Atmosphere

Review of literature on residential firewood use, wood-smoke and air toxics

Technical Report No. 4
Environment Australia, June 2002
ISBN 0 6425 4868 4

Discussion and conclusions

4.1 Review outcomes

4.1.1 Emissions of particles and air toxics

The Australian literature demonstrates the growth in popularity of woodheaters commencing in 1978 and reaching a peak in about 1992 when 25.4% of Australian households used firewood for heating (18.5% as the main heating fuel and 6.9% as a secondary heating fuel). Some states had a much higher proportion of firewood using households, such as Tasmania 70% and Western Australia 40%. The proportion of households using firewood has been dropping slowly since 1992, with 15.5% of Australian households using firewood as their main heating fuel in 2000.

Estimates of total firewood used for residential space heating in Australia in 1999/2000 range from 3 to 5.5 million tonnes. This uncertainty of almost a factor of two results from different surveys of households and whether or not the quantity of firewood purchased is a true measure (one study suggests households receive 23% less wood by weight than they pay for).

Community and government concern about urban wood-smoke has been increasing since the mid-1980s. International studies linking mortality and morbidity to fine particle concentrations significantly increased this concern since emission inventories for Australian cities showed residential wood heating was a significant source of PM10 and PM2.5 particles. Action to mitigate wood-smoke emissions has focussed on community education campaigns to encourage responsible woodheater use and requirements that new heaters should meet the emission limit specified in the Australian Standard AS4013. There is little evidence on the success or otherwise of community education campaigns. Enforcement of AS4013 compliance has been very limited (i.e. independent checking of appliance emissions has not occurred).

The significance of air pollution emissions from residential firewood use in Australia has focussed almost entirely on fine particles. Over 300 models of woodheater have been tested for compliance with AS4013. The average emission factor for all models tested is 3.3 g/kg. This only represents actual emissions if it is assumed appliances are correctly operated and air-dry firewood is burnt. However, it is generally acknowledged that 'real-world' emission factors will be higher, possibly by up to a factor of 3, because an unknown proportion of households operate their heaters poorly and use wet firewood.

Australian emission inventories have used estimates of average emission factors because no measurements of emission factors for appliances operating in people's homes have been made in Australia. Extrapolation of laboratory measurements of emission factors to field use may introduce significant errors, although these extrapolated 'real-world' emission factor estimates are similar to measured 'real-world' emission factors in the United States. Table 4.1 summarises particulate emission factors for open fireplaces and woodheaters.

Table 4.1: Particulate emissions, all values shown as emission factors (g/kg)
  Woodheater Range Open fireplace Range
AS4013 tested heaters (laboratory) 3.3 0.8 to 5.5    
Australian non-certified woodheaters 11 7 to 15    
Estimated 'real-world' WH for Australia 12.5 9 to 13.6    
DeAngelis et al. 1980 9.1 1 to 28 13 2.4 to 26
Cooper 1980 8.5 1 to 24 9.1 7.2 to 12
AP42 conventional (real-world) 15.3   17.3  
AP42 Certified (real-world) 9.8 7.3 to 12.9    
AP42 Catalytic (real-world) 10.2 8.1 to 12.1    

Measurements of 'old' and 'new' carbon in particles in cities where wood-smoke is thought to be significant have confirmed the importance of biomass sources as well as the presence of fossil fuel sources. There is little doubt that residential firewood use is a significant source of fine particles in many Australian cities and towns, although the precise contribution is subject to uncertainty.

Emission factors for other priority air pollutants have not been determined for woodheaters or open fireplaces in Australia. The international literature provides figures that are often used in Australian emission inventories. Table 4.2 shows emission factors for carbon monoxide, oxides of nitrogen and oxides of sulfur.

