Proposed residential indoor air quality guidelines for formaldehyde

3. Indoor Sources

Extensive reviews of formaldehyde emissions sources have been published by the World Health Organization (WHO 1989), and Environment Canada and Health Canada (2001). Sources that influence indoor levels of formaldehyde can be divided into two broad categories: combustion and off-gassing. Combustion sources include cigarettes and other tobacco products, and open fireplaces. Off-gassing sources include wood products such as particle board and other building materials made with adhesives containing formaldehyde, varnishes, paints, carpeting, drapes and curtains.

3.1 Combustion

Formaldehyde is released into the air by incomplete combustion of organic matter, especially wood. Formaldehyde emissions from residential stoves were assessed with birch wood and spruce wood, under normal and air-starved conditions. Under normal conditions, combustion of birch and spruce emitted 0.058 g and 0.041 g formaldehyde per kg wood, respectively. Air-starved conditions (i.e. with air supply almost completely shut) strongly increased formaldehyde production: birch and spruce combustion emitted 1.722 g and 0.255 g formaldehyde per kg wood, respectively (Ramdahl et al. 1982). In another study, aldehyde emissions from wood stoves were assessed with four types of wood: jack pine, cedar, red oak and green ash. Formaldehyde emissions ranged from 0.089 to 0.708 g per kg wood, and accounted for 8% to 42% of total aldehyde emissions (Lipari et al. 1984).

When a residential wood stove and a residential charcoal-fueled heater were tested under similar controlled conditions, charcoal combustion produced less formaldehyde than wood combustion. Under normal conditions (i.e. without air starvation), it emitted 0.0012 g formaldehyde per kg charcoal (Ramdahl et al. 1982).

Health Canada's Tobacco Control Programme (unpublished data) determined total formaldehyde emitted in mainstream smoke (smoke inhaled and exhaled by the smoker) and in sidestream smoke (released directly by the burning end of a cigarette) from cigarette brands marketed in Canada. Under standard testing conditions, the formaldehyde content of mainstream smoke of 20 cigarette brands tested ranged from 11 to 128 μ g per cigarette with a mean of 53 μ g per cigarette, and that of sidestream smoke of 5 brands tested ranged from 327 to 440 μ g per cigarette, with a mean of 367 μ g per cigarette (Table 2).

Table 2. Formaldehyde content in smoke from cigarette brands marketed in Canada

Smoke	Mainstream		Sidestream
Test conditions	ISO	intense	ISO	intense
Number of brands tested	20	20	5	5
Formaldehyde (μg/cigarette)
Minimum	10.7	46.8	327	275
Maximum	128	248.3	440	334
Mean	53.4	139.7	367	302
Standard deviation	32.8	47.8	44	22
Smoke generated by smoke machines. ISO conditions: 35 ml/puff, 2-second puffs each 60 seconds, ventilation holes unobstructed. Intense conditions: 55 ml/puffs, 2-second puffs each 30 seconds, ventilation holes obstructed. Source: Final Report: Cigarette Tobacco and Cigarette Smoke, Toxic Emission Information: Assessment, Characterization and Verification. 2002. Health Canada Contract Number H4097-015017/001/SS, work performed by Labstat International Inc. for Health Canada.

3.2 Gaseous Emissions

3.2.1 Wood products

Formaldehyde is released from pressed wood products made with urea-formaldehyde resins (e.g. particle board, hardwood plywood, medium-density fibreboard), and at lower levels from wood products with phenol-formaldehyde resins (e.g. softwood plywood, oriented strand board). Concerns about potential health impacts from these emissions led the wood products industry to adopt voluntary standards on formaldehyde emissions from particle board (ANSI 208.1) and medium-density fibre (MDF) board (ANSI 208.2) in the 1990s (Composite Panel Association 1999; 2002).

Kelly et al. (1999) assessed formaldehyde emissions from several wood products in a chamber over 24 hours. Emissions from coated urea-formaldehyde wood products (e.g. melamine, laminate) ranged from <2.7 to 55 μg/m² h with the exception of one product emitting 460 μg/m² h, and emissions from bare phenol-formalde-hyde wood products ranged from 4.1 to 9.2 μg/m² h. Among bare urea-formaldehyde wood products, emissions from plywood products ranged from 8.6 to 103 μg/m² h, emissions from particle board products ranged from 104 to 1,580 μg/m² h, and emissions from MDF products ranged from 210 to 364 μg/m² h.

