| Front page | | Contents | | Previous | | Next |

Update on Impact Categories, Normalisation and Weighting in LCA

6 Photochemical ozone formation

• 6.1 Summary
• 6.2 Description of impact category
• 6.3 Substances contributing to the impact category
• 6.4 Methodology
• 6.5 Normalisation references
• 6.6 Recommendations for future updating
• 6.7 References
• Appendix A.1: Calculation of normalisation reference for Denmark.
• Appendix A.2: Calculation of normalisation reference for EU-15
• Appendix A.3: Calculation of a worldwide normalisation reference

Karsten Fuglsang, dk-TEKNIK ENERGY & ENVIRONMENT

6.1 Summary

This chapter summarises the presently available data on emission of substances to the Danish and the European environment. The emissions include:

• nmVOC - non methane volatile organic carbons
• CO - carbon moNO_xide
• CH₄ - methane

Emission data for these substances are generally available for Denmark and EU-15 as well as for a number of other European countries. For the time being, worldwide emission data are available for 1990 and therefore the worldwide normalisation reference is based on 1990-data. This fact is exceptional for photochemical ozone formation. The normalisation references are presented in Table 6-1.

Table 6-1
Summary of normalisation references for photochemical ozone formation for Denmark, EU-15 and Worldwide in 1994.

1 1990.

2 Based on the extrapolation method outlined in section 3, 2, the normalisation reference can be calculated to 20 kg C₂H₂/capita/year.

The worldwide normalisation reference can be calculated to 20 kg C₂H₂/capita/year by the extrapolation method outlined in 3.2,General exstrapolation method. The extrapolated value is nearly the same as the value calculated by using the worldwide emission data for 1990. nmVOC accounts for 70-80% of the effect potential contributing to the normalisation reference. The Danish and EU-15 normalisation references are relatively certain, as the relevant emissions to air have been measured through many years according to the Convention on Long-range Transboundary Air Pollution and the following Protocol on nmVOC (adopted in 1991).

6.2 Description of impact category

6.2.1 Tropospheric ozone formation

Ozone is formed in the troposphere under the influence of sunlight when nitrogen oxides are present. When VOCs are also present, peroxy radicals can be produced. Peroxy radicals are highly reactive and toxic compounds, and the presence of peroxy radicals can result in an increase of the concentration of ozone through a complex reaction pattern. The primary reactions responsible for the formation of ozone in the troposphere are described in 6.3. Ozone is a secondary pollutant, as there is practically no ozone present in source emissions derived from human activity.

Tropospheric ozone, or ground level ozone, has been recognised as an important environmental impact on the regional scale. At high concentrations it is hazardous to human health, but already at lower concentrations it causes damage to the vegetation. Ozone is a trans-boundary pollutant, and it can be produced or consumed by other pollutants during transport over long ranges. The health problems caused by ozone have generally been considered to be an effect of the very high peaks of ozone concentration, known as ozone episodes. Increased background levels of ozone causes damage to vegetation, and thereby ozone also imposes an economic threat through a potential reduction of crop yield. It is assumed that anthropogenic emissions have resulted in a rise in the global background of ozone concentration from around 10 ppb in the year 1900 to around 20 ppb in 1975 (Fenger 1995).

6.2.2 POCP - impact category

In the EDIP model, the photochemical ozone formation is described through POCP, the photochemical ozone creation potential, as an individual impact category (Hauschild & Wenzel 1998). POCP is used in Europe for the ranking of VOCs according to their ability to produce ozone. In the US, a slightly different approach is used for the same purpose: Incremental Reactivity (Carter et al. 1995).

POCP describes the production of ozone from a VOC emission through computer modelling of a complex series of chemical reactions in the atmosphere over a given scenario. A large amount of input data is required for the calculation of POCP by the model. The input data consist of the following principal components (Derwent et al. 1996):

• Emission inventories for VOCs and NO_x
• Chemical and photochemical data
• Dry deposition rates for different surfaces
• Meteorological data

The model describes the chemical composition of primary pollutants during the transport away from their sources and of secondary pollutants during the transport towards the sensitive receptors where environmental damage may occur. In the model, the chemical composition of parcels of air is followed as they travel trans-boundary across Europe. Emissions of NO_x, CO, SO₂ and VOCs are introduced into the air parcels in a series of trajectory studies. The trajectories are meant to be illustrative of the general situation during photochemical episodes in Europe, and they illustrate the photochemical production of ozone over 1-5 days (Derwent & Jenkins 1991). For a given VOC, POCP is calculated as the average of the results of the three scenarios. Most of the VOCs are oxidised more than 95% after 4-5 days, so that the calculated POCP represents the total ozone creation potential.

