity, operation of fans and windows (de Dear & Brager 1998). Amongst the various adaptive actions performed by building occupants, adjustment in clothing is one of the most practiced action to achieve thermal comfort (Feriadi & Hien 2004). There are multiple field studies from different geographical regions suggesting that indoor and outdoor temperatures are important determinants of clothing behaviour. In a metaanalysis of more than 21,000 from four continents covering a wide range of climate zones, 66% of the variance in clothing insulation worn indoors could be accounted for by the regression model where the independent variable was mean indoor operative temperature (de Dear & Brager 1998). In the same study, 40% of the variance in clo values was explained by variations in outdoor effective temperature in an exponential decay curve while a straight regression model accounted for 44% of variance in naturally ventilated buildings. A field study in Australia provides evidence that outdoor temperatures significantly influence clothing levels (Morgan et al. 2002). A study done in Libya found a correlation between clothing insulation and both the outdoor temperature (R 2 =0.492) and the indoor globe temperature (R 2 =0.519), with average clo values ranging from 0.55-0.62 across the three locations (Akair & Bánhidi 2007). A field study in Tunisia documented a large scatter of clothing insulation ranging from 0.7 clo in summer to 1.8 clo in winter and reported a more robust correlation between clothing insulation and indoor temperature (R 2 =0.52) as compared to outdoor temperature (R 2 =0.5) (Bouden & Ghrab 2005). Clothing insulation values of male and female respondents was found to be similar in a field study in Seoul, Korea, but the clo value of female respondents was slightly higher than that of male respondents in winter. Female respondents also seemed to change their clo value gradually with the seasonal changes, unlike make respondents. The study also reported a decrease in clo value in both summer and winter from 1980 to 2009 (Bae & Chun 2009). A field experiment in Taiwanese classrooms reported that students adjusted their clothing according to the indoor temperature. AC classrooms, which also had cooler conditions, had higher clothing levels than NV classrooms. More importantly, it showed that female students adjusted their clothing levels more swiftly in response to the indoor temperature (Hwang et al. 2006). Another field study reported that occupants would accept the thermal environment by adjusting clothing insulation value for an operative temperature up to 29 C in residential apartments Hong Kong (Lai et al. 2009). A field study done in 25 office buildings in a warmhumid and a composite climatic zone location reported overall clothing insulation values ranging from 0.49 to 0.97 clo units. The study was conducted in summer and south-west monsoon seasons. Clothing insulation correlated rather poorly with temperature in both the cities (Indraganti et al. 2013). Recognising importance of clothing in providing insulation to human body, thermal comfort guidelines and standards documents such as ANSI/ASHRAE 55-2013, TM52 by CIBSE, ISO 7730: 2005, ISO 9920:2007 and BS EN 15251-2007 have regarded clothing insulation values to determine appropriate thermal conditions. Indians wear a wide range of clothing attire that stems for the country s cultural, socioeconomic and climatic diversity as well as from the variety of clothing material and degree of customization on offer. Over the years, international clothing styles and norms have been seamlessly assimilated in the day-to-day life, both at home and work. An extensive project Windsor Conference 2016 - Making Comfort Relevant - Proceedings 404 of 1332

with multiple field studies in Indian offices was undertaken by the authors of this paper which resulted in an India-specific adaptive thermal comfort model (Manu et al. 2016) for naturally ventilated and mixed mode buildings and demonstrated that Fanger's static PMV model consistently over-predicts the sensation on the warmer side of the 7-point sensation scale even in AC buildings. In order to implement these models, it is important to understand the clothing practices in Indian work spaces and variation in clothing insulation with season and building type. It is also important to understand how clothing behaviour changes with indoor and outdoor environmental conditions. This paper focuses on these aspects of adaption in Indian office buildings. 