State of knowledge report
Environment Australia, 2001
ISBN 0 6425 4739 4
Part B: Indoor air quality (continued)
Indoor air quality is influenced strongly by specific indoor sources, including gas appliances, furnishings (eg reconstituted wood-based panels), drycleaning fluids, open fires, cigarette smoke, carpet dust and products such as aerosols and solvents used for cleaning. Very little is known about indoor VOC concentrations in Australia. The NHMRC1 has recommended an indoor air quality goal for total VOCs for Australian buildings (see Table 6.3). However, individual VOC levels have received very limited investigation to date (Brown 1996a).
CSIRO Building Construction and Engineering have assisted in investigations of buildings from several States regarding adverse health effects associated with poor indoor air quality. Brown (1998c) provides a brief review of cases where indoor air quality problems occur in Australia. Occupants were asked to complete a questionnaire, with particular attention to symptom frequency and association with occupation of the building. Buildings were inspected to gather information about building faults, air ventilation, cleaning regimes and known pollutant sources. All pollutant sampling was of short-term duration (approximately 1 hour) and several sampling methods were used.
|Building type||Age (years)||Problems reported||Indoor air pollutant concentration (gases and particles µg/m³, microbial CFU/m³, mites/m² carpet)||Probable cause|
|Office||50||Odour, ill-effects||TVOCs 450–920, toluene 35, tetrachloroethylene 80–10 000, CO2 450 ppm||Shared building with drycleaners|
|Laboratory||20 (recently renovated)||Ill-effects||TVOCs 400–3200 (hexane, acetone, chloroform, toluene, texanol), formaldehyde 120–150, fungi 85, bacteria 150||Formaldehyde fumigation of laboratory, paints and coatings|
TVOCs = total VOCs;
CFU = colony-forming unit; ppm = parts per million
Source: Brown (1998c).
Air ingress and distribution were important factors that caused indoor air pollution in areas of buildings remote from pollutant sources. The case studies showed that pollution of indoor air environments could arise from a wide variety of causes, including VOCs in new or renovated buildings, formaldehyde from particleboard flooring and other products over periods of years, odourants and other pollutants from furniture items, and bacteria from materials exposed to standing water. Two case studies showed that some practices (eg fumigation of laboratories with formaldehyde and the location of semi-industrial processes in office buildings) need greater care to prevent long-term indoor air pollution (Table 10.1).
NSW Health conducted a survey of indoor air quality in 140 homes across the state in the winter of 1999. Qualitative measures of PM10, nitrogen dioxide, formaldehyde, nicotine, dust mite aeroallergen and ventilation rates were made. NSW Health will publish this study in late 2000.
Several studies have been conducted in Australia to investigate how the emissions of formaldehyde and VOCs affect indoor air quality. Brown (1998c) studied houses and office buildings in Melbourne, Victoria, to quantify emissions and assess their impact. His first study centred on VOCs in homes and office buildings (Brown 1996a). The houses used in the study were typical of Melbourne suburban houses but also included one house on the rural east coast of New South Wales. Some houses had occupants who had complained of mucosal irritation, headaches, nausea or bad odours, and are referred to as ‘complaint houses’. The houses used in this part of the study were all more than 12 months old and none had been renovated recently. In addition, detailed measurements over prolonged periods were made in two houses: one had a room addition completed a few days previously; the other had one room cleaned with typical products (polish, disinfectant, deodoriser). The office building investigated had two storeys; the upper level had been renovated recently and several employees had reported headaches and nausea. Table 10.2 reports the VOC concentrations found in the houses investigated.
|VOC||Concentration (mg/m³) in air sampled from:|
|Typical houses (20)a||Complaint houses (10)a||Outdoors (14)a|
AM = arithmetic mean;
GM = geometric mean;
LQ–UQ = lower quartile to upper quartile range;
VOC = volatile organic compound;
TVOC = total VOCs
Source: Brown (1996a).
