PM 2.5. an overview of composition, sources & health effects. Mathew Heal School of Chemistry, University of Edinburgh

Transcription

1 PM 2.5 an overview of composition, sources & health effects Mathew Heal School of Chemistry, University of Edinburgh Scottish Air Quality Database and Website Annual Seminar, 18 th March 2013

2 SNIFFER report: PM 2.5 in the UK SEPA report: PM 2.5 in Scotland or contact John Lamb Defra AQEG report: Fine particulate matter in the United Kingdom WHO report: Review of evidence of health aspects of air pollution

3 The potential hazard: deposition in the respiratory system as a function of particle size Tracheobronchial deposition efficiency PM 2.5 Alveoli deposition efficiency PM 2.5

4 Health effects of PM 2.5 are associated with both short- and long-term exposures Exposure Outcome Expert view on causality Short-term Mortality Causal Manifest as illness, hospital admission & death Cardiovascular effects Respiratory effects Central nervous system Causal Long-term Mortality Causal Likely to be causal Inadequate evidence Manifest as decrease in life expectancy & chronic illness Cardiovascular effects Causal Respiratory effects Likely to be causal Cancer, genotoxicity Suggestive Reproductive & neurodevelopment Suggestive USEPA (2009)

5 The quantitative impacts on health from Short-term exposure to PM 2.5 Few studies for PM 2.5 but plenty for PM 10 Long-term Daily outcome Relative risk per 10 g m -3 PM 10 All-cause mortality 0.6% ( %) Respiratory mortality 1.3% ( %) Cardiovascular mortality 0.9% ( %) Hospital admissions 0.8% Outcome Relative risk per 10 g m -3 PM 2.5 All-cause mortality 6% (2 11%) WHO (2004) meta-analysis Cardiopulmonary mortality 9% (3 16%) Lung cancer mortality 8% (1 16%) COMEAP (2009) (from US ACS and 6-cities studies) Health impacts dominated by long-term exposures

6 What do these health risks mean in practice? (for the UK) Short-term For PM 10 levels in the UK in 2001, the IGCB estimate: 6,800 deaths brought forward 6,700 each for respiratory and cardiovascular hospital admissions Long-term For PM 2.5 levels in the UK in 2008, the IGCB estimate: ~6 months average loss of life expectancy Other average loss of life expectancies: 1-3 months due to road traffic accidents 2-3 months due to passive smoking

7 Is there a safe level for exposure? None yet identified at population level But variability in individual susceptibility very likely - effects expected to be greatest amongst children, the elderly and those with pre-existing disease

8 Are specific components/sources of PM 2.5 more toxic? A cautious yes for example: - redox-active and/or toxic transition metals - redox-active and/or toxic organic species - ultrafine (< 0.1 m) particle numbers - particles from road traffic or combustion 10 m particle The current WHO review says more on this 2.5 m particle 1 m particle 0.1 m particle 5 m

9 Exposure: PCM-modelled PM 2.5 in the UK Populationweighted PM 2.5 exposure g m 3 Scotland 5.5 Wales 8.3 Northern Ireland 6.4 Inner London 14.1 Outer London 13.4 Rest of England 10.6 (2010)

10 PM 2.5 across Europe Populationweighted PM 2.5 exposure g m 3 UK 10.3 Sweden 11.0 France 13.2 Germany 15.3 Spain 16.1 Italy 23.8 (2010)

11 Composition and sources

12 DISPERSION Mixture of primary & secondary inorganic and organic particles PRIMARY PARTICLE & GAEOUS EMISSIONS GAS & PARTICLE CHEMISTRY Sources: anthropogenic & natural, local & distant Receptor RECEPTOR MODELLING CHEMICAL-TRANSPORT MODELLING

13 Typical urban background PM 2.5 composition Primary, mechanically generated Secondary inorganic Primary and secondary carbonaceous Measurements in Birmingham