2 Methods This study is based on the data collected for the larger IMAC (India Model for Adaptive Comfort) field study conducted from 2011-2014 across India to develop an India specific model for adaptive thermal comfort (Manu et al. 2016). More than 6000 Right here, right now surveys were administered along with concurrent indoor environmental measurements in 16 office buildings in India. These buildings were located across five Indian cities that were selected as representative locations within five distinct climate zones of India (Bansal & Minke 1995; Bureau of Indian Standards 2005). The surveys were repeated in three seasons summer, monsoon and winter in each building. 2.1 Surveys The right here, right now surveys were administered to gather a respondent s assessment of her/his immediate thermal environment at work space at the time of the survey in three office building types naturally ventilated (NV), mixed-mode (MM) and air conditioned (AC) buildings. The survey questionnaire included the ASHRAE 7-point thermal sensation scale of warmth ranging from cold (-3) to hot (+3) with neutral (0) in the middle. This was a continuous scale allowing non-integer ratings, however very few respondents used that option. The other questions were related to thermal acceptability, preference and general comfort. Clothing and activity of each subject was also recorded on the questionnaire. The survey was administered in an interview format where the field researcher read the questions to the subject and noted the responses on the form manually. The respondents were observed unobtrusively and the clothing garment checklist was filled-in on the questionnaire by the researchers. The interview and physical measurements were completed in about 5-10 minutes per subject. The respondents were interviewed on the questionnaire at the same time as their workstation environment was being measured using the hand-held instruments. 2.2 Measurements Indoor climate measurements were recorded using hand-held equipment at each subject s workstation while the survey response was taken. That meant that each set of measurements was spatially and temporally coincident with the occupant location. Based on the categorization of field studies by Brager & de Dear (Brager & de Dear 1998), IMAC study was a Class II investigation. Extech HT30 Heat Stress WBGT Meter was used to measure three indoor environmental parameters - air and globe temperature and relative humidity. It is a hand-held instrument that can measure (black) globe temperature in the range of 0 to 80 C with an accuracy of ±2 C using a black globe of 40mm diameter. It measures air temperature in the range of 0 to 50 C Windsor Conference 2016 - Making Comfort Relevant - Proceedings 405 of 1332

with an accuracy of ±1 C and relative humidity in the range of 0 to 100% with an accuracy of ±3%. For globe temperature measurements, the globe was give 5~10 minutes to reach equilibrium. TSI VELOCICALC Air Velocity Meter 9525 was used to measure indoor air velocity. It is a handheld instrument and uses a telescopic probe to measure air velocity in the range of 0 to 50 m/s with an accuracy of ±3% of the reading. Two measurements were taken at each position, the first one parallel to the ceiling (sensor was horizontal) and second parallel to wall (vertical) according to wind source. U12-012data loggers were used to measure and store outdoor air temperature and humidity data for each location. In each city, a safe location, an office or a residence was identified to install the loggers. Periodic checks were performed either by the field researchers or the owners of the property. The positioning of the loggers was chosen based on the daily activities so it won t be disturbed by the owners. They were installed in semi-open spaces shielded from direct solar radiation. 2.3 Clothing insulation An extensive clothing garment checklist was prepared for the survey questionnaire. It included Indian garments for women and men, such as sari, kurta, pajama, etc. For each garment, the field researchers also indicated if it was light weight, medium weight or heavy weight, on the questionnaire. Clothing insulation (clo) values were assigned to each garment based on the lists published in ASHRAE Standard 55-2010 (ASHRAE 2010). For garments not listed in the standard, clo values were interpolated from those of the existing garments. The total clo for each respondent was calculated by adding the clo values of individual garments and undergarments (Table 2). To account for insulation provided by a cushioned chair, a clo value of 0.15 was added to the total clo from the garments. Table 3 lists the classification of