Although the sample size was limited, the results from Table 10.2 indicate that the indoor total VOC concentrations were well below the recommended level set by the NHMRC1 (500 µg/m³ over one hour) (NHMRC 1992). Furthermore, the concentrations were similar in both typical and complaint houses, and were significantly higher than those taken outdoors. Some VOC concentrations were found to be quite high in typical houses; the reasons for this are unknown (Brown 1996a). For example, benzene concentrations ranged from 15 to 81 µg/m³ in four of the houses, whereas adjacent outdoor levels ranged up to 5 µg/m³. At another of the houses, elevated VOCs such as freon-11, hexane, benzene, toluene and total VOCs were recorded. The detection of elevated levels of these VOCs may be explained by the former use of the adjacent site as a bus depot.
Where concentrations of VOCs were measured directly after intensive cleaning, deodorising and vacuum cleaning of carpet, and the use of wood polish, cleaners and disinfectants, it was found that VOCs were elevated for a short period of time. In comparison, renovations on a house and on an office resulted in more significant elevations of VOCs over a longer period of time. VOC concentrations were shown to be elevated above the NHMRC recommendations (500 mg/m³ over one hour) for several weeks after the renovations were completed and were shown to be elevated in other parts of the building away from the renovation. Table 10.3 reports the results of VOCs measured in a house and an office after renovations were completed.
|Building renovation||Period after renovation (days)||TVOC concentration (µg/m³)||Major compounds|
|Renovation area||Other areas|
|Extension to house||5 9 18 33||1190 590 300 180||920 410 170 130||n-Butylacetate, xylene, texanol, trimethylbenzenes, toluene, 4PC|
|Ground floor office renovation||During 35 80||10 000 540 180||2300 –10 600 – –||Branched alkanes, n-alkanes, trimethylbenzenes, xylene, toluene|
Note: TVOC = total volatile organic compounds
Source: Brown (1996a).
A study performed by Monash University looked at the effect of and formaldehyde emissions in 80 homes across the Latrobe Valley, Victoria. levels were monitored using passive samplers and analysed using spectrometry (Garrett et al 1999). At least half of the houses had at least one asthmatic child (a total of 53), while the other half had none (95), making a total of 148 children included in the study. On six occasions, passive samplers were exposed in the bedroom, living room, kitchen and outdoors to measure NO2 concentrations in each household over one year. On two of these sampling visits, each child was interviewed and required to complete a questionnaire. Frequency of symptoms during the previous year and parental allergy and asthma incidents were recorded; in addition, skin prick tests were performed on each child.
Note: a Number of source units.
Source: Garrett et al (1999).
Overall mean indoor NO2 levels were higher than outdoor levels, and levels in winter were higher than in other seasons. The NO2 levels recorded ranged from < 0.7 to 246 µg/m³, with a mean of 11.6 µg/m³. The primary contributor was gas stoves, with vented gas heaters and tobacco smoking also being major sources of indoor NO2. Multiple regression showed that 67% of the variation in indoor NO2 levels could be explained by the presence of major sources, house age and outdoor levels. However, different patterns of appliance use in the various houses resulted in considerable variation in NO2 levels between the houses. Levels of NO2 recorded in different rooms by major sources in the house are shown in Table 10.4.
The health outcomes of children (total 148 children 7–14 years of age, 53 of whom were asthmatic) resident in the homes were also studied using a respiratory questioner, skin prick tests, and peak flow measurements (Garrett et al 1998). Respiratory symptoms were more common in children exposed to a gas stove (odds ratio 2.3 [95% CI 1.0–5.2], adjusted for parental allergy, parental asthma, and sex). However, NO2 exposure was a marginal risk factor for respiratory symptoms. That is, gas stove exposure was a significant risk factor irrespective of NO2 levels (odds ratio 2.2 [95% CI 1.0–4.8]).