Brown (1999) assessed formaldehyde emissions from particle board panels and MDF panels in different small chambers and room chambers for several months, starting 7 days after manufacturing. Emissions factors from all the products tested were approximately 300 to 400 μg/m² h in the first few weeks and 80 to 240 μg/m² h after 6 to 10 months.

3.2.2 Varnishes, floor finishes, and paints

Varnishes are also known to emit formaldehyde. Three conversion varnishes tested by the U.S. Environmental Protection Agency still emitted detectable levels of formaldehyde more than 720 hours (one month) after application; and one of the three varnishes emitted 170 μg/m² h formaldehyde 2,762 hours (about 115 days) after application (Howard et al. 1998). Formaldehyde was still emitted 3,300 hours (about 138 days) after varnish application and the cumulative formaldehyde emission to then was about 700% to 800% of the free formaldehyde amount present in the varnish at the time of application, indicating that formadehyde was formed during the curing process (McCrillis et al. 1999).

Two commercially applied floor finishes were tested by Kelly et al. (1999). In typical conditions, a base coat emitted 1,050,000 μg/m² h formaldehyde immediately after application, and 10,800 μg/m² h 24 hours later; a top coat emitted 421,000 μg/m² h immediately after application and 4,660 μg/m² h 24 hours later.

Water-based paints also emit formaldehyde. In a chamber study by Chang et al. (1999) of four interior water-based paints (water content 40.7%-55.4%) advertised as "low-VOC," two of the paints tested emitted significant amounts of formaldehyde after application; formaldehyde emissions from one paint were still detectable 50 hours after application. Actual emission rates were not shown in the paper. Additional studies were conducted with the paint that had the highest formaldehyde emission (Chang et al. 2002). It was shown that formaldehyde emissions can be characterized by three stages: an initial "puff" of instant decay, a fast decay phase, and a slow decay phase lasting more than 300 hours post-application (emission levels and duration not specified in the paper). Elimination and replacement of the biocide (not specified) from the paint resulted in a 55% decrease in formaldehyde emissions.

3.2.3 Other consumer products

Some carpets emit formaldehyde into the air. The Canadian Carpet Institute (CCI) has established a voluntary emission standard of 50 μg/m² h. In a chamber experiment, the time-course of VOC emissions from four different carpets wasdetermined, but only one was found to release aldehydes: formaldehyde emission rates were 57.2 μg/m² h after 24 hours and 18.2 μg/m² h after 168 hours (Hodgson et al. 1993).

Some textile fabric finishes such as dimethylol-dihydroxyethyleneurea (DMDHEU), melamine resin, and wax water repellent have been found to emit formaldehyde; emissions were decreased but not eliminated by curing (Martin et al. 1998). In another study, formaldehyde emissions from cottons treated by DMDHEU-based finishes were measured in a dynamic chamber; emissions reached a peak after about 2 hours, and decreased to a non-detectable level within 4 days (Kottes Andrews and Trask-Morrell 1997).

3.3 Secondary Production from Ozone

Formaldehyde may also be formed by the chemical reaction of ozone with some building and surface materials. A chamber study showed that the presence of ozone increased the release of formaldehyde from plaster, plywood and fitted carpet (Moriske et al. 1998). Formaldehyde is also formed through the oxidation of R-(+)-limonene, a VOC that is common in indoor environments, by ozone (Clausen et al. 2001). Indoor ozone-releasing devices such as photocopiers and laser printers have been found to release formaldehyde, and this is thought to result from the reaction of ozone with aliphatic hydrocarbons. When a single dry-process photocopier was sent to four different laboratories for chamber experiments, formaldehyde emissions rates ranging from 1.3 to 4.7 g/h of operation were measured (Leovic et al. 1998). Emission from laser printers were also assessed, and were found to range from non-detectable to 0.3 g/h of operation (Tuomi et al. 2000).