all chair types documented during the field surveys and the corresponding insulation values assigned to each type. It is important to note here that the IMAC estimates of sari ensemble insulation range from 0.61-0.77 clo (lightweight heavyweight). This range is very similar to the results from the manikin experiments done by Havenith (Havenith et al. 2015) where the sari ensemble insulation was 0.74 clo and those conducted by Indraganti where she proposed ensemble values of 0.65 (pleated pallu) and 0.74 (unpleated pallu) (Indraganti et al. 2015). Table 1 Clothing garments checklist and insulation (clo) values Clothing Light weight Medium weight Heavy weight Description from ASHRAE Standard 55-2010 Petticoat 0.15 0.15 0.15 Baniyan/undershirt 0.06 0.06 0.06 Short sleeved shirt/kurta 0.19 0.24 0.28 Short-sleeve dress shirt Long sleeved shirt/kurta 0.25 0.3 0.34 Long-sleeve dress shirt and Long-sleeve flannel shirt Pants 0.15 0.2 0.24 Straight trousers (thin and thick); 3 for jeans Pajama 0.12 0.16 0.21 Pajama/salwar/churidar 0.13 0.18 0.22 Scarf/dupatta 0.04 0.08 0.13 Hijab 0.06 0.1 0.15 Windsor Conference 2016 - Making Comfort Relevant - Proceedings 406 of 1332

Clothing Light weight Medium weight Heavy weight Description from ASHRAE Standard 55-2010 Blouse (for sari) 0.12 0.16 0.19 Sleeveless/scoop-neck blouse and Short-sleeve dress shirt Sari 0.3 0.35 0.39 Dress 0.33 0.4 0.47 Long-sleeve shirtdress (thin and thick) Skirt 0.14 0.19 0.23 Skirt (thin and thick) Long sleeved sweater 0.25 0.3 0.36 Long-sleeve (thin and thick) Vest/waistcoat/sleeveless 0.13 0.18 0.22 Sleeveless vest (thin and thick) sweater Jacket 0.36 0.4 0.44 Single-breasted (thin and thick) Shawl 0.27 0.3 0.35 Shorts 0.08 0.13 0.17 Walking shorts Dhoti 0.15 0.15 0.15 Turban 0.08 0.12 0.17 Tie 0.01 0.03 0.05 Thermal underwear - 0.2 0.2 0.2 upper Thermal underwear - 0.15 0.15 0.15 lower Socks 0.03 0.05 0.06 Calf-length socks, Knee socks (thick) Stockings 0.02 0.04 0.06 Pantyhose/stockings Shoes 0.02 0.06 0.1 Shoes and Boots Chappals/sandals 0.02 0.02 0.02 Sandals Clo values in highlighted cells have been interpolated from ASHRAE 55-2010 values. Others have been taken as they were in ASHRAE 55-2010. Table 2 Undergarment Clo values Clothing clo Bra 0.01 Panties 0.03 Men's briefs 0.04 Clo values have been taken from ASHRAE 55-2010 Table 3 Chair insulation Chair Code Chair Description insulation (clo) P1 Plastic moulded chair 0 P2 Plastic moulded chair used with a cushion for seat 0.15 W1 Wooden framed chair with open mesh weave for seat and back rest 0 Windsor Conference 2016 - Making Comfort Relevant - Proceedings 407 of 1332

Chair Code Chair Description insulation (clo) W2 Wooden framed chair with open mesh weave for seat and back rest used with a cushion for seat/back rest and hand rest 0.15 M1 Metal framed chair with open mesh weave for seat and back rest 0 M2 Metal framed chair with open mesh weave for seat and back rest 0.15 O1 Revolving chair 0 O2 Revolving chair used with a cushion for seat/back rest and arm rest 0.15 3 Results Table 4 presents a statistical summary of the clothing ensemble insulation (including chair insulation) for the three seasons. Clo value ranged from 0.38 to 2.24, with an average of 0.79 clo units. Mean clo in summer and monsoon was 0.64 and 0.69 which is higher than the ASHRAE 55-2010 assumed summer value of 0.5 clo. The winter season s average was 1.05 clo. The standard deviation (SD) in clo was higher in winter (0.42 clo), almost thrice the SD in summer and monsoon (0.1-0.12 clo), indicating greater variability in clothing ensemble in winter as increased number of layers lead to more freedom in clothing adjustments. Female respondents wore lower clothing insulation on an average in summer when compared to male respondents. Similar trends were evident in monsoon and winter. Lower insulation in response to warmer temperatures indicates fewer layers of clothing resulting in diminished opportunity to adjust clothing further. In spite of wearing lower insulation than male respondents, the variability in clo was slightly higher among female respondents in summer and monsoon. In winter, however, clothing insulation among female respondents showed less variability as compared to the male respondents. Table 4 Clothing insulation statistical summary by season and gender Summer Winter Monsoon Male Female Combined Male Female Combined Male Female Combined N 1467 636 2103 1434 633 2067 1452 708 2160 Avg. 0.66 0.57 0.64 1.08 0.97 1.05 0.70 0.65 0.69 Max. 1.02 1.00 1.02 2.24 2.07 2.24 1.29 1.35 1.35 Min. 0.40 0.38 0.38 0.43 0.38 0.38 0.40 0.38 0.38 