By: Ian Longley
EECA report: “Indoor combustion in New Zealand homes: health effects and costs”
Our commentary and opinion
A new report has shown how emissions from gas stoves and unflued gas heaters pose a burden on health that is comparable in size to the burden from outdoor air pollution. But the report also contains much more information and subtlety that might be lost in the inevitable drive to turn the report into headlines. Later in this blog we’ll discuss the issue of uncertainty, but first we address the important distinction between “indoor” and “outdoor” air pollution.
On 24th September, EECA (the Energy Efficiency and Conservation Authority) published a report (prepared by Emissions Impossible Ltd) on the health costs associated with air pollution from residential cooking and heating appliances, mainly by comparing gas and wood burning devices.
As well as the finding for gas stoves, the report also notes that leakage of smoke from woodburners into the home also pose an additional risk not accounted for in the “HAPINZ” family of outdoor air quality risk assessments.
What the report also notes – but is easily missed – is that human exposure to the contribution of woodburners to “outdoor” air pollution mainly occurs indoors, through the infiltration of outdoor smoke into the home.
Press coverage of the report so far has focussed on the results for gas stoves as being the main story (“Pollution from gas stoves linked to hundreds of deaths a year in New Zealand” – headline, Stuff.co.nz, 24th September) with industry group GasNZ questioning the data.
Yet, if you use the report’s own data to calculate the contribution of each appliance studied to the total health costs (combining indoor and outdoor components) of $12.1 billion per year, you find that 65 % comes from woodburners, and most of that (43 %) from burners that comply with National Environmental Standards. Gas stoves contribute 27% and unflued gas heaters contribute 8 %.
Whereas exposure to the “indoor” component is limited to those living with the appliance, the “outdoor” component is distributed across the community. Using the report’s own figures, the total cost of “outdoor” pollution (which comes entirely from wood-burning appliances) is approximately $6.3 billion per year. Just assuming that burden is spread equally across New Zealand’s 1.8 million homes, that represents $3,400 per home. But it won’t be evenly distributed – it will be distributed in proportion to local woodburner prevalence and dispersion, for which outdoor PM levels are a proxy. According to the HAPINZ 3 report (by the same team) annual mean PM2.5 associated with domestic heating in New Zealand varies from 0.9 to 12.6 mg m-3 following a skewed distribution with a median of 2.0 mg m-3 and mean of 2.7 mg m-3. 10 % of the population are exposed to double the mean level and 4 % to three times. So, for 180,000 homes, the burden posed by outdoor air pollution from appliances is $6,800 each, or more (up to $16,000 in the most polluted locations).
This means that for homes in more polluted airsheds (which are generally more polluted because woodburners are more common) the impact arising from infiltration of outdoor woodsmoke is comparable to the impact from direct pollution from a non-NES woodburner, and also similar to the direct impact from a gas stove. So, in these situations, woodburners have roughly double the impact of gas stoves.
The other issue that leaps out from the report for us is uncertainty. The study’s scope was not to generate original raw data, but to base a health economics analysis on existing published data. That published data is very sparse, mostly from overseas, some of it quite old and not particularly representative of NZ homes or appliances. The studies available were not designed with EECA’s questions in mind. This is, of course, no fault of the study but it does mean that the uncertainty in the analysis is very large. The report acknowledges this in its text but presents results tables with numbers quoted to 4 or 5 significant figures with no indication of uncertainty. Thus, it is easily inferred that the estimates are far more accurate and precise than they actually are.
To take just one example, we suspect that the estimates for woodburner leakage might be too high. They are derived from studies comparing PM levels in homes with woodburners with those without, assuming that higher levels in the former can be explained by leakage. But this is not the only possible mechanism, however.
Data on woodburner leakage – including how common it is – is very rare and contentious. In our experience of monitoring well over 100 homes with wood burners we have only observed this phenomenon a handful of times. On the other hand, we have observed how woodsmoke readily infiltrates into every home we have ever studied. But the degree of infiltration from outside is directly related to outdoor air quality, which, in New Zealand, is most strongly driven by the prevalence of wood-burning appliances in the neighbourhood. In other words, if your home has a woodburner you are more likely to live in a neighbourhood with a higher density of woodburners, hence poorer outdoor air quality, hence more infiltration, hence poorer indoor air quality.
Overall, this report presents a welcome first step to filling a significant knowledge gap, namely the impact of poor indoor air quality on health and wellbeing in New Zealand. But we think as a matter of scientific integrity, the very large inherent uncertainties need to be acknowledged. Without this there will never be any incentive to reduce those uncertainties. And whereas reducing error in absolute estimates is important for informing cost-benefit analyses that often drive or inhibit action, reducing relative errors between the indoor and outdoor pathways is also important as each implies different mitigation strategies. And finally, the relative importance of each pathway depends on geography and will be quite different between Invercargill and Auckland, potentially implying different solutions in different locations.
Our final point is that our recent research has largely been focussed on the method development needed to capture the raw data that would substantially reduce those uncertainties, at much reduced costs that was possible only a few years ago. Keep an eye on our website as we will be sharing how we’ve achieved that and how this new capability could be used.
