Available online at www.sciencedirect.

com

Building and Environment 39 (2004) 1115 1123

www.elsevier.com/locate/buildenv

Comparative study of the indoor air quality of naturally ventilated and air-conditioned bedrooms of residential buildings in Singapore
Nyuk Hien Wong , Bernard Huang
Department of Building, School of Design and Environment, National University of Singapore, 4 Architecture Drive, Singapore 117566, Singapore Received 2 November 2002; received in revised form 25 February 2003; accepted 8 January 2004

Abstract Indoor air quality (IAQ) measurements were carried out in 3 residential dwellings in Singapore to nd out the indoor air quality (IAQ) of the bedrooms. It was found that when the subjects utilized air-conditioners while sleeping, there was a considerable build-up of carbon dioxide (CO2 ) ( 1000 ppm) for all types of air-conditioners. These CO2 levels were substantial higher as compared to naturally ventilated bedrooms. A survey was also conducted to investigate whether occupants exhibited sick building syndrome (SBS) symptoms while sleeping in air-conditioned (AC) as well as naturally ventilated (NV) bedrooms. A comparison showed that almost all occupants who used air-conditioners while sleeping exhibited one or more SBS symptoms and these occupants usually displayed more SBS symptoms after using air-conditioning than when they utilized natural ventilation. The survey also revealed that the frequency and duration of usage of air conditioning has important impact on the exhibition of the SBS symptoms. ? 2004 Elsevier Ltd. All rights reserved.
Keywords: Indoor air quality; Thermal comfort; Natural ventilation

1. Introduction During the past few decades, various symptoms and illnesses have increasingly been attributed to non-industrial indoor environments. In general, exposure indoors to noxious, chemical, physical and biological hazards is at a low level. However, unlike industrial or accidental exposures, such exposures are very common and usually sustained. Problems associated with the indoor environment are a common environmental health issue that is faced by doctors and health practitioners. This sick building syndrome (SBS) has received much attention over the years and, as originally dened by the World Health Organization (WHO) [1], refers to non-specic symptoms including eye, nose, and throat irritation, mental fatigue, headaches, nausea, dizziness and skin irritation, which seemed to be linked to the indoor climate [2]. Although not life-threatening nor disabling, the incidence of SBS among working people does have an economic repercussion [38]. If SBS can a ect the productivity of working people, then SBS that is experienced at home may also

Corresponding author. Fax: +65-67755502. E-mail address: [email protected] (N.H. Wong).

be carried forward to the workplace, thereby increasing the e ects of SBS on the persons productivity at their workplaces. SBS has over the years been identied as a significant problem in the UK and other European countries as well as in Canada, Australia, Japan and America. Singaporeans are becoming more a uent as the nation transits from a developing to a developed country status and can a ord more luxurious goods such as air-conditioners. In 1998, 57.7% of Singaporeans possess air-conditioners in their homes as compared to 19.4% in 1988 [9]. About 86% of Singaporeans live in public Housing and Development Board (HDB) ats [9] while the remaining 14% stay in high-rise condominiums or apartments and landed properties. As can be seen from the statistics, air-conditioners are no longer just restricted to the use by the rich who stay in houses and condominiums. Increasingly, Singaporeans staying in HDB ats are also buying more air-conditioners and there seems to be a trend towards the use of air-conditioners to alter the thermal conditions within the home, especially while sleeping. An average man spends up to 90% of his time indoors [10] means that both time spent at workplace and at home is included. Therefore, the indoor air in our dwellings, o ces, schools and other enclosed premises are of decisive importance for our health. The air-conditioners that are used in domestic homes are di erent from the ones

0360-1323/$ - see front matter ? 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.buildenv.2004.01.024