Table 4.2: Gaseous emissions, all values shown as emission factors (g/kg)
  CO CO Range NOx NOx Range SOx SOx Range
DeAngelis et al. 1980 WH 180 91 to 370 0.49 0.2 to 0.8 0.2 0.16 to 0.24
DeAngelis et al. 1980 OFP 67 15 to 140 2 0.84 to 4.3    
Cooper 1980 WH 160 83 to 370 0.5   0.2  
Cooper 1980 OFP 22 11 to 40 1.8      
AP42 OFP 126.3   1.3   0.2  
AP42 conventional WH 115.4   1.4   0.2  
AP42 Phase II WH 70.4   -   0.2  
AP42 catalytic WH 52.2   1   0.2  

Notes:
WH = woodheater
OFP = open fireplace
AP42 = US EPA (1996a) emission factors

Emission factors for air toxic species have not previously been determined for woodheaters or open fireplaces used in Australia. The emissions research, of which this literature review is a part, will contribute much new information on this issue.

International measurements of emission factors for air toxics are summarised in Table 4.3. The Table includes reference to the limited number of Australian studies that have identified, but not quantified, emissions of air toxics in wood-smoke.

Table 4.3: Emission factors for various compounds for woodheaters selected from US EPA (1996a)
Compound g/kg Aust. Khalili et al. *
Ethane 0.735    
Ethylene 2.245    
Acetylene 0.562    
Coronene   b,c  
Propane 0.179    
Propene 0.622    
i-Butane 0.014    
n-Butane 0.028    
Butenes 0.596    
Pentenes 0.308    
Benzene 0.969 A  
Toluene 0.365    
Furan 0.171    
Methyl Ethyl Ketone 0.145    
2-Methyl Furan 0.328    
2,5-Dimethyl Furan 0.081    
Furfural 0.243    
o-Xylene 0.101    
PAH total <0.250    
Acenaphthene 0.005   0.0001
Acenaphthylene 0.016   0.026
Anthracene 0.005 a,c 0.005
Benzo(a)anthracene <0.0005    
Benzo[a]pyrene 0.003 b,c 0.0029
Benzo(e)pyrene 0.001   0.0028
Benzo[b]fluoranthene 0.004 b 0.0003
Benzo[k]fluoranthene <0.001 b 0.0006
Benzo(g,h,i)fluoranthene 0.014    
Benzo[g,h,i]perylene 0.028 b  
Benz[a]anthracene <0.001 b,c 0.0003
Dibenzo[a,h]anthracene 0.004    
Biphenyl 0.011 a  
Chrysene 0.005 b,c 0.0005
7,12-Dimethylbenz(a)anthracene 0.002    
Fluoranthene 0.004 a,b,c 0.001
Fluorene 0.007   0.002
Indeno[1,2,3,cd]pyrene 0.010 b  
12-Methylbenz(a)anthracene 0.001    
9-Methylchlolanthrene 0.002    
3-Methylchlolanthrene <0.0005    
1-Methylphenanthrene 0.015    
Naphthalene 0.072 A 0.006
Nitronaphthalene BDL    
Perylene 0.001 c  
Phenanthrene 0.059 a,c 0.003
Phenanthrol BDL    
Phenol <0.001    
Pyrene 0.004 a,b,c 0.001

Notes:
Value are for non-catalytic or conventional woodheaters.
The column headed Aust. indicates compounds found in Australian studies of wood-smoke.
a identified in woodheater emissions in Australia (Chesterman 1984, Quraishi et al. 1984)
b identified in open fireplace emissions in Australia (Freeman and Cattell 1990)
c identified in woodheater and open fire emissions (Gras et al. 1992)
*Khalili et al. (1995) express their results as a percentage of PM10. The figures presented assume an emission factor for PM10 of 10 µg/m³.
BDL = below detection limit.