SD 0.09 0.10 0.10 0.42 0.40 0.42 0.11 0.14 0.12 The IMAC dataset was disaggregated into building operation types NV, MM and AC in Table 5 to present a statistical summary of the clothing ensemble insulation. Average Clo in AC buildings was lower than in NV and MM buildings. The variability (SD) in clo was much higher in NV and MM buildings as compared to AC buildings, for both male and female respondents. This indicates that the respondents changed clothing frequently to adapt to the varying indoor and outdoor conditions in NV and MM buildings. Respondents in AC buildings experienced similar indoor temperatures across the year which may have resulted in low variability in clo. All AC buildings in the dataset were corporate offices with a business attire dress code resulting in restricted opportunities for adaptive clothing behaviour. These buildings also maintained the indoor temperatures with a narrow range conditioning the occupants to similar thermal environment throughout the year, therefore, making personal adaptive measures, such as changing clothing insulation, redundant. Windsor Conference 2016 - Making Comfort Relevant - Proceedings 408 of 1332

Table 5 Clothing insulation statistical summary by building operation type and gender NV Buildings MM Buildings AC Buildings Male Female Combined Male Female Combined Male Female Combined N 1312 693 2005 1875 601 2476 1166 683 1849 Avg. 0.82 0.75 0.80 0.83 0.78 0.82 0.77 0.66 0.73 Max. 2.16 2.06 2.16 2.24 2.07 2.24 1.66 1.93 1.93 Min. 0.40 0.38 0.38 0.40 0.38 0.38 0.55 0.53 0.53 SD 0.36 0.32 0.35 0.35 0.38 0.36 0.15 0.16 0.16 Seasonal clothing insulation trends are similar between NV and MM buildings as shown in Figure 1. Winter clo values in AC buildings are lower than NV and mm. All buildings dataset has pooled survey responses from NV, MM and AC buildings. Average clo for male respondents was always higher compared to female respondents across all building operation types and seasons, except in winter in MM buildings. In addition, the variation in average clo values between seasons was least in AC buildings. For instance while there was a reduction of 0.21 clo from winter to summer for males and 0.16 clo for females in AC buildings, the corresponding change in NV buildings was 0.5 for males and 0.43 for females, and 0.48 for males and 0.61 for females in MM buildings. Average clo was highest in winter in MM buildings 1.15 for male and 1.16 for female respondents. While the average clo values were not too dissimilar in NV buildings (1.13 and 1.01 for male and female respectively), these values dropped to 0.09 and 0.76 for male and female in AC buildings. Standard deviation in clo insulation, was the highest for male respondents in NV buildings in winter (SD=0.47) and lowest for male respondents in AC buildings in summer (SD=0.05). (a) (b) (c) (d) Figure 1 Clothing insulation mean and standard deviation for NV, MM and AC buildings by season and gender Windsor Conference 2016 - Making Comfort Relevant - Proceedings 409 of 1332

3.1 Distribution of clothing garments Figure 2 plots the percentage distribution of selected clothing garments for female and male respondents for the three seasons. This distribution presents an interesting view into the clothing patterns and behaviour of office workers in India. From Figure 2a shows that less than 15% of the total female respondents (n=1977) were wearing a sari at the time of the survey. This percentage did not vary significantly from one season to another. Salwar/chudidar was the most widely worn garment, worn with short or long sleeved kurta/shirt to form a complete ensemble. 75% were wearing salwar/chudidar in summer, 73% in monsoon and 65% in winter. Reduced numbers of respondents wearing salwar/chudidar in winter may be explained by an increase in those wearing pants in that season (27%). Another important garment seems to be the scarf or the dupatta, worn by more than 60% of the female respondents in summer and around 45% in monsoon and winter. Sandals seem to be the footwear of choice with more than 90% respondents wearing them in summer and monsoon. This percentage was understandably lower in winter (61%) where 39% were reported wearing shoes. All male respondents were wearing pants at the time of the survey (Figure 2b). 70% of them were wearing long sleeved shirt with pants in summer and monsoon. This percentage increased to 87% in winter. More than 80% of the total male respondents (n=4353) were wearing shoes in summer and monsoon and 93% in winter. (a) (b) Figure 2 Percentage distribution of female and male clothing garments 3.2 Clothing insulation adjustments with change in indoor temperature Figure 3 presents weighted linear and exponential trend lines plotted between the mean level of clothing insulation worn for a building + season aggregate and its mean indoor temperatures for male and female respondents. This aggregate has data points from a specific seasonal campaign within a building, leading to a total of 48 aggregates from 16 buildings and 3 seasons. The trend lines were plotted for NV, MM, AC and All buildings datasets. Windsor Conference 2016 - Making Comfort Relevant - Proceedings 410 of 1332