As part of the larger study, Garrett et al (1996) also looked at formaldehyde levels in 80 households on four occasions between March 1994 and February 1995. Home surveys were conducted to identify potential formaldehyde sources, and this was linked to health data (described above). The presence of particleboard, fibreboard, plywood and new furnishings was recorded and potential combustion sources such as gas stoves, heaters and smoking were also noted. Formaldehyde levels had a median of 15.8 µg/m³ and ranged from below detection limits to 139 µg/m³ (Garrett et al 1996). This was significantly higher than the outdoor levels. Summer and early autumn had higher indoor formaldehyde levels than autumn and early spring. Higher formaldehyde levels were associated with the presence of fibreboard in the sampled room and high temperature at the start of the sampling. Open windows at the start of the sampling period were associated with lower formaldehyde levels. Table 10.5 includes the results of the formaldehyde concentrations measured indoors over the study period.
|Month||Season||Median||25th percentile||75th percentile|
Source: Garrett et al (1996).
While exposure to NO2 emissions was found to increase the risk of respiratory symptoms, formaldehyde exposure was found to increase the risk of asthma and allergy in children. There was an association between high formaldehyde levels and homes of asthmatic children: a household was 2.8 times more likely to have an asthmatic child if the formaldehyde levels ever exceeded 40 ppb (1.49–5.20 [95% CI 1.0–5.2], adjusted for parental allergy, parental asthma, ventilation practices and NO2 levels).
Formaldehyde emissions correlated with the presence of fibreboard in the sampled room and high temperatures at the start of sampling (Garrett et al 1996). When windows in the sampled room were open, formaldehyde concentrations were lower than in rooms with closed windows. Furthermore, an increase in indoor humidity and the presence of particleboard flooring tended to correlate with higher formaldehyde levels (Garrett et al 1996).
In households where formaldehyde emissions exceeded 40 ppb, there was an increase in incidence of asthma in children. Similarly, diagnosed asthma and the number of positive skin prick tests correlated with formaldehyde levels. As formaldehyde concentrations in the home increased, respiratory and total symptom scores also increased, implying a dose-response relationship (Garrett et al 1996). Table 10.6 shows the results of the formaldehyde exposure and health effects in children. The results indicated that low-level exposure to indoor formaldehyde might increase the risk of allergic sensation to common aeroallergens in children. That is, exposure to the formaldehyde levels frequently encountered in homes may increase the risk of allergy and asthma in children.
|<16 ppb||16-40 ppb||>40 ppb|
|Number of houses||31||76||40|
|Proportion of children asthmatic||16%||39%||43%|
|Total symptom scorea||6.42 (3.91–8.93)||9.74 (7.82–11.65)||11.6b (8.73–14.48)|
|Respiratory symptom scorea||1.1 (0.44–1.76)||2.22 (1.70–2.75)||2.56b (1.68–3.47)|
|Number of positive skin prick testsa||1.29 (0.37–2.21)||3.47b (2.62–4.33)||3.83b (2.75–4.91)|
|Maximum relative wheal sizea||0.5 (0.21–1.79)||1.05b (0.83–1.26)||1.28b (0.99–1.58)|
aFigures in brackets are 95% confidence intervals
bSignificantly different (p<0.05) from lowest category using Bonferroni's multiple comparisons test.
Source: Garrett et al (1996).
A pilot study in Hobart, Tasmania examined VOC exposure in nonindustrial environments using 11 known ‘sick buildings’, identified in an earlier study, and a control (Mesaros 1998). The buildings ranged from modern city highrise to one-storey buildings. VOCs were monitored using activated charcoal tubes suspended 1.5 m from the floor and exposed continuously over seven days, in office buildings under normal working conditions. To complement the study, 256 office workers were surveyed using a questionnaire based on a combination of the Danish Research Building Institute questionnaire and WHO building investigation procedures. The questionnaire addressed such building issues as floors, walls, ceilings, heating, lighting, room density, furniture, smoking, cleaning regimes, airconditioning and other environmental determinants. The study also compared indoor VOC emissions in winter and summer. A quantitative and qualitative analysis was undertaken in each building every week for five weeks in February–March and five weeks in July–August.