6.2.3 Definition of POCP

POCP is generally presented as a relative value where the amount of ozone produced from a certain VOC is divided by the amount of ozone produced from an equally large emission of ethene:

The unit of POCP is grams of ethene equivalents per gram of gas (g C₂H₄/g VOC). Ethene has been chosen as a reference gas as it is one of the most potent ozone precursors of all VOCs.

By definition, the calculated POCP values are not absolute values. POCP will be a function of the scenarios chosen, i.e. from one geographical area to another. As data for e.g. the chemical and photochemical reactions are often not known in great detail, their representation in the model will often be a compromise. Therefore, even for the same scenario, the POCP values can be calculated with higher precision when more accurate input data and more powerful computer tools are available.

The most commonly used POCP values have been calculated by Derwent and Jenkin (1991) and Derwent (1996), primarily from scenarios for the transport of ozone to the United Kingdom. IVL (Instituttet för Vatten- och Luftvårdsforskning) in Sweden has calculated values for POCP under Scandinavian conditions using lower NO_x concentrations and lower VOC emission rates. Because of the importance of the NO_x-level for the ozone formation, the low- NO_x values for POCP under Scandinavian conditions are found to be slightly different, but generally a good agreement with the high- NO_x values (as calculated by Derwent for UK) is found.

6.3 Substances contributing to the impact category

The principal precursors of tropospheric ozone are

• NO_x
• VOCs including CH₄
• CO

Reactions (I)-(III) govern the background level of ozone in the troposphere:

NO₂ + hn NO + O (I)
O + O₂ O₃ (II)
NO + O₃ NO₂ + O₂ (III)

If VOCs are also present, they are oxidised to produce peroxy radicals. Peroxy radicals can either consume NO or convert it to NO₂ and thus compete with ozone produced by reaction (II). Less ozone is thereby destroyed through reaction (III), and the ozone concentration will then increase.

The ozone creation potential for several VOCs is dependent on the NO_x concentration. In the calculation of a 1990 normalisation reference for POCP by Hauschild and Wenzel (1998), it was chosen not to differentiate between high-NO_x and low-NO_x values in the calculation of a normalisation reference for Europe. Because of the dependency on NO_x concentration, POCP values (and the calculated of the normalisation references) will differ between areas of high NO_x and low NO_x concentrations. For this reason, high-NO_x and low-NO_x values for POCP were introduced. High NO_x values are generally used for densely populated and highly industrialised areas, where high NO_x concentrations occur. The European Union was in this study defined as a high-NO_x area except for the Scandinavian countries and Ireland, as shown in Table 6-7.

6.4 Methodology

The calculation of normalisation references for 1994 has been carried out according to the methodology described in Wenzel et al. (1997). For photochemical ozone formation (POF), the normalisation reference for 1994 is calculated as

where Normref (POF)₁₉₉₄ is the normalisation reference for the area in question for the year 1994 [g C₂H₄-equivalents per person]

IP(POF)₁₉₉₄ is the impact potential for photochemical ozone formation for the area in question for the year 1994 [g C₂H₄-equivalents]

P₁₉₉₄ is the population of the area in question [persons]

The total impact potential IP (POF) from substances contributing to the impact category is calculated according to

where E_{n, 1994} is the total emission of substances from source category n to the atmosphere in 1994 for the area in question [kt per year]

POCP_n is the photochemical ozone creation potential for source category n [kt C₂H₄-equivalents per kt]

Because of the influence of NO_x on the formation of ozone from VOCs, the POCP values should be chosen according to the background concentration of NO_x in the area in question.

6.4.1 Normalisation reference for 1990

For the calculation of a 1990-normalisation reference for Denmark, Hauschild and Wenzel (1998) chose low-NO_x POCP values in the EDIP model. For the calculation of a 1985-normalisation reference for EU, high-NO_x POCP values was used. For practical reasons, the European Union was not separated into regions according to the background concentration of NO_x. The POCP values used have been calculated for each individual source category. These POCP values were calculated as weighted averages of individual POCPs for major nmVOC sources as shown in Table 6-2. A discussion of the methodology is given in section 6.6.

Table 6-2
Average POCP values (g C₂H₄/g VOC) for nmVOC mixtures from source
categories as used in the EDIP model for calculation of normalisation
references for Denmark in 1990 and for EU in 1985 (Hauschild & Wenzel
1998).

6.4.2 Emission data: CORINAIR

For the purpose of updating normalisation references to 1994 for European countries, data for emissions of VOCs, CO and CH₄ were taken from the CORINAIR 1994 database (Ritter 1997). In the CORINAIR 1994 database, national emissions are summarised in main source sectors according to main source sectors as shown in Table 6-3. These main source sectors differ slightly from the source categories used in the EDIP 1985 study for EU and 1990 study for Denmark.