(a) (b) (c) (d) Figure 3 Male and Female clothing insulation inside buildings (mean ±stdev) as a function of mean indoor operative temperatures The graphs indicate a statistically significant relationship between clothing insulation and mean indoor operative temperature (TOP) for NV and MM buildings indicating a gradual decrease in clo values with increase in indoor TOP. For these building types, the exponential model provided a better fit than linear regression. The model for AC buildings failed to achieve significance possibly due to the narrow range of indoor temperatures encountered in these buildings as compared to the NV and MM building aggregates. The error bars on either side of the plotted points in the figure represent ±1 standard deviation around the within-aggregate mean. The standard deviation bars indicate the variability of clothing insulation which decreased as the indoor temperature increased. This probably shows the diminished freedom to adjust clothing as the number of garments in the ensemble reduced to the socially acceptable minimum dress standards. The regression models indicate that across all building types, male respondents wore higher clothing insulation as compared to the female respondents, irrespective of the prevalent indoor temperatures. In NV buildings, the linear regression models were significant and explained 53% variance in clothing insulation of male and female respondents with change in indoor operative temperatures. Mean clothing insulation decreased, on average, by 0.05 clo units for male respondents and 0.04 clo units for female respondents for every 1K increase in the mean Windsor Conference 2016 - Making Comfort Relevant - Proceedings 411 of 1332

indoor temperature, very similar to the de Dear s regression model for NV buildings from the RP-884 meta-analysis of global database (de Dear et al. 1997). The regression models for MM buildings were less robust for male respondents and more robust for female respondents compared to NV buildings. Linear models accounted for about 47% variance in clothing insulation of male respondents and 59% for female respondents with change in indoor operative temperatures. There was no discernable relationship between clothing insulation and indoor temperature in AC buildings. The slope of the regression models was marginally lower for female respondents compared to their male counterparts in NV buildings, while in MM buildings it was similar. With the increase in indoor temperature there was steeper decline in male clothing insulation indicating more layered clothing as well as a faster adjustment of clothing insulation to address the change in indoor temperatures as compared to the female respondents. At the same time, however, there was lower variability in the male clothing insulation at higher indoor temperatures in NV and MM buildings. The reason might be because the male respondents may be wearing the lowest clothing insulation that social convention may allow and any further reduction may not be possible in a work environment. 3.3 Clothing insulation adjustments with change in outdoor temperature Since the clothing insulation has a strong relationship with the indoor temperatures, clothing decisions and behaviour may also be expected to be influenced by outdoor weather conditions. Figure 4 presents weighted linear and exponential trend lines plotted between the mean levels of thermal insulation for each building + season aggregate against 30-day outdoor running mean air temperatures. The trend lines were plotted for NV, MM, AC and All buildings datasets. The graph for NV buildings indicates a statistically significant relationship between clothing insulation and outdoor temperature. The linear regression model accounts for 71% variance in the dependent variable for male respondents and 65% variance for female respondents. Exponential models provide a better fit than the straight line for male respondents explaining 76% of the variance in clo values. The