Atmosphere

Emissions from domestic solid fuel burning appliances (wood-heaters, open fireplaces)

Technical Report No. 5
J. Gras, C.Meyer, I. Weeks, R. Gillett, I. Galbally, J. Todd, F. Carnovale, R. Joynt, A. Hinwood, H. Berko and S. Brown.
Environment Australia, March 2002
ISBN 0 6425 4867 6

8. Discussion - Emissions observed in this study compared with national and international standards and guidelines

Combustion of wood is a complex process with many factors governing its progress. In partially-controlled conditions, such as the certification testing of wood heaters and in this and other emission studies, apparently 'random' variations in emission levels still occur, resulting from the less-well-controlled or even uncontrolled factors. National and international emission guidelines are generally set against a background of a 'standard' measurement procedure, for good reason. Such 'standard' procedures have been developed to allow unbiased comparison between different appliances and have generally been set with a view to capturing features that the standards developers have identified as the principal determinants of heater performance or operation. In order to achieve reproducibility, many of the known variables are controlled and in addition constraints such as allowable variance in some parameters (e.g. efficiency) included. How well, or poorly, a particular method captures or represents typical operation is a legitimate and frequent subject for criticism of test procedures (e.g. Gilmour and Walker 1995; Houck and Tiegs 1998).

Factors known to influence the amount and nature or composition of emissions include fuel properties, e.g. fuel species, calorific value, density, water content, trace material content in fuel, loading of the appliance, e.g. size of logs, disposition/location, spacing, volume of chamber used, chamber temperature at loading, size of coal bed, the design of the appliance itself, e.g. presence of baffles, residence time of gases in the combustion chamber, airflow path, secondary flow, thermal mass; as well as the mode of operation, e.g. airflow rates, frequency of loading, method of kindling, etc. Geographic factors including altitude may also be important but less so for Australia than other countries with populations living at high altitudes. So many individual variations of the basic independent variables are possible in homes that no single test methodology can realistically hope to fully capture the whole range of elements that might constitute 'typical' operation. Not surprisingly, different compromises have been made in different jurisdictions in an attempt to develop reproducible, affordable and meaningful test methodologies. Unfortunately this seriously complicates comparisons of test results and the evaluation of test results against standards that are written around different test methods.

The present study was conducted, where possible, under the Australian/New Zealand test regime. Departures were made to accommodate specific burn conditions, such as high moisture content, fuel density, overloading etc and due to the limited scope of the project there were fewer repetitions than specified for the AS4013 standard. Also, limits on acceptable variance in the efficiency were not included. Australian and New Zealand test procedures for determination of emissions from most wood burning appliances are given in AS4013 and associated standards (AS4012, AS4014). These standards and the history of their development are detailed more fully in Technical Report No. 4: Review of Literature on Residential Firewood Use, Wood-Smoke and Air Toxics. Briefly, AS4013 is a dilution tunnel method, it employs dry hardwood of specified density and size and incorporates measurements at three different flow settings (high, medium and low airflow rates) with specified repetitions and conditioning burns. Emissions are determined as particle mass (effectively TSP sampled orthogonally to the dilution tunnel stream). Initially, in 1992, the emission standard was set at 5.5 g/kg, and this was lowered to 4 g/kg in 1999. For appliances with catalytic converters the current limit for compliance is 2.25 g/kg.

US Federal EPA certification emission limits have also evolved with time. Certified Phase I heater emission limits were set at 8.5 g/h (5.5 g/h for catalyst equipped) coming into effect from July 1 1988 with compliance for new appliances mandatory after July 1, 1990. These appliances then became exempted from local burn bans. At this date also Phase II regulations came into effect with an emission limit of 7.5 g/h (and 4.5 g/h for catalyst equipped appliances) with mandatory compliance for new appliances by July 1, 1992. For a first approximation to converting from the US emission rates to emission factors, a typical burn rate of around 1 kg/h is often used (e.g. Gilmour and Walker 1995). Testing is to US EPA method 28, for fuel and appliance operation with methods 5G and 5H defining the emissions sampling procedures. Method 28 requires the use of air-dried Douglas fir 2x4 or 4x4 inch timber (16–20% moisture wet weight) constructed into fixed dimension cribs. Testing is conducted at four burn rates. Emissions may be sampled using a dilution tunnel (Method 5G) or directly from the heater flue (Method 5H). These procedures are summarised and various deficiencies and strengths discussed by Houck and Tiegs (1998). Local codes e.g. Washington State administrative Code 150-31-200 and Colorado Regulation 4 may apply stricter requirements on the operation of wood heaters.

The European Commission Comit. 'Europ' de Normalisation Technical Committee (CEN/TC 295) working group WG 5 in Belgium has prepared the draft European (CEN) standard prCEN – 13240. This method samples under nominal loading (Pettersson, 2001). Proposed emission standards considered are 2500 mg/m³ for CO, 150 mg/m³ for particle mass and 100 mg/m³ for total hydrocarbons.