Table 4.3 provides a basis for estimating total emissions of various air toxics from residential firewood use in Australia. If annual firewood consumption is taken as 4 million air-dry tonnes (estimates range from 3 to 5.5 million tonnes), the oven-dry weight will be about 3.4 million tonnes. Multiplying the emission factors in Table 4.3 by 3400 will yield an estimate of total annual emissions, in tonnes, for each species (e.g. for benzo(a)pyrene: 0.003x3400 = 10.2t). However, the results of such calculations provide estimates that are extremely uncertain. All papers dealing with measurement of individual chemical species emphasise the variation that occurs through changes in fuel properties, appliance design, burn rate and other operating variables.

4.1.2 Australian government policy on wood-smoke

The Australian Government and all state and territory governments now have policies in place aimed at reducing emissions of wood-smoke from residential heating. Rapid development of policy on firewood collection and woodheater use has occurred since the mid-1990s, with new legislation now in place or proposed for all state and territory governments. In all cases, a combination of education on correct use of woodheaters, information on air pollution caused by woodheaters, and requirements for new heaters to meet emission factor limits is being used. Some authorities are regulating moisture content of firewood and adopting a code of conduct for firewood cutters to minimise ecological impacts. Proposed Tasmanian legislation to define smoke nuisance (visible smoke impacting on neighbouring properties for two minutes or more continuously) may give greater power to pollution control authorities to deal with problem cases.

4.1.3 Ambient air measurements of wood-smoke

Identification of wood-smoke particles as a proportion of total particles (e.g. TSP, PM10 or PM2.5) has been carried out in Australia and overseas using various 'tracers'. These studies have confirmed the significant woodheater and fireplace contribution to urban air pollution in communities where firewood use is common. Together with emission inventory techniques, some cities in Australia, e.g. Armidale and Launceston, have been identified as having winter PM10 concentrations dominated by wood-smoke. These cities experience PM10 concentrations well above the national goal of 50 µg/m³ (24 hour average) not to be exceeded more than five times per year. In Launceston, for example, this goal is exceeded 30 to 40 times per year.

4.1.4 Health impacts of wood-smoke

Much scientific analysis of epidemiological studies linking PM10 and PM2.5 to near-term mortality and morbidity has been carried out in many countries, including Australia. In summary,

  1. there is considerable support for the notion that particulates (PM10, PM2.5) have adverse effects on human health;
  2. particulate pollution appears to be associated with adverse health effects at all levels;
  3. certain subgroups in society are particularly vulnerable to adverse health effects from particulate pollution; and
  4. the precise nature of the biological cause and effect relationship between particulate levels and observed human health affects is unknown.

Wood-smoke is known to contribute a significant proportion of respirable and fine particles to the urban atmosphere in Australia, and so almost certainly contributes to these observed health impacts. Preparation of estimates of near-term mortality and morbidity attributable to wood-smoke in Australia is possible, but the large uncertainties in all parameters mean that credible estimates are not yet available.

The long-term impact of wood-smoke on human health is difficult to measure. Exposure to very high concentrations of wood-smoke (up to 100 times higher than ambient conditions measured in Australia) indoors in rural households in many developing countries provides evidence of serious health impacts. The limited evidence available suggests a correlation between the extent of exposure and health impacts.

The observed mutagenicity of wood-smoke and the presence of respiratory irritants and known carcinogens in wood-smoke support the hypothesis that long-term health impacts from wood-smoke exposure are likely. However, the risk associated with typical Australian exposure is not known.

4.2 Information shortages

It is likely that the majority of woodheaters in use in Australia will require replacement over the next 5 to 10 years. Thus, this is an opportune time to take action to reduce wood-smoke. Masonry fireplaces, due to their life-of-building operating life, may require different control strategies. In order to better inform policy decisions for wood-smoke reduction, certain information gaps, or conflicting sources of information, need to be overcome. This section highlights these information shortages and suggests options for surmounting them.

4.2.1 Estimating emissions

No data on actual emissions from woodheaters and open fireplaces in homes in Australia is available. Use of overseas information, mainly from USA, risks introducing errors due to differences in firewood species burnt, design of appliances and operator behaviour. Collection of data establishing particulate and CO emission factors for field operation of woodheaters and linking this to the large database of laboratory measurements of particulate emission factors for Australia seems important. This information would greatly improve emission inventory methods for modelling existing and future wood-smoke contribution to air toxics.