regression models are very similar for male and female respondents. The linear regression models indicated a decrease of 0.04 clo units in mean clothing insulation for every 1K increase in the mean outdoor temperature. Regression models for MM buildings suggest that outdoor temperature is an important determinant of clothing insulation but the relationship is not as strong as in the case of NV buildings. In MM buildings, the linear regression models explained 60% of the variance in clothing insulation of male respondents and 65% of the variance in female respondents with change in mean outdoor temperatures. Mean clothing insulation decreased, on average, by 0.03 clo units for male and female respondents for every 1K increase in the mean outdoor temperature. Exponential models yielded better regression coefficients. Regression models for female group were stronger than the male group The linear regression models for AC buildings explained 60% of the variance clothing insulation worn by male and female respondents. The linear regression models predicted a reduction of 0.01 clo in mean clothing insulation for male and 0.02 clo for female respondents for every 1K increase in the mean outdoor temperature. Windsor Conference 2016 - Making Comfort Relevant - Proceedings 412 of 1332

(a) (b) (c) (d) Figure 4 Male and Female clothing insulation inside buildings (mean ±stdev) as a function of outdoor temperature 4 Discussion This paper offers a number of insights towards a better understanding of clothing insulation as a behavioural response to changes in the environment that are summarised below: The findings show that clo value of the IMAC database ranged from 0.38 to 2.24, with an average of 0.79. There was a greater variance in the clothing ensemble in winter owing to more clothing layers offering a greater possibility of adjusting clothing as a mechanism for thermal adaption. Female respondents wore lower clothing insulation on an average across the year as compared to male respondents. In summer and monsoon, clothing insulation among female respondents showed more variability (SD) as compared to the male respondents. Although the annual average clothing insulation did not vary significantly from one building type to another, the variability (SD) in clo was much higher in NV and MM buildings as compared to AC buildings, for both male and female respondents. On the other hand, the change in clothing insulation across seasons was more pronounced in NV and MM buildings than in AC buildings. Average clo for male respondents was always higher compared to female respondents across all building operation types and seasons except in winter in MM buildings. But Windsor Conference 2016 - Making Comfort Relevant - Proceedings 413 of 1332

the difference between the two genders was more pronounced in winter season and in AC buildings. The graphs indicate a statistically significant relationship between clothing insulation and mean indoor operative temperature (TOP) for NV and MM buildings indicating a gradual decrease in clo values with increase in indoor TOP. For NV and MM buildings, an exponential decay curve explained the relationship between clothing insulation and indoor temperature better than a straight-line model. With the increase in indoor temperature the decline in clothing insulation (slope of the regression model) was similar for male and female respondents. In NV buildings, the relationship between clo and outdoor temperature was stronger than the relationship between clo and indoor temperature, for both male and female respondents. This indicates that outdoor conditions play a very important role in clothing behaviour in NV buildings. Across all buildings types, the gradient of linear models between clo and outdoor temperature was similar for female and male respondents. As previously noted (Manu et al, 2016) fan and window operation and change in clothing are significant adaptive measures seen in office buildings in the dataset where Indian respondents were shown to tolerate a wider range of temperatures than would be predicted by Fanger s static PMV model in all building types. Nevertheless, the study has shown adaptive clothing behaviour to be the least evident in AC buildings. The impact of regulating clothing via business dress code coupled with a regulated and narrow range of indoor temperatures across the year as seen in the AC buildings, serves to restrict adaptive behaviour and opportunity. If AC buildings continue to be designed and operated along western standards, they have the potential to create a vicious circle of