The National Atmospheric Emissions Inventory

1
Uncertainties in the UK Heavy Metal Emissions
Inventory
Chris Dore AEA Energy and Environment
UK Emissions Inventory Programme Funded by Defra
RMP2106
2
Contents

• 1. Principles of Uncertainty
• 2. Combustion Sources
• 3. Non-Combustion Sources
• 4. Missing Non-Anthropogenic Sources
• 5. Mapping Emission Estimates
• 6. Conclusions

3
1. Principles of Uncertainty

• Combination of uncertainties

• Point sources- combination of random independent
  errors
• Area sources- one EF, prone to bias.

4
1. Principles of Uncertainty
5
2. Combustion SourcesSolid

• Coal, Coke, Anthracite SSF
• Important for As, Cd, Cr, Pb, Be and a major
  source of Mn, Hg Se
• Point Sources
• Well characterised, and emissions data are
  reported. Hence no bias is expected.
• But possible issues associated with extrapolation
  of few measurements

6
2. Combustion Sources Solid

• Coal, Coke, Anthracite SSF
• In 2005, over 2/3 of UK steam coal was imported
• 80 from Russia/South Africa, 13 from
  Columbia/Indonesia.
• Distribution with the ranges shown is not known.
• Fugitive emissions from Coke Ovens
• Area Sources
• Median values taken from a wide range
• Fuels assumed to be the same as coal
• Limited information on PM control
• Potential for significant bias

7
2. Combustion Sources Petroleum

• Petroleum Fuels
• Important for all metals except Cr and Hg, major
  sources of Be, Cd, Ni, Se, V Zn
• Metal emissions are primarily associated with
  petroleum coke, waste lubricants fuel oil
  (higher metal contents)
• However, large volumes of gas oil, DERV and
  petrol are consumed, giving notable contribution,
  despite their lower metal contents.
• Point Sources
• Variability would result in little impact on
  emissions total
• Area Sources
• Metal content taken from literature values
• Fuel oil variability of 2-4 times the mean value
• Gas oil/DERV very variable metal content data
• Waste lubricants (10 of Pb) factor of 10
  variability metal content

8
2. Combustion Sources Pb in Petrol

• Pb in Petrol, and Unleaded Petrol
• UK uses EFs based on measurement data
• Pb content of unleaded 0.04 mg/l (UKPIA 2003)
• 70 assumed to be released to air
• Small when compared to the limit value (5 mg/l)

9
2. Combustion Sources

• Burning of CCA Treated Wood
• CCA Cu, Cr As treated wood.
• Major source for As, important for Cr
• Data available on As consumed in CCA
  preservatives.
• But v difficult to estimate the quantities of
  wood burned
• Cr and Cu emissions extrapolated from As data
• Estimates could be out by a factor of 20.
• Cremation
• Well characterised in the UK
• Uncertainties of Hg would give a maximum impact
  of 10 increase to the UK total.
• Other Fuels
• Scrap tyres, MSW and wood in power stations
• Not a particularly large source

10
3. Non-Combustion Sources

• Metal Industry Processes
• Important for most metals
• Includes processes at steelworks (sintering and
  blast, basic oxygen, electric oxygen
  furnaces), foundries, 1 Al production, 2 Pb
  Al production, and various other non-ferrous
  metal processes
• Variety of point source data and literature data.
• Estimates are likely to only include stack
  emissions, and fugitives are therefore not
  accounted for.
• Chloroalkali Processes
• Important for Hg
• The main source is associated with the
  ventilation air from the cell room
• Very difficult to asses for a variety of reasons,
  assume that the emission could be underestimated
  by a factor of five (also used for other fugitive
  emissions)

11
3. Non-Combustion Sources

• Tyre Break Wear
• Important for Zn and Cu
• Estimates are a fixed fraction of PM10 emission
  from these sources
• Tyre wear is easy to estimate, but PM10 emission
  less so. UK specific data. Metal concentrations
  in tyres are highly variable (less so for HGVs).
• Metal content of brake linings is fairly well
  characterised.
• Emissions per vkm vary by nearly an order of
  magnitude. UK specific data.
• Odd and Ends
• A variety of other sources included in the
  inventory (fireworks, glass manufacture, waste
  incineration etc.)

12
4. Missing Non-anthropogenic Sources

• Missing Sources
• Accidental/malicious fires
• dwellings, factories, other buildings, vehicle
  fires
• Demolition
• Corrosion/abrasion of metal structures
• Galvanizing
• Non-thermal processing of scrap metals
• e.g. shredding of scrap metals
• Part B industrial processes
• e.g. cement batching, quarrying, powder coating
• Abrasion of road surfaces by motor vehicles

13
4. Missing Non-anthropogenic Sources

• Non-anthropogenic Sources
• Natural Sources
• not currently included or estimated. Estimates
  available from Ilyin Travnikov (2005)
• Marine Sources
• not currently included or estimated. Estimates
  available from Ilyin Travnikov (2005)
• Resuspension
• Estimates available from Ilyin Travnikov (2005)
  suggest significant contributions for Pb and Cd
  (trebling the 2004 Cd emissions in the UK).
• However estimates from Vincent and Passant (2006)
  for Cd, Pb, As, Ni suggested resuspension was not
  a major source.

14
5. Mapping Emission Estimates
15
6. Conclusions

• Conclusions
• There are areas where improvements need to be
  made
• However we are currently limited by data
  availability
• Recommendations for Future work
• Point Sources Obtain more information on whether
  fugitive emissions are included in current
  estimates
• Combustion Sources Obtain more comprehensive
  data on metal content of fuels
• Brake and Tyre Wear Review and consolidate
  existing literature information
• Missing Sources Make some initial estimates by
  improving PM10 estimates (not straightforward!)
• Natural Sources Incorporate estimates into
  emissions inventory
• Validation Verification
• After conducting these improvements, reassess the
  estimates derived from modelling in light of
  updated emissions inventory estimates.

16
THANK-YOU FOR YOUR ATTENTION