Dioxins in ambient air in Australia

Extract from the Executive summary

This study was a component of the National Dioxins Program that was tasked to quantify and assess the concentrations and relative chemical compositions of dioxin-like chemicals in ambient air in Australia.

This report is part of the first phase of the National Dioxins Program, which involves the determination of representative levels of dioxins in Australia. Dioxins are predominantly generated as unintended by-products of combustion processes and are usually emitted into the atmosphere. Consequently, atmospheric transport represents the primary route for transport of dioxins into the environment. A recent review has shown that, potentially, combustion sources could be expected to contribute up to 95% of the total air-borne dioxin emissions, with up to 75% from fossil fuel and biomass combustion alone (Environment Australia, 2002). Dioxin-like PCBs have not been subject to similar review in Australia; in general they can be expected to have a different evolution, including sources, transport and, potentially, fate.

The overall objective of this report was to characterise dioxin levels in ambient air in Australia, at various locations including metropolitan, agricultural and remote reference sites. The primary aims of the study included consolidation of the current state of knowledge on dioxin levels in ambient air and achieving a greater understanding of dioxin levels in ambient air, by direct sampling.

Sampling for this report was conducted over a twelve-month period to establish possible seasonal variations in dioxin concentration, related, for example, to emissions from sources such as wood heaters and bushfires. The basic sampling period was one month. The ten measurement sites represent metropolitan, agricultural and remote area air sheds, with eight of the ten sites situated at excisting air-monitoring facilities. Samples were collected using high-volume samplers incorporating particulate and gas phase traps. Particulate phase collections were based on one-week exposure periods. Analyses, at the ultra-trace level, of toxic polychlorinated dibenzodioxins (PCDDs), polychlorinated dibenzofurans (PCDFs or furans) and co-planar polychlorinated biphenyls (PCBs) were conducted by the Australian Government Analytical Laboratories (AGAL) Sydney. Data were collated and processed at CSIRO Atmospheric Research.

Samplers used in the ambient air study were designed and constructed to collect both gas and particle phase dioxins and dioxin-like PCBs using established semi-volatile species collection methodologies. Samplers utilised open face quartz fibre filters for condensed phase. The gas traps, used in conjunction with the filters, were constructed from a medium-density PUF (polyurethane foam) plug, a layer of XAD^-2 resin and second PUF plug in a glass housing. The first PUF plug in each trap was pre-spiked using a range of isotopically-labelled surrogate standards for sampling loss determination.

Eleven samplers were deployed overall, including a duplicate sampler at one site (Alphington). The majority of samplers were operated at flow rates between 160-200 std. L min-1 (std. refers to standard volume, or reduced to STP) and an additional sampler, operating at approximately 1 m3 min^-1 was constructed for Cape Grim and operated from March 2003 to August 2003. Samplers in Queensland were operated at around 73 std. L min^-1. The sampled air volume was determined weekly, using built-in calibrated gas meters (or orifice plate for the Cape Grim high-volume sampler).

Sampling sites used for the ambient sampling project, represent the three broad geographic regions of northern Australia, south-eastern Australia and south-west Western Australia, and priority air sheds represented include Darwin, south-east Queensland, Sydney, Port Phillip and Perth, with the later addition of Adelaide.

The ten sites used for the study were:

• Wattleup, in the Kwinana area, Perth, WA (industrial)
• Duncraig, Perth, WA (mid-sized urban)
• Berrimah, Darwin, NT (small urban within a remote rural region)
• Eagle Farm, Brisbane, south-east Qld (light industrial)
• Mutdapilly, south-east Qld (agricultural/grazing)
• Westmead, Sydney, NSW (major urban, some light industry impact from mixed industrial/urban air shed)
• Boorolite, lower north-east Vic (agricultural/grazing)
• Alphington, Melbourne, Vic (major urban area, Yarra Valley)
• Cape Grim, Tasmania (pristine remote)
• Netley, Adelaide SA (light industrial)

The nominal start date for the ambient program was 1 September 2002 and the end date for sampling was 31 August 2003 with trap changes nominally at the end of each calendar month. Some latitude in change dates (a few days) was allowed to meet local logistical requirements. The sampling program for Netley, SA, was incorporated as an addition to the main sampling program after the start of the main program, and was commissioned in January 2003 and ran until January 2004.

Quality control procedures were established and documented before the start of the sampling program and included a wide range of checks on both sampling and analytical procedures. In general, data collection and analysis went smoothly and most data quality goals were attained. Two filters from Queensland were lost in the mail - these represented only small fractions of the monthly samples for those sites and resulting concentrations could be substantially corrected. Also two months' samples (June - July) from the Sydney site, were analysed with incorrect filter and gas trap combinations; however, these data were also substantially correctable using separate gas and filter analyses for the complementary filter and trap combination, retaining effective monthly samples for that site (and providing some limited qualitative information on phase partitioning).

Further information

• More about dioxins