1116

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

used in o ce buildings. They are mainly multi-split or window unit types unlike the central air-conditioners, which have fresh air intake. Most domestic air-conditioners have no fresh air intake and this could a ect the indoor air quality (IAQ) of the bedrooms signicantly. The average Singaporean usually spends 8 h at work and almost 8 h sleeping at home. If the air-conditioner is switched on at night while sleeping, it will usually be on the whole night or at least for 46 h. This can result in a considerable build-up of CO2 in the bedrooms and perhaps other pollutants. Thus, there is a need to measure the IAQ of residential ats in Singapore to determine the impact of the use of air-conditioners while the occupants are sleeping and also in naturally ventilated conditions. Thus, the main objectives of this study are: 1. To determine the IAQ status of the residential bedrooms in both air-conditioned (AC) and naturally ventilated (NV) bedrooms. 2. To nd out whether SBS is more prevalent in NV or AC environments in residential homes. 3. To identify the common SBS symptoms that the occupants exhibit in both NV and AC environment. 4. To determine the relationship between the frequency and duration of usage of air-conditioner and the types or prevalence of SBS symptoms. 2. Data collection 2.1. Objective measurement The eld experiment was performed in 3 similar residential units at 3 locations from June to July 2001. The bedrooms had di erent types of air-conditioner so as to nd out the e ects of the type of air-conditioner on the IAQ of the bedrooms. Table 1 shows the information about the 3 residential units under investigation. For each residential unit, the following equipments were deployed to measure the various IAQ parameters [11]. These include: Q-Trak IAQ Monitor Model 8550/8551to measure the carbon dioxide, relative humidity, carbon monoxide and room temperature. Sampling of the parameters were done using automatic logging every minute over at least 8 h overnight. Dust Trak Model 8520to measure the particulate content within the room. Sampling was done using automatic logging every minute over at least 8 h. Anemomaster Model 8061to measure the wind velocity within the bedroom. Sampling was taken hourly. Single stage Anderson samplerto collect the sample for bacteria and fungi. Sampling was done at 3 intervals: when the air-conditioner was switched on, 3 h after the air conditioning was switched on and at 8am in the morning.

All the equipments were positioned in the middle of the room at a height of 1:2 m according to the measuring parameters set out by the guidelines for indoor air quality in o ces for small areas [12]. For each residential unit, the measurement was conducted simultaneously in two similar bedrooms, one with the air conditioning on and the other naturally ventilated. In order for the experiments to be consistent, some factors were kept constant where applicable: i. The bedrooms under investigation were all of similar size. ii. Each bedroom was to have two occupants. iii. Temperature of the air-conditioner was set at 22 C wherever applicable. iv. Windows were opened wide wherever possible for natural ventilation. v. The speed of the air-conditioner was set at high. vi. The measuring equipments were set up in the middle of the bedroom.

2.2. Subjective measurement For the subjective measurement, a questionnaire survey was formulated to nd out the IAQ perception of the occupants and whether they exhibited any SBS symptoms in NV dwellings and air-conditioned dwellings [1316]. The objectives of the survey are summarized as follows: i. To nd out whether NV homes have lesser SBS occurrences than in AC homes. ii. To nd out the common SBS symptoms su ered by Singaporeans in NV and AC Homes. iii. To nd out whether the duration of the air-conditioner being switched on has an e ect on the prevalence of SBS symptoms. A total of 300 questionnaires were sent out and 105 occupants surveyed responded. The basic demographic data is summarized in Table 2. 3. Results 3.1. Objective measurements 3.1.1. Carbon dioxide (CO2 ) proles and concentration levels Figs. 1ad show the CO2 proles and concentration levels of the bedrooms utilizing the di erent air-conditioning systems. Each of the gure also shows the corresponding CO2 proles and concentration levels of the bedroom utilizing only NV.

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123 Table 1 Data on the residential units under investigation Case 1 Location Telok Blangah Rise Types of air conditioning system Vertical casement (no fresh air intake) Photos of air conditioning system Floor height 8th

1117

Measuring period and time June 22nd25th 10pm8am

2A/2B

Bukit Merah Central

Window unit Air vent opened Air vent closed

18th

July 2nd & 3rd 10pm8am

Bukit Merah Central

Portable air conditioner (with fresh air intake)

18th

July 2nd & 3rd 10pm8am

Yishun Avenue 7

Multi-split air conditioner (no fresh air intake)

16th

July 5th & 6th 10pm8am

Table 2 Demographic data for the samples of respondents Type of Ventilation Gender Age 20 Natural ventilation only Air-conditioner and natural ventilation Male Female Male Female 1 4 0 2 2130 49 45 32 22 3140 3 1 0 0 4150 1 1 1 1 50 0 0 0 0 54 51 33 25 105 58 Subtotal Total

From the gures, it can be seen that the CO2 levels of those bedrooms utilizing AC were consistently higher than those utilizing NV. For those air-conditioners with no fresh air intake, the CO2 level could reach about 1600 ppm. The gures also show that for the window unit, when the air vent was opened, the CO2 level reduced from about 1200 to about 1000 ppm. In the case of portable air-conditioner, the CO2 levels hovered around 750800 ppm. For NV bedrooms, the CO2 levels for all bedrooms were about 550600 ppm.

