Environment Australia, October 2000
EPA (Victoria) Conference Room, Melbourne, Tuesday 3 October 2000
In June 1998, the National Environment Protection Council (NEPC) made the National Environment Protection Measure for Ambient Air Quality, which sets standards for the six criteria pollutants; namely carbon monoxide, nitrogen dioxide, ozone, lead, sulfur dioxide and particles (measured as PM10).
At that time, the NEPC noted that a PM2.5 or PM1.0 standard might soon be a better indicator of public health impacts and agreed to commence a review of the particles standard by 2001, in particular, the need for a standard for particles less than 2.5 micrometres.
The Airborne Particles Health Forum was designed to assist this review and guide the Commonwealth's future work program in particle research and management.
The objectives of the forum were:
- to identify key issues and knowledge gaps associated with setting a particles standard that provides the best indicator of public health impacts; and
- to develop priorities for national action to improve our understanding and management of particles.
To meet these objectives, the forum was asked the key questions:
- does the current PM10 standard provide the best indicator of human health risk? If not, what will?
- does the research evidence support this view? Or is further research required?
- can we readily measure the preferred indicator? If not, what are the problems and can they be resolved?
- can we readily identify the sources of the preferred particle indicator? If not, what work is required to enable source identification?
- do feasible source management options exist?
The Airborne Particles Health Forum is a Clear the Air initiative which is part of the Commonwealth Government's Natural Heritage Trust funded Air Pollution in Major Cities Program.
Further information can be obtained by contacting the Air Quality Section of Environment Australia, telephone 02 6274 1890, facsimile 02 6274 1172, e-mail firstname.lastname@example.org or Internet www.environment.gov.au/atmosphere/airquality.
Dr Lyn Denison, EPA (Victoria)
Dr Denison gave a summary of the recent research into the health effects of particles. Exposure to particles has been associated with increases in daily mortality, increases in hospital admissions for respiratory and cardiovascular disease, exacerbation of existing disease (including asthma) and increased use of medication. Groups within the population that are sensitive to the effects of particles include people with existing respiratory or cardiovascular disease, the elderly, and children. The issue of what particle size fraction is responsible for observed health effects is still open for debate. The role of particle composition is also still unclear. Studies have shown that both combustion particles and crustal particles are associated with adverse health effects. Recent studies conducted in Detroit and the Coachella Valley in the USA have shown that PM10–2.5 is associated with increases in daily mortality. In the Detroit study the effects for ischemic heart disease and stroke were stronger for PM10–2.5 than for PM2.5. Similar results were obtained in the Coachella Valley (California) where the association between increases in daily mortality was as strong for PM10–2.5 as it was for PM10. In contrast, studies from Boston, where approximately 80% of the PM10 is PM2.5, the effects of PM2.5 were stronger than those observed for PM10–2.5 or PM10. The contribution of the coarse fraction to PM10 levels in Detroit and the Coachella Valley is greater than that observed in Boston. The role of ultrafine particles in the health effects attributed to exposure to particles is also still strongly debated.
Recent research from the US, conducted as part of the National Morbidity, Mortality and Air Pollution Study (NNMAPS), has shown that across the 20 cities examined in Phase I of the study significant variability in results was obtained. The associations between increases in daily mortality and PM10 levels were strongest in Southern California and Texas. A combined effect estimate across these cities of 0.34% increase in daily mortality per 10 µg/m³ increment in PM10 was obtained. Phase II of the analysis, conducted across 90 US cities, found similar effects with a combined effect estimate of 0.5% increase in daily mortality per 10 µg/m³ increment in PM10 was obtained.
An analysis examining the extent of life shortening associated with exposure to particles found that both a long-term effect and 'mortality displacement' were important in the observed increases in mortality.
A number of studies have been conducted in Australia to investigate the effects of air pollution on public health. In general the effects observed in these studies are similar to those observed in studies conducted in the US, the UK and Europe. Earlier studies conducted in Sydney and Brisbane showed strong associations between fine particles (measured by nephelometry) and increases in daily mortality and hospital admissions. More recent studies from Melbourne and Perth found that the associations observed were not as strong as those observed in Brisbane and Sydney. The importance of these differences is being examined as part of a National health study coordinated by Griffith University.
Dr Malcolm Brown, The Shell Company of Australia
Dr Brown raised concerns with the existing NEPM; in particular that the stages in the development of the NEPM were done concurrently rather than sequentially. This resulted in a 'dogs breakfast' that need not meet the requirements of many stakeholders. The method of prioritisation within the NEPM is uncertain. The view of industry is that there is a clear need to look at a standard for particle sizes lower than PM10.