dependence on energy intensive means for regulating temperatures to achieve thermal comfort in these buildings. On the other hand, the results presented in this paper demonstrate the critical role clothing can play as an adaptive response. The ability to vary their clothing is an important factor in occupants ability to adapt to changing outdoor climate and indoor temperatures. As seen here, this is particularly relevant to deliver comfort in the case of NV and MM buildings where thermal conditions vary to a greater extent. Along with other aspects such as fan and window operation, it will be critical that contemporary workplaces offer the ability for adaptation through clothing to suit personal preferences and the Indian climatic and cultural context in contrast to a standardised business dress code. Such a user and climate responsive approach to building design and operation would go a long way in ensuring a sustainable future for the subcontinent. References Akair, A. & Bánhidi, L., 2007. Thermal comfort investigation in Libya. Periodica Polytechnica Mechanical Engineering, 51(1), p.45. ASHRAE, 2010. ANSI/ASHRAE Standard 55-2010 - Thermal Environmental Conditions for Human Occupancy, Atlanta: American Society of Heating and Air-Conditioning Engineers, Inc. Bae, C. & Chun, C., 2009. Research on seasonal indoor thermal environment and residents control behavior of cooling and heating systems in Korea. Building and Environment, 44(11), pp.2300 2307. Bansal, N.K. & Minke, G. eds., 1995. Climatic Zones and Rural Housing in India., Forschungszentrum Jülich GmbH, Zentralbibliothek. Windsor Conference 2016 - Making Comfort Relevant - Proceedings 414 of 1332

Bouden, C. & Ghrab, N., 2005. An adaptive thermal comfort model for the Tunisian context: a field study results. Energy and Buildings, 37(9), pp.952 963. Brager, G. & de Dear, R., 1998. Thermal adaptation in the built environment: a literature review. Energy and Buildings, 27(1), pp.83 96. Available at: http://www.sciencedirect.com/science/article/pii/s0378778897000534. Bureau of Indian Standards, 2005. National Building Code of India 2005 M. Kisan et al., eds., New Delhi: Bureau of Indian Standards. de Dear, R. & Brager, G., 1998. Developing an Adaptive Model of Thermal Comfort and Preference. ASHRAE Transactions, 104(1), pp.154 167. de Dear, R., Brager, G.S. & Cooper, D., 1997. Developing and Adaptive Model of Thermal Comfort and Preference. Final Report ASHRAE RP-884, Sydney, Australia. Feriadi, H. & Hien, N.H., 2004. Thermal comfort for naturally ventilated houses in Indonesia. Energy and Buildings, 36(7), pp.614 626. Government of India, 2015. India s Intended Nationally Determined Contribution, Available at: http://www4.unfccc.int/submissions/indc/published Documents/India/1/INDIA INDC TO UNFCCC.pdf. Havenith, G. et al., 2015. A database of static clothing thermal insulation and vapor permeability values of non-western ensembles for use in ASHRAE Standard 55, ISO 7730, and ISO 9920 CH-15-018 (RP-1504). ASHRAE Transactions, 121(1), pp.197 211. Available at: https://dspace.lboro.ac.uk/dspace-jspui/bitstream/2134/16724/3/2015 havenith et al transactions ASHRAE-D-CH-15-018.pdf. Hwang, R.-L., Lin, T.-P. & Kuo, N.-J., 2006. Field experiments on thermal comfort in campus classrooms in Taiwan. Energy and Buildings, 38(1), pp.53 62. Indraganti, M. et al., 2015. Thermal adaptation and insulation opportunities provided by different drapes of Indian saris. Architectural Science Review, 58(1), pp.87 92. Available at: http://dx.doi.org/10.1080/00038628.2014.976540. Indraganti, M., Ooka, R. & Rijal, H.B., 2013. Field investigation of comfort temperature in Indian office buildings: A case of Chennai and Hyderabad. Building and Environment, 65, pp.195 214. Available at: http://www.sciencedirect.com/science/article/pii/s0360132313001108. Lai, A.C.K. et al., 2009. An evaluation model for indoor environmental quality (IEQ) acceptance in residential buildings. Energy and Buildings, 41(9), pp.930 936. Manu, S. et al., 2016. Field studies of thermal comfort across multiple climate zones for the subcontinent: India Model for Adaptive Comfort (IMAC). Building and Environment, 98, pp.55 70. Available at: http://www.sciencedirect.com/science/article/pii/s0360132315302171 [Accessed January 23, 2016]. Morgan, C., de Dear, R. & Brager, G., 2002. Climate, Clothing and Adaptation in the Built Environment. In H. Levin, ed. Indoor Air 2002. Monterey, California, pp. 98 103. Available at: https://www.irbnet.de/daten/iconda/cib7766.pdf. Windsor Conference 2016 - Making Comfort Relevant - Proceedings 415 of 1332