14 NAEI emissions of primary PM 2.5 in the UK Waste Agricultural Residential Industrial 2008

15 Industrial sources of primary PM 2.5 in the UK 2008 Primary sources are many and diverse multi-sector abatement required

16 PM 2.5 Polar annulus plots (direction and time-of-day) NO x Southampton

17 Air-mass backtrajectory analysis for PM 2.5 Important influence of continental sources for elevated PM 2.5 (2009)

18 PM 2.5 source apportionment for west-east transect across London (PCM model) 20 Annual mean PM2.5 transect across London for 2009 from PCM model Urban increment Regional (nonurban) PM component (ugm -3) Except close to strong local sources, e.g. major roads, concentration dominated by background traffic area sources non-traffic area sources point sources urban dust rural dust regional primary secondary organic secondary inorganic residual sea salt 0 Henley Easting (m) Southend

19 Formation of secondary inorganic aerosol nitrate aerosol NaNO 3, NH 4 NO 3 sulphate aerosol NH 4 HSO 4, (NH 4 ) 2 SO 4 NaCl, H 2 O N 2 O 5 HNO 3 H 2 SO 4 SO 4 2 aq oxidation oxidation NO, NO 2 NH 3 SO 2 Combustion point sources; transport Agriculture Combustion point sources; shipping PRECURSOR GAS EMISSIONS

20 Formation of secondary organic aerosol CO 2 oxygenated organic products OH, O 3, NO 3, sunlight oxygenated organic products Secondary organic aerosol OH, O 3, NO 3, sunlight NMVOC Industry; transport; biogenic PRECURSOR GAS EMISSIONS

21 Reductions in UK SIA PM 2.5 from 15% reductions in SO x, NO x or NH 3 emissions in different regions Reduction in UK PM 2.5 / g m SOx NOx NH UK Ireland Atlantic & North Sea France & Benelux Scandinavia inc. Baltic Sea Source country or region Germany & Central & East Europe Regions further afield & BCs

22 Sensitivity of PM 2.5 to 30% reduction in all Scottish emissions of primary PM 2.5, SO x, NO x & NH 3 Reduction in PM 2.5 (µg m -3 ) % reduction in PM 2.5 Max reduction ~10%; for most of Scotland reduction is only 3-7%

23 Sensitivity of primary PM 2.5 to 30% reduction in Scottish emissions of primary PM 2.5 % reduction in primary PM 2.5 Reductions up to 23% in central belt, 11-15% over most populated areas Response of PM 2.5 to precursor emissions reductions is highly non-proportional Significant contribution to UK SIA from non-uk emissions UK SIA is most sensitive to reductions in NH 3 and SO x insensitive to NO x emissions reductions on their own Greatest local leverage on PM 2.5 is via reduction in primary PM 2.5

24 Source apportionment of carbonaceous PM Biogenic SOA 38% Fossil EC 18% Fossil POA 8% Fossil SOA 13% by carbon-14 (plus some assumptions) by organic molecule tracers and PMF-CMB Vegetative detritus + other noncombustion contemp C 8% Biomass OM (wood, liq biofuel, cooking) 14% Biomass EC (wood, liq biofuel, cooking) 1% Biogenic SOA 34% Diesel engines 29% Significant non-fossil carbon contributions: biomass/biofuel, cooking, primary biological material, and in particular biogenic SOA Vegetative detritus 6% Wood smoke 3% Gasoline engines 3% Coal combustion 3% Smoking engines / other Yin et al. (2010) 22%

25 Some final remarks Exposure varies across the UK but is dominated by the background, except adjacent to strong sources ~6 month reduction in life expectancy from current exposures; toxic component(s) remain uncertain, but any reduction in PM 2.5 has potential (and significant) health gain Sources of PM 2.5 are many and diverse making control a challenge; reductions in primary PM 2.5 remain an effective local lever, reductions in SIA and SOA require considered transnational action There are co-benefits from emissions reductions aside from on PM 2.5 (e.g. on O 3 generation, and on eutrophic, acid & metal deposition)