The main source sectors of CORINAIR 1994 was defined as the source categories to be used in updating normalisation references for European countries. Because these source categories are not identical with the source categories used in the calculations of normalisation references for Denmark and EU, this means that the weighted averages for POCP should be carefully transferred from the previously calculated values. If transfer is not justifiable, new POCP values should be calculated from an analysis of the composition of the emissions of the individual subsectors. Table 6-3 shows the main source sectors of the CORINAIR 1994 database and the equivalent sources of nmVOC applied for use of POCP values.

Table 6-3
Main source sectors as defined in CORINAIR 1994 and source categories applied for calculation of POCP from nmVOC.

1 Road transport is included as emissions from petrol-engine cars (assumption in EDIP
1985/1990 inventory).

2 Best estimate for non-methane VOC from waste treatment and disposal.

3 luwc: land use and wood stock change

Table 6-4
Main source sectors used in the CORINAIR 1994 (Ritter 1997) emission inventory and the POCP values for nmVOC assigned to these source categories (Hauschild &Wenzel 1998).

1 Low value: road transport, Denmark; high value from petrol/diesel engine car, exhaust.

2 luwc: land use and wood stock change

6.5 Normalisation references

6.5.1 Denmark 1994

The calculation of a normalisation reference for Denmark is shown in detail in Appendix A. The calculated impact potential for nmVOC, CO and CH₄ is shown in Table 6-5.

Table 6-5
Impact potentials calculated for 1994 for nmVOC, CO and CH₄ for Denmark.

The normalisation reference for Denmark in 1994 is calculated in Table 6-6.

Table 6-6
Impact potentials calculated for 1990 and 1994 for nmVOC, CO and CH₄ for Denmark.

1 Hauschild and Wenzel (1998).

6.5.2 European Union 1994

The total impact potential IP(POF) from substances contributing to the impact category for a region like the EU having both areas of low-NO_x and high-NO_x background concentrations is calculated according to the formula

where E_{n, low-NOx, 1994} is the total emission of substances from source category n to the atmosphere in 1994 for the given area of low-NO_x background concentration [kt per year]

E_{n, high-NOx, 1994} is the total emission of substances from source category n to the atmosphere in 1994 for the given area having high-NO_x background concentration [kt per year]

POCPn, _low-NOx is the photochemical ozone creation potential for source category n in low-NO_x areas [kt C₂H₄-equivalents per kt]

POCP_{n, high-NOx} is the photochemical ozone creation potential for source category n in high-NO_x areas [kt C₂H₄-equivalents per kt]

For the area covered by EU-15, areas of high- and low-NO_x have been defined as shown in Table 6-7.

Table 6-7
Areas defined as low NO_x and high NO_x background concentration for calculation
of the impact potential in EU-15.

The calculation of a normalisation reference for EU-15 is shown in detail in Appendix B. The calculated impact potential for nmVOC, CO and CH₄ is shown in Table 6-8.

Table 6-8
Impact potentials calculated for 1994 for nmVOC, CO and CH₄ for EU-15.

The normalisation reference is calculated for EU-15 in 1994 as shown in Table 6-9.

Table 6-9
Impact potential calculated for nmVOC, CO and CH₄ for EU-15. Comparison
with 1985 values.

1 Hauschild and Wenzel (1998). Based on high-NO_x POCP values for the entire EU.

The low-NO_x countries contribute to the EU-15 normalisation with a higher POCP-emission per capita, as shown in Table 6-10.

There are three major source categories contributing to the total impact of nmVOC: Road transport, solvent and other product use, and agriculture and forestry. From Table 6-10, it is clear that the difference between the POCP emission per capita in the low-NO_x countries compared to the high-NO_x countries is due to differences in the CORINAIR 1994 emission inventory for nmVOC for the source category "Agriculture and Forestry". This is probably due the difficulties in estimating nmVOC emissions from this category.

Table 6-10
Comparison between normalisation references for low-NO_x and high NO_x countries in EU-15, with and without nmVOC emissions from source category "Agriculture and forestry".

6.5.3 World 1990

One of the aims of this project has been to calculate a global normalisation reference for 1994. A methodology for the calculation of such a "world proxy" is proposed in chapter 3, Development of normalisation references for different geographic areas. This methodology applies to the most frequent situation for impact categories: global emission data cannot be found in an acceptable form or quality.

In the case of POCP, anthropogenic emissions of the key substances (nmVOC, methane and CO) can be found for 1990 in the EDGAR database (Olivier et al. 1996). RIVM has not yet any published global emission data for 1994. Given the data from 1990, it was decided to calculate a normalisation factor for 1990 instead of 1994. Because of the undoubtedly larger uncertainties associated with the calculation of the global emissions from GDP, it is suggested that a 1994 normalisation reference for POCP is calculated when an update of the EDGAR database is at hand.