A total of 12–15 VOCs were detected in all the sampled buildings (Table 10.7), and are consistent with those types found by other studies (eg Brown 1996a). The weekly total VOC levels ranged from 0.0001 to 1.934 mg/m³ in the 11 test buildings. All buildings with high total VOC levels had been refurbished within the previous five years. Decane, nonane and octane were found to have originated from the floor coverings in the buildings. VOC concentrations were elevated in winter, and overall hydrocarbon levels were twice as high in the winter as in the summer. However, elevated concentrations of ethanol were evident only in the summer months.
|Chemical name||Range µg/m³||Chemical name||Range µg/m³|
|m-, p- and o-Xylene||4–48||Decane||7–137|
|1 2 4 Trimethylbenzene||3–294|
Source: Mesaros (1998).
The Hobart study emphasised the complexity of effective indoor air quality analysis. There seemed to be no direct relationship between seasons, sick building syndrome symptoms, ventilation, and VOC concentrations. Elevated ethanol concentrations in the summer months may have been due to solvents and fuel; high VOC levels in winter months may have been due to heating of the buildings. The degree of ventilation had no direct correlation with the symptoms experienced by the occupants of the buildings.
The use of total VOC as an effective risk index or guide to the relationship between occupant illness in nonindustrial building is questionable. In the Hobart study, preliminary analysis has not shown clear correlations between total VOC levels and reported sick building syndrome. Individual concentrations of VOCs were well below Australian exposure standards for atmospheric contaminants in the occupational environment, although symptoms among building occupants were high. A large proportion (60–100%) of employees reported at least one sick building syndrome symptom. Therefore, the level of total VOC was not considered to be a reliable indicator of occupant ill health.
As part of an international study on personal exposure to tobacco smoke funded by the Center for Indoor Air Research in the United States, fixed site monitoring for four VOCs both inside and outside buildings was undertaken in Sydney (Phillips et al 1998). The study involved sampling indoor air through adsorption tubes for a 24-hour period and later analysing it by mass spectrometry. The houses were classified as smoking if a resident smoked cigarettes, pipes or cigars in the communal areas of the house.
Table 10.8 presents the VOC results of the fixed site monitoring. Median toluene levels were the highest of all VOCs measured. This was consistent irrespective of the city studied. Similar levels of 1,3 butadiene were found both indoors and outdoors. Interestingly, there was no significant difference between the median VOC concentrations monitored in ‘smoking’ and ‘nonsmoking’ households, probably due to the contribution of non-environmental tobacco smoke sources, including from road traffic emissions.
|Compound (µg/m³)||Indoors smokinga||Indoors nonsmokinga||Outdoorsa|
|Benzene||2.3 (4.2)||3.4 (6.5)||3.7 (7.8)|
|Toluene||20 (65)||19 (36)||17 (30)|
|1,3-Butadiene||0.14 (0.72)||0.12 (0.54)||0.13 (0.32)|
|Isoprene||0.19 (0.46)||0.16 (0.38)||0.12 (0.25)|
Note: a Brackets refer to the 90th percentile.
Source: Phillips et al (1998).
Nitschke et al (1998) analysed indoor nitrogen dioxide levels by selecting a cohort of 129 asthmatics and ensuring they wore diffusion badge monitors while indoor gas appliances are being used, which is seen as potentially the greatest period of exposure. Levels were expressed at time-average nitrogen dioxide concentration, because daily badge exposure times differed (mean 4.5 hours, sd 2.4). Comparisons were also made between the types of household appliances used. Personal time averaged exposure ranged from below detection limits to 1760 ppb. The presence of gas appliances doubled the exposure of nitrogen dioxide when compared to households with electric appliances. Mean indoor exposure levels were higher than the average level outdoors.