Dr Brown encouraged that any future development of a standard for particle sizes less than PM10 be done in a method that is both logical and practical. In particular he emphasised the need for a clear, transparent, and logical analysis of epidemiological studies. Criteria used in the assessment of such studies need to be clearly stated. He identified exposure data as being the main weakness in studies to date and strongly recommended that the limitations of these studies be exposed.
The uncertainty within any risk based health assessment needs to be crystal clear to the lay person and, in particular, the politician. Individual studies used in the risk assessment should be given a rating of uncertainty to quantify the accuracy of the findings.
As part of the NEPM development industry studies, such as those listed below, should be included in any review:
- CONCAWE's 1999 report, The health effects of PM2.5 particles (including ultrafine particles). Downloadable from http://www.concawe.be/Html/Reports.htm
- The NSW Minerals Council report on the health effects of particles.
Management options should be addressed in the initial phase of any NEPM development to ascertain what is the most economically practical solution. Budget requirements may mean that the most practicable avenue forward is that the existing NEPM undergo a process of continual review. Alternatively it may be economically viable to establish a new PM2.5 standard using reputable overseas studies as a foundation.
Ms Bronwen Machin, Environment Victoria
Ms Machin, who is a Clean Air campaigner with Environment Victoria, presented community views on air quality standards. Ms Machin stressed that a 'triple bottom line' philosophy was necessary in order to achieve the most beneficial outcome. In other words social and environmental concerns should not be sacrificed or compromised for economic returns.
Short term economic costs should not be considered a relevant deterrent to development of a new standard, since the standard development process will take a considerable amount of time to implement, thus giving adequate time for industry to meet the long term objectives of the standard.
The lack of sufficient 'rocket science' and exposure levels should not preclude the development of a standard. The standard should not be 'set in concrete', rather it should be of malleable format, easily adaptable as more exact scientific data is made available.
Speaking for environment groups in general Ms Machin believes that 'they were robbed' at the last NEPM, with the PM2.5 standard not being included in the NEPM. Ms Machin supports the scientific research into PM2.5 but insisted that lack of research should not preclude the development of a new PM2.5 standard.
There was considerable discussion on the health effects associated with particles, focussing on the state of knowledge of chemical and physiological mechanisms for the observed health effects. There was also discussion on the chemical aspects of fine particle activity.
While it was recognised that particles have health effects with no apparent threshold and suggested that we are close to realising a physiological mechanism, the main point, clearly expressed, was that current research into health effects of particles has not been able to clearly attribute these effects to specific size fractions ie. PM10, PM2.5, PM1 or ultrafines.
There exists a comprehensive body of knowledge in relation to PM10 however this is not as extensive for smaller size fractions such as PM2.5. This is due to the fact that PM10 is monitored more widely than the other size fractions. Being a subset of PM10 makes it very hard to separate the related health effects of PM2.5 from those attributed to PM10. Lack of monitoring data for smaller size fractions makes health studies difficult. Health studies within Australia may be limited due to the lack of substantial population densities.
If a second particle size standard were to be developed the main question would be, what is the most appropriate and practicable particle standard to choose at this point in time? There was discussion that at this time there is not sufficient evidence to support a change from a PM10 to a PM2.5 standard, however, consideration could be given to introducing a PM2.5 standard in addition to the existing PM10 standard.
Dr Greg Ayers, CSIRO Division of Atmospheric Research
Dr Ayers presented an overview of the characteristics of particles within the air shed and the complexities in choosing the appropriate measurement methodology to achieve reliable and representative data.
Dr Ayers explained that different aerosol particle measuring methodologies will reflect different segments of the aerosol size distribution. The chemical composition of particles differs with particle size. Different particle measurement methods respond differentially to size distribution properties.
The variations in relative humidity greatly affect the measured particle properties. Aerosol water condensed at high humidity can be driven off effectively by sample heating to 50°C. However in some cases, semi-volatile material contributing significantly to particle mass is also lost from aerosol samples at this temperature. This is an issue with the use of TEOMs for particle measurements. An alternative to heating the sample is to dry the aerosol at ambient temperature. This technology although not yet demonstrated is technically feasible.
Current practice shows that there is a comparable difference in terms of accuracy and precision between the different PM10 measurement systems. At present these systems fall short of the 'expectation' listed in AS 3580.9.6-1990. In general measurements using similar methodology probably compare better then measurements using different methods.