3.1.2. Temperature and relative humidity (RH) proles Figs. 2ad show the temperature and RH proles of the bedrooms investigated. For NV bedrooms, the temperatures recorded ranged from 29 to 30 C and the RH 7080%. For AC bedrooms, the temperatures varied from 23 to 27 C even though all the air-conditioners were set at 22 C. Fig. 2b shows that when the air vent was opened, the initial temperature was higher than when the air vent was closed. However, after few hours, the temperature became similar. For the portable air-conditioner, Fig. 2c shows that

1118

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

(a)

(b)

(c)

(d)

Fig. 1. Comparison of CO2 proles of bedrooms using di erent air-conditioning systems and natural ventilation: a: Vertical casement air conditioner; b: Window unit (vent opened and closed); c: Portable air conditioner; d: Multi-split air conditioner.

the temperature was only marginally lower than the NV bedrooms. For RH, the uctuation was very high for most air-conditioners showing the constant dehumidication by the air-conditioners to remove the moisture produced by the occupants. 3.1.3. Particulate proles Figs. 3ad show the particulate levels of the bedrooms. The results show that the particulate levels were higher for NV bedrooms than AC bedrooms. For NV bedrooms, the levels varied from 40 to about 80 g= m3 . For AC bedrooms, the levels varied from 40 to about 70 g= m3 . Fig. 3c also shows that the particulate level for NV bedrooms was only marginally higher for AC bedrooms with fresh air intake. 4. Discussion of results 4.1. Objective measurements Table 3 shows the IAQ status of the NV bedrooms, which acts as a basis for comparison with the AC bed-

rooms (see Table 4). From Table 4, it could be observed that most type of air-conditioners except for the portable air-conditioner caused a buildup of the CO2 levels which were more than the recommended level of 1000 ppm. Although the levels of CO2 in these cases were not life-threatening or hazardous, headaches, nausea, dizziness and fatigue can occur. A sensation of lack of air may be noticed at concentrations as low as 600 800 ppm, particularly at low relative humidity. ASHRAE 62-1989 limits the average concentration in ambient air to 1000 ppm for a comfortable environment and acceptable air quality [14]. From Table 3, it can be seen that for NV bedrooms, the temperature and relative humidity were consistently higher than that of AC bedrooms. Furthermore, the wind speeds recorded were relatively low in most cases. As such, mechanical fans are usually utilized to improve the thermal conditions. As for the AC bedrooms, the temperatures were within the thermal comfort zone except for the bedroom using Portable air-conditioner where the average temperature reached about 26:5 C. This shows that the air-conditioner

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

1119

Fig. 2. Comparison of temperatures and RH of bedrooms using di erent air-conditioning systems and natural ventilation: a: Vertical casement air conditioner; b: Window unit (vent opened/vent closed); c: Portable air conditioner; d: Multi-split air conditioner.

was unable to achieve the required set point temperature especially in this case where fresh air provision was allowed. As for the RH, the multi-split air conditioner resulted in average RH of about 44%. Such low humidity may result in the exhibition of the SBS symptoms such as dry skin, dry throat and rashes. For particulate levels, those in NV bedrooms were shown to be higher than in AC bedrooms. However, for those AC with fresh air provision, the particulate levels were similar to NV bedrooms. This indicates that the main source of particulates was from outdoor and that the ltration systems of the air-conditioners may not be e ective. However, in both cases, the levels were far below the threshold level of 150 g= m3 . For bacteria and fungi, it is interesting to note that the bacteria and fungi counts in AC bedrooms were consistently higher than in NV bedrooms though the di erence is marginal. In both cases, the counts were well below the threshold level of 500 CFU= m3 . 4.2. Thermal comfort comparison In order to compare the thermal comfort in the bedrooms which are air-conditioned with the air vent closed and opened as well as the NV bedroom, Fangers equation was used for the computation of the PPD thermal comfort index [17]. The PPD was computed using the parameters obtained from the measurements which include air temperature, relative humidity and wind velocity. The MRT is assumed