The percentage of measurements that exceeded the WHO outdoor guideline was 3.5%, recorded in 33 households, of which 30 had a gas appliance of some description (Figure 10.1). Households with unflued gas heaters had significantly high mean nitrogen dioxide levels compared to those with flued heaters. Important, but not significant, differences were found between households with increased cooking times (> 1 hour) resulting in an increase in exposure to nitrogen dioxide. Also active ventilation (eg extractor fans) in the 33 households with gas cooking decreased the mean nitrogen dioxide values.
Source: Nitschke et al (1998).
Brown (1998a) reviewed emissions from office furniture contributing to concentrations of formaldehyde and VOC in indoor air. High concentrations of formaldehyde have been shown to occur in indoor air, largely from reconstituted wood-based panels such as particleboard and medium density fibreboard (Brown 1998a). However, very little is known about VOC emissions from Australian reconstituted wood-based panels. The NHMRC1 has recommended an indoor air quality goal for total VOCs for Australian buildings, but individual VOC levels have received very limited investigation to date (Brown 1996a).
Pollutant emissions from reconstituted wood-based panels were assessed concurrently in small room chambers for up to 20 days. Table 10.9 presents the results of formaldehyde concentrations in chamber emissions. The results of the study showed that the NHMRC indoor air quality goal for formaldehyde (130 mg/m³) was exceeded in chamber air, and it was predicted that the goal would be exceeded for months to years. Additionally, the formaldehyde measurements for particleboard show varying emissions within each chamber; they were much more stable for medium-density fibreboard.
The effect of the ventilation rate on formaldehyde emission from particleboard was also measured; the emission factor was found to increase as the ventilation rate increased. The relationship between emission factor and ratio to ventilation rate was approximately linear (Brown 1998a).
|Panel||Time (hours)||Small chamber 1||Small chamber 2||Small chamber 3||Room chamber|
|Emission model parametersa||M0(?g/m²)||920,000||590,000||440,000||7.7x108|
|Emission model parametersa||M0(?g/m²)||440,000||190,000||410,000||380,000|
a factor (mg/m².h) = k1.Mo.exp(-k1.t).
RH = relative humidity;
ACH = air changes per hour
Source: Brown (1998a).
Identifiable VOC emissions from medium-density fibreboard specimens were at very small quantities but consisted of butanal, furfural, limonene and a-terpineol. In comparison, particleboard emitted a wide range of VOCs, as presented in Table 10.10. The range of VOCs emitted were similar to those found by Koontz and Hoag (1995): the major VOCs emitted from unfinished particleboard and medium-density fibreboard from North America were (in approximate order of quantity) acetone, hexanal, pentanal, benzaldehyde, pentanol, heptanal, pinenes, nonanal and octanol.
|Chamber||Time (days)||Concentration (µg/m³) of VOCs|
|Small chamber 1||1||540||120||21||52||6.2||2.6||130|
|Small chamber 2||1||620||120||17||76||10||3.2||160|
|Small chamber 3||1||520||81||17||38||4.4||2.2||97|
VOCs = volatile organic compounds;
TVOCs = total VOCs
Source: Brown (1998a).
All VOCs except hexanal were found to decay more rapidly than formaldehyde emissions (Brown 1998a). Furthermore, VOC emissions from small samples across the sheet were similar to each other but were significantly lower than emissions in a room chamber using full size sheets.
Table 10.11 presents the results for the major VOCs emitted from office furniture. Formaldehyde concentrations and emission factors from office furniture were in excess of the NHMRC1 goal of 120 µg/m³. The concentration and emission factor for formaldehyde were very similar despite most of the surfaces being laminated. It was also found that a large number of VOCs were emitted from the furniture, which was largely composed of medium-density fibreboard.
The major VOCs found to be emitted from furniture were methanol and acetone, which were also the major VOCs emitted from particleboard. Many other VOCs were emitted from furniture so the total VOC emission, relative to formaldehyde emission, was significantly higher for the furniture in comparison with unfinished reconstituted wood-based panel products.
|VOC||Concentration (µg/m³)||EF (µg/m².h)|
|4 hours||1 day||4 hours||1 day|
|Methyl isobutyl ketone||200||210||250||260|
|m- and p-Xylene||59||48||74||60|
Source: Brown (1998a).