Individual national standards exist for a number of western European countries including Norway, Austria, the U.K and Sweden (Skreiberg et al. 1997). In Denmark, CO emissions are regulated with a limit of 0.3% CO at 7.5% CO2 in the flue gas.

The Swedish test method has been based on (Sveriges Provnings och Forskningsinstitut) SP-1425 for wood stove emissions, using CO and hydrocarbons at three fire output levels, low power, 3–5 kW and maximum power. This also includes a test point after insertion of a log (Pettersson 2001). Limits for particle emissions had been based on 40 mg/MJ, for the combined aerosol refractory and tar components (Skreiberg et al. 1997). In a revision to SP-1425 the limit is now based on gaseous organic compounds (OGC), using the carbon content derived from the total hydrocarbon concentration. The limit is for a maximum of 250 mg OGC per cubic metre, dry gas, at 10% CO2 (L. Gustavsson, personal communication, 2001).

The Norwegian method measures aerosol mass using a dilution tunnel and also requires determination of thermal efficiency. Emission limits are 5 g/kg for catalyst-equipped appliances and 10 g/kg for non-catalyst appliances.

Since 1999 the Austrian wood heater emission limit has been set at 60 mg/MJ based on standard EN 303-5 (following Art. 15a BV-G, Lasselsberger 1999). This specifies limits for small hand-stoked furnaces. For CO the limit is 1100 mg/MJ, NOX 150 mg/MJ, HC 80 mg/MJ and dust 60 mg/MJ. For this testing method dust refers to the refractory component of the aerosol. Allowing 20 MJ/kg for fuel heat capacity, this equates to 1.2 g/kg for refractory aerosol or around 2–6 g/kg for total aerosol using a reasonable range of estimates for the refractory fraction.

The German standard for wood stoves is designated DIN 18891 and the standard for open fireplaces DIN 18895. Emission limits are based on flue CO concentration normalised to 13% O2. For DIN 18891 the CO limit is 0.4%, and for DIN 18895 the CO limit is 0.5%. Local requirements may be more restrictive, for example in Stuttgart the requirement is for emissions of CO less than 0.2% and in Regensburg less than 0.12% CO, both relative to 13% O2.

In the UK, Section 20 of the Clean Air Act of 1993 prohibits the emission of smoke in smoke control areas. Within these areas use of only smokeless fuels, or exempt low-emission appliances are allowed. The relevant UK standard for room heaters burning wood logs is BS 7256:1990 with a permitted mass emission level of 5.5 g/kg.

Recent attempts to produce an international standard include the draft standard ISO 13336, which, based heavily on the Australia-New Zealand standard AS/NZ 4013, is also an aerosol dilution tunnel, calorimeter room approach. At the recent international voting this standard was not ratified (J. Crouch, personal communication, 2001).

A direct measurement comparison between draft standard ISO 13336 and the draft European standard prEN 13240 was reported by Gaegauf and Macquat (2000). These authors show equivalence of results at three burn rates for efficiency testing using calorimeter room and stack-loss methods (CO2, CO). Emission factors based on CO determination were also shown to be equivalent using the two methodologies although the ISO method, as developed, is an aerosol emission method. The need for further testing and harmonisation of test methods was emphasised. In relation to compatibility between the US testing methods and other international methods, Houck and Tiegs (1998) express the opinion that the ISO 13336 test is too different to the US tests (EPA methods 28 with method 5G and 5H) to give anything but qualitative correlation. Appliances with low emissions for one test will also be low on the other. Because of the similarities of ISO 13336 and AS/NZ 4013 the comments of Houck and Tiegs can be taken to also apply equally to comparisons between the US and Australian standards. Canadian Standards Association emission standard (CSA) B-415 reflects the US EPA standard although recent versions are slightly more inclusive, having removed exemptions for cook stoves and central heating systems (CSA B-415.1).

In comparing emissions data from different studies or assessing emission data against jurisdictional guidelines or standards it is imperative that the methodologies followed in those studies are also comparable. If results are assessed against an emission guideline, the background 'standard' measurement protocol relevant to that jurisdictional guideline also needs to be known. The degree of variance that results with wood combustion-emission studies is perhaps not generally appreciated, but can be very large. To some extent the methodologies adapted for 'standardised' testing in any particular jurisdiction mask this variability. One very significant study in this respect is the round-robin measurement series reported by Skreiberg et al. (1997). For their study a single (catalyst equipped) wood heater was shipped to nine different countries where it was tested using the relevant national protocols, where these exist. Emissions tested included aerosol mass, CO, hydrocarbons and NOX, although not all parameters were measured in all countries. For all of these emissions the typical variance is around a factor of five (and in some cases more than an order of magnitude) in concentration (as g/kg dry fuel mass) for any given fuel consumption rate across the range employed (approx. 0.7–3 kg/h dry). Similar, or greater, variance must be expected for aerosol trace composition measurements such as dioxins, PAHs etc. where much more complicated sampling, extraction and analysis procedures are involved. This is a significant factor that must be taken into account in assessing or comparing emission factors such as those given in this study and summarised in Tables 8 and 9, or when using emission values reported from studies in other jurisdictions.