to be the same as the air temperature. The metabolic rate used is 0:8 met (for sleeping) and the cloth value is 0.38. Fig. 4 shows the comparison of the thermal comfort for the bedrooms which are air conditioned with vent closed (PPD(ACVC)), vent opened (PPD(ACVO)) and the NV bedroom (PPDNV). It is interesting to note that the PPD computed for the bedrooms which are air-conditioned are extremely high, indicating that the bedrooms are overcooled. The air temperatures are found to be in the range of 19:8 C to about 22 C. By opening the air vent, the PPD decreases initially due to the increase in the air temperature. However, it reaches the same value as that with the air vent closed after about 3 h. For the NV bedroom, the PPD are substantially lower. The results show that perhaps by increasing the air movement with the use of fan, the naturally ventilated bedroom should be able to achieve the required thermal comfort requirement. 4.3. Subjective evaluations results 4.3.1. Comparison of SBS symptoms between occupants in NV and AC bedrooms Fig. 5 shows the comparison of SBS symptoms between occupants in NV and AC bedrooms. It can be discerned that NV homes had lesser occurrences of SBS symptoms than AC homes. In order to portray the statistics more accurately, the comparison was made in terms of percentage of the total sample size that was in NV and AC conditions, which was 105 and 58, respectively. It was interesting to note that there

1120

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

Fig. 3. Comparison of particulate levels of bedrooms using di erent air-conditioning systems and natural ventilation: a: Vertical casement air conditioner; b: Window unit (vent closed); c: Portable air conditioner; d: Multi-split air conditioner.

Table 3 IAQ status of naturally ventilated bedrooms Case Case Case Case Case 1 2A/B 3 4 CO2 (ppm) 640 669 566 595 Temp ( C) 28.9 29.8 28.5 28.4 RH (%) 77 73 75 69 Wind Speed (m/s) 0.05 0.05 0.05 0.04 Particulate ( g= m 3 ) 73 40 80 70 Bacteria (CFU= m3 ) N.A N.A 12 3 Fungi (CFU= m3 ) N.A N.A 27 28

Table 4 IAQ status of air-conditioned bedrooms Case Case 1 Case 2A Vent Closed Case 2B Vent Open Case 3 Case 4 CO2 (ppm) 1100 1146 1003 778 1164 Temp ( C) 24.8 20.5 21.5 26.5 25.1 RH (%) 58 53 53 60 44 Wind Speed (m/s) 0.02 0.02 0.02 0.02 0.02 Particulate ( g= m 3 ) 39 35 42 72 42 Bacteria (CFU= m3 ) N.A N.A N.A 36 44 Fungi (CFU= m3 ) N.A N.A N.A 39 43

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

1121

Comparison of thermal comfort (PPD) for AC Bedroom and NV Bedroom

PPD (%)

50 40 30 20 10 0 10:30 PM 11:00 PM 11:30 PM 12:00 AM 12:30 AM 1:00 AM 1:30 AM 2:00 AM 2:30 AM 3:00 AM 3:30 AM 4:00 AM 4:30 AM 5:00 AM 5:30 AM 6:00 AM 6:30 AM 6:47 AM

Time PPD(AC-VC) PPD(AC-VO) PPD-NV

Fig. 4. Comparison of the thermal comfort (PPD) between air-conditioned bedrooms and naturally ventilated bedroom.

Fig. 5. Comparison of the SBS symptoms in AC and NV bedrooms.

was a signicantly higher percentage of respondents in NV homes, who reported that they had no symptoms (48%) as compared to those in AC homes (18%). The gure also shows that the most common SBS symptoms su ered by the occupants in AC homes are Blocked

Nose, Running Nose, Dry Throat, Flu-like Symptom and Dry Skin. Fig. 6 shows the relationship between the number of days of usage of AC per week and the occurrence of SBS symptoms. It is intriguing to discover that for those who switched on their air-conditioners daily, they actually

1122

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123

100.00% 90.00% 80.00% 70.00% Percentage 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% Blocked Nose 0.00% Dry Eyes Watery Eyes Dry skin Running Nose Dry throat Rash No symptoms Chest tightness Headaches Lethargy Flu like

SBS Symptoms

Fig. 6. E ect of frequency of usage of AC on SBS symptoms.