Indoor air quality is influenced by two major components: the amount and quality of outdoor air getting in, and indoor sources of pollutant emissions. There are a number of sources of indoor air pollutants, including building operations and construction materials, household products, external factors and various human indoor activities. Specific examples of these indoor air pollutants include petrol and car exhaust entering building air from attached garages, timber stains, paints and coatings, carpets and furniture, cleaners, disinfectants and detergents, and environmental tobacco smoke.
The optimum strategy for the control and management of indoor air is pollutant source control, that is, the selection and use of low-emission products in the construction and operation of buildings. Ventilation is also recognised as a central factor for the management of the great majority of identified indoor air pollutants. This is especially the case for mechanically ventilated buildings constructed in the 1980s and residential buildings constructed in recent years. Ventilation acts by diluting the concentration of the contaminant and, as a part of the process of dilution, removes some of the contaminant to outdoor spaces. The 1990 Building Code of Australia requires all buildings to have adequate ventilation and air quality by providing permanent openable windows or mechanical ventilation to Australian Standard levels. A balance needs to be established in building ventilation rates between addressing indoor air quality issues and reducing energy consumption and greenhouse emissions.
Many other substances that are found indoors may act as allergens and have a significant impact on health, especially for individuals who are predisposed to hypersensitivity reactions such as hayfever and asthma. Indoor allergens with the potential to have an adverse effect on health include the house dust mite, animal dander, cockroaches and fungal spores.
We have only limited knowledge about indoor air quality and the effects of household pollutants on human health, but a number of studies have been conducted to gain a better understanding of the some of the issues surrounding air quality in the home and the office. Studies have centred on examining some of the compounds present and the concentrations emitted from construction materials and furniture, and surveys of residents and office workers to correlate health effects.
Results of these studies indicate that indoor total VOC concentrations in established buildings are well below the recommended level set by the NHMRC1. In one study where indoor and outdoor VOC levels were compared, the indoor concentrations were significantly higher than outdoor levels. Cleaners and disinfectants were found to elevate total VOC concentrations for short-periods of time; renovations resulted in more significant elevations over a longer-period of time. Increased nitrogen dioxide and formaldehyde levels were associated primarily with gas appliances and the presence of reconstituted wood-based panels, respectively.
The major VOCs emitted from furniture and particleboard were methanol and acetone; the total VOC concentration from furniture was higher than that from reconstituted wood-based products. In a study conducted on emissions from reconstituted wood-based products, it was found that formaldehyde emissions exceeded the NHMRC indoor air quality goal for long periods after installation. Additionally, formaldehyde concentrations and emission factors from office furniture were found to be in excess of the NHMRC goal. Furthermore, formaldehyde emissions in homes were found to correlate with asthma incidence in children.
Overall, studies on indoor air quality emphasise the complexity of factors that influence the link to human health effects. Changes in the activities that occur both within and outside the buildings, cleaning, heating and renovation regimes, the degree of ventilation and the presence of products that emit VOC all influence the concentrations of air pollutants indoors. Further studies of indoor air quality and in-vehicle environments are required to achieve a greater understanding of individuals' total exposure to air toxics in Australia.
1 On 19 March 2002, the National Health and Medical Research Council rescinded its publication "Ambient Air Quality Goals and Interim National Indoor Air Quality Goals". The Council has made this publication available on its Internet Archives site as a service to the public for historical and research purposes only.
The publication is available at: http://www.nhmrc.gov.au/publications/synopses/eh23.htm
The Internet Archives site also contains the following statement made by the National Health and Medical Research Council:
Rescinded publications are publications that no longer represent the Council's position on the matters contained therein. This means that the Council no longer endorses, supports or approves these rescinded publications.
Links to another web site
Opens a pop-up window