The chemical composition of PM10–2.5 and PM2.5 can be significantly different. However PM2.5 and PM1 are of similar chemical composition. At this point in time PM2.5 measurement methodologies are more precise then those developed for PM1, and the PM2.5 measurements were a good compromise (or surrogate) for PM1 measurements. PM1 is not routinely monitored.
Nephelometry has proven to be a useful method for measuring fine particles with substantial data having been collected in Victoria over the past twenty to thirty years. This data could provide a useful database to guide the setting of a standard for PM2.5 however for this to occur it is necessary for correlations between the nephelometry data and PM2.5 mass measurements to be determined for each airshed.
The particles that make up the PM2.5 size fraction have longer lifetimes in the atmosphere compared with reactive gases and PM2.5 levels may be more uniformly distributed than reactive gases in some airsheds.
The general consensus was that there is a need to assess whether long-term, annual average standards are necessary for PM2.5 and PM10. The question arose to whether there was a need for personal exposure data and if so, to differentiate between indoor and outdoor air environments in relation to personal exposure. It was agreed that in the absence of indoor sources, there was good correlation between indoor and outdoor values. Further to this there is a need to ensure monitoring occurs over time periods reflective of relevant health effects.
Further discussion was had on the imprecision of PM1 measurements, in particular the problems with 'noise' in Victoria. The consensus was that PM2.5 should be measured with continuing research into PM1 measurement.
Increased loss of volatiles while sampling PM2.5 using TEOMs was raised as a concern, particularly since some jurisdictions have made substantial investments into TEOMs.
There is a need to assess whether standards set using the results of epidemiological studies that have utilised hi-volume sampling data can be monitored against using TEOMs, knowing that there is a difference between the mass measured by these two methods. It is necessary to ensure that particle sampling methods reflect the air that people are breathing.
Long range transportation of particles may cause anomalies in sample data, there needs to be further research into this phenomenon within Australia.
Dr David Cohen, ANSTO Physics Division
Dr Cohen presented an overview of an ongoing program in NSW investigating the composition of particles and identification of the major sources. As mentioned by Dr Ayers the chemical composition of particles differs systematically with particle size. The chemical constituents of PM1 and PM2.5 differ significantly from that of PM10.
Through the use of reliable particle composition data it is possible to trace the source of particles within the airshed back to particular industries. Performing a chemical mass balance of chemical elements within a sample allows us to determine the proportion of chemical constituents from each industry within the airshed. This data can then be used to guide the development of new particle standards.
In summary Dr Cohen expressed the need to look more closely at the new goals that maybe set and to draw on the knowledge of national and international research. Large scale seasonal and regional variations between different sources and different samples makes for an interesting challenge ahead.
There was no discussion here.
1. Does the current PM10 standard provide the best indicator of human health risk? If not, what will?
The view of the workshop participants was that it would be of considerable benefit to introduce a PM2.5 standard in addition to the existing PM10. The general consensus within the forum was that the PM10 standard should remain with the addition of a PM2.5 standard as the health research cannot clearly attribute health effects to one size fraction. Maintaining a standard for both size fractions also allows for different management options to be implemented to account for the different sources of particles in each size fraction.
2. Does the research evidence support this view? Or is further research required?
Currently there is enough evidence to show that PM10 particles have a considerably different chemical composition to PM2.5. Due to the present lack of monitoring of smaller size fractions conclusive evidence is not yet available that health effects are mainly due to the smaller size fractions. At present evidence from epidemiological studies suggests that adverse health effects are observed with both PM10 and PM2.5 supporting standards for both size fractions.
3. Can we readily measure the preferred indicator? If not, what are the problems and can they be resolved?
We can readily measure both PM10 and PM2.5, however problems with loss of volatiles from the particle mass do exist in using TEOMs. This loss may be more significant for PM2.5 than for PM10. With substantial capital having already being invested in some jurisdictions into TEOMs it is paramount that research be conducted to resolve these measurement issues and ascertain the significance in such losses.
4. Can we readily identify the sources of the preferred particle indicator? If not what work has to be done to enable source identification?
Yes as shown by the results of the work presented by Dr Cohen, sources of particles within an airshed can be identified. The monitoring of PM2.5 will allow a more precise identification of combustion related sources.
5. Do feasible management options exist?
Management options will very from case to case with the level of complexity being related to the types of sources within an airshed. It is not appropriate to apply a broad management approach across the different size fractions. Feasible management options do exist for most sources of particles.
Environment Australia would like to thank all of the speakers that gave their time to present at the Forum, in particularly Dr Neville Bofinger (Clean Air Society of Australia and New Zealand) for acting as facilitator.
Special thanks are also extended to the EPA Victoria for providing meeting facilities and minute taker.