had lesser occurrences of SBS symptoms than for those who switched on their air-conditioners irregularly (other than daily). For those who switched it on daily, they had the highest percentage of people who displayed no symptoms (27.78%). 5. Conclusion From the objective measurement of the 3 residential bedroom units utilizing both NV and di erent types of air-conditioning systems, it can be observed that: CO2 levels of bedrooms using AC are consistently higher than those utilizing NV. For those AC with fresh air provision, the CO2 levels are lower. RH and air temperature of NV bedrooms are higher than those of AC bedrooms. Thermal comfort comparison of the AC bedrooms and naturally ventilated bedrooms indicate that the AC bedrooms are usually substantially overcooled, resulting in extremely high PPD. For NV bedrooms, the utilization of fans should be su cient to achieve the required thermal comfort. For particulate levels, those in NV bedrooms are shown to be higher than in AC bedrooms. However, for those AC with fresh air provision, the particulate levels are similar or marginally lower than NV bedrooms. This indicates that the main source of particulates is from out-door and that the ltration systems of the air-conditioners may not be e ective. From the subjective evaluations, it can be seen that: The occupants utilizing AC tends to exhibit more SBS symptoms than those utilizing NV.

The frequency of usage of AC has important impact on the occurrence of SBS symptoms. Those who utilize AC daily tend to exhibit less symptoms indicating some form of acclimatization have taken place.

References
[1] WHO. Indoor air pollutants: exposure and health e ects. EURO reports and Studies, vol. 78. Copenhagen: WHO Regional O ce for Europe; 1983. [2] Bholah R, Fagoonee I, Subratty AH. Sick building syndrome in Mauritius: are symptoms associated with the o ce environment? Indoor+Built Environment 2000;9:4451. [3] Clements-Croome DJ, Baizhan L. Productivity and environment. In: Olli S, Jorma S, editors. Healthy Buildings 2000:Exposure,Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 62934. [4] Hagstrom K, Kosonen R, Heinonen J, Laine T. Economic value of high quality indoor air climate. In: Olli S, Jorma S, editors. Healthy buildings 2000: Exposure, Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 6716. [5] Hannula M, Niemela R, Rautio S, Reijula K. The e ect of indoor climate on productivity. In: Olli S, Jorma S, editors. Healthy Buildings 2000: Exposure,Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 65964. [6] Singh J. Health, comfort and productivity in the indoor environment. Indoor+Built Environment 1996;5(1):2233. [7] Wargocki P, Wyon DP, Fanger PO. Productivity is a ected by the air quality in o ces. In: Olli S, Jorma S, editors, Healthy Buildings 2000:Exposure, Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 63540. [8] Wargocki P, Wyon DP, Baik YK, et al. Perceived air quality, sick building syndrome symptoms and productivity in an o ce with two di erent pollution loads. Indoor Air 1999;9(3):16579. [9] Department of Statistics of Singapore. Ownership of household consumer durables Spore. Singapore: Department of Statistics.

N.H. Wong, B. Huang / Building and Environment 39 (2004) 1115 1123 [10] Jantunen, M. When and where are people exposed to pollutants. In: Olli S, Jorma S, editors. Healthy Buildings 2000: Exposure,Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 1522. [11] Hill BA, Craft BF, Burkat JA. Carbon dioxide, particulates, and subjective human responses in o ce buildings without histories of indoor air quality problems. Applied Occupational Environmental Hygiene 1992;72:10111. [12] Ministry of Environment. Indoor air quality guidelines for o ces. Singapore: Ministry of Environment. [13] Norback D, Bjornsson E, Janson C, et al. Asthmatic symptoms and volatile organic compounds, formaldehyde, and carbon dioxide in dwellings. Occupational and Environmental Medicine 1995;52(6):38895.

1123

[14] Sekhar SC, Tham KW, Cheong KW, Wong NH. A study of indoor pollutant standard index (IPSI) and building symptom index (BSI). In: Olli S, Jorma S, editors. Healthy Buildings 2000: Exposure,Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 14550. [15] Soon, A. What causes sick building syndromesick workroom or sick dwelling? In: Olli S, Jorma S editors. Healthy Buildings 2000:Exposure, Human Responses and Building Investigations Proceedings, vol. 1, 2000. p. 838. [16] Concordia University. Indoor air quality for o ces and classrooms. Canada: Concordia University; 2001. [17] ASHRAE Standard 55:thermal environmental conditions for human occupancy. America Society of Heating, Refrigerating and Air-conditioning Engineers; 1992.