There are virtually no data available on toxic components of wood-smoke from eucalypt firewood. The present study, running in parallel with this literature review, should help fill this gap.

If emission factors, including air toxic components of total particulates, are better known, there is still significant uncertainty in the total mass of firewood burnt each year, especially within problem air-sheds. Surveys that rely on householders' opinions of how much firewood they consume in a year might be seriously in error (over-estimated) if 'short-weight' or wet firewood deliveries are common. More information is needed to establish a 'correction factor' between perceived firewood use and actual use.

Analysis of ambient particulates in different seasons and cities, with chemical tracers used to estimate the contribution of biomass combustion should continue. These measurements help assess the accuracy of inventory methods in predicting the contribution of wood-smoke to total particulate concentrations.

4.2.2 Estimating effectiveness of policy options

If they are effective, community education campaigns offer one of the more cost efficient means of reducing wood-smoke quickly. However, lack of information on the actual emission reductions across a whole air shed as a result of an education campaign means this control option is difficult to incorporate in an overall policy. There is some information available on how many people remember seeing an advertisement about correct heater operation, but not on actual behaviour change. A small air shed monitored in detail before and after an education campaign would help assess education options.

Woodheater certification requirements may be less effective than laboratory figures suggest because (a) woodheater models sold may differ from the models tested in the laboratory, and (b) retailers or householders may modify the heater, for example, to achieve longer burn times between refuelling. The reduced effectiveness of certification, if it is occurring, makes mitigation programs harder to plan. Some random tests of woodheater models purchased on the retail market, to confirm compliance with certification, would discourage unscrupulous manufacturers and put complying manufacturers on an equal economic footing. Identification of modified woodheaters presents problems of invasion of a householder's privacy, although some information might be gained when inspecting households creating a nuisance due to wood-smoke emissions (i.e. where access to the property occurs because of a known problem).

4.2.3 Health, economic and social factors

On-going epidemiological studies in Australia, examining links between air quality and health are desirable. However, evidence of new correlations, or impact of confounding factors, is most likely to come from large epidemiological studies overseas. Direct evidence of wood-smoke health impacts is likely to come from studies of extreme wood-smoke exposure of cooks using open fires in developing countries.

A detailed and objective theoretical study (with uncertainties well defined) identifying near-term health impacts of wood-smoke, using the best available local and international data, would be valuable. This would place the urgency of wood-smoke mitigation in context with other health issues in Australia.

A study of alternative heating options for different socio-economic groups, especially low-income families would be useful. This would allow informed discussion of policy actions aimed at discouraging woodheater use. The health impacts of inadequately heated houses, especially on vulnerable groups such as young children and the elderly, should also be addressed.

If firewood use is to be rapidly reduced, the impact of such a change on firewood cutters and their families should be investigated.

4.2.4 General research

In addition to the research and information gathering outlined above, several areas of research might yield useful outcomes for overall wood-smoke reduction.

4.3 Conclusions

The principle reason for studying air toxics and the contribution residential wood heating makes to these types of pollutants is to inform policy makers so that action can be taken to improve the health of Australian citizens. The adverse health impacts linked to wood-smoke require more detailed quantification. Simultaneously, the economic and social impacts of measures to reduce wood-smoke must also be quantified. The available evidence overwhelmingly suggests that wood-smoke in urban areas should be reduced, but the best method for achieving this goal is more elusive.

In broad terms, the options available to policy makers are as follows.

The Australian and international literature dealing with emissions of particulates, including air toxics, provides information to support some, but not all, these initiatives. Gaps in the information base have been identified.

Before you download

Most publications are available as PDF files. Adobe Acrobat Reader  is required to view PDF files.

If you are unable to access a publication, please contact us to organise a suitable alternative format.

Key

   Links to another web site
   Opens a pop-up window