Appendix C shows the calculation of POCP emissions for the geographical areas as defined in EDGAR. For nmVOC, a division between high-NO_x and low-NO_x areas is done. High-NO_x areas are in Appendix C marked as "HN" (USA, Western Europe, Eastern Europe, Japan) and low-NO_x are marked as "LN" (Canada, Latin America, Africa, Soviet Union (CIS), Middle East, India, China, East Asia, Oceanic areas, Int. shipping). Table 6-11 shows the result of the calculation for POCP in kt C₂H₄ total and the calculated normalisation reference for 1990.

Table 6-11
Calculated normalisation reference for POCP for the world, 1990.

Table 6-11 shows a very large contribution to the worldwide effect potential from CO emissions. The global CO emission per person calculated from the emissions in the EDGAR database is relatively large compared to the CO emission per person in the CORINAIR database. As low temperature combustion is assumed to be the major source of CO emissions, the reason for this deviation could be the widespread use of uncontrolled combustion in households etc. in the third world. A further assessment of the emission factors for CO used in the EDGAR database has not been pursued.

By using the extrapolation method proposed in chapter 3 the worldwide normalisation reference for photochemical ozone formation is estimated to 20 kg C₂H₄-eq./capita/year. The normalisation reference found by using the extrapolation method is slightly lower than the value found by using the 1990 emission inventory.

6.5.4 Comparison of updated normalisation references with previously used normalisation references

In Table 6-12 the normalisation references for photochemical ozone formation for 1994 estimated in this study are compared with the previously estimated normalisation references (Hauschild & Wenzel 1998).

Table 6-12
Summary of normalisation references for photochemical ozone formation estimated in the present study compared with previously normalisation references.

1 Hauschild and Wenzel (1998).

2 1985.

3 1990.

The normalisation reference for Denmark is approximately the same as the normalisation reference for 1990 whereas the normalisation reference for EU-15 has increased by 25% from 1985 to 1994. The worldwide normalisation reference has been estimated to 22 kg C₂H₄-eq./capita/year based on a worldwide inventory for nmVOC, CO and CH₄.

6.6 Recommendations for future updating

6.6.1 Quality of emission data

Any updating of normalisation references should focus on uncertainties of the data used from emission inventories. The CORINAIR database used in this study is prepared by the European Environment Agency and should as such be considered as the most validated database for European Union. The UN-ECE database is also an option as a considerable number of countries have ratified the Convention on Long-range Transboundary Air Pollution (UN-ECE, 1979) and the protocol on VOC (UN-ECE, 1991). As apparent from the differences found in section 6.5.2 between POCP (nmVOC) emissions per capita for Scandinavian countries compared to the rest of Europe, the countries reporting nmVOC data to CORINAIR have encountered difficulties in assessing the emissions from source category no. 10 (agriculture and forestry). A future updating should generally focus on the quality of the emission data used, and a special attention should be paid to data from the source category "agriculture and forestry".

Considerable discrepancy was found between emission data from the European countries in the CORINAIR 1994 database and in the UN-ECE database ^[8] for 1994. This is especially the case for methane. The reason for these discrepancies has not been found.

6.6.2 NO_x inclusion/exclusion?

POCP is a parameter developed to express the potential of VOCs to create ground level ozone, and emissions of NO_x will thus not be included in the impact assessment for photochemical ozone formation in the EDIP model. The impact of NO_x is in EDIP present through the impact of acidification, nutrient enrichment and human toxicology. However, model calculations have shown that reductions of emissions of NO_x will in many cases have a larger effect than VOC reductions on the reduction of the ozone concentration levels (Fenger 1995). Hauschild and Wenzel (1998) has chosen POCP as the best existing and the most internationally accepted model for the calculation of photochemical ozone potential for emissions of individual substances, and NO_x can not be assigned as a separate factor within this model. The use of POCP as a characterisation factor is still considered as the best internationally accepted method available. The error introduced hereby should however be evaluated.

To give the highest possible ozone reduction, it is necessary not only to reduce VOC and CO emissions, but also NO_x emissions. As the impact of NO_x on ozone production is not included in EDIP, a false impact from this characterisation factor could be expressed in the assessment of a product system by EDIP. The magnitude of this error will be site dependent.

6.6.3 Focus on most potent individual VOCs?

McBride et al. (1997) has suggested that a more efficient ozone reduction could be achieved if the emission reductions focused on the most potent ozone producers instead of reducing all VOCs regardless of the species. In EDIP, the VOCs are partly divided, not according to their ozone creation potential, but according to source categories. For a more efficient abatement strategy, it might be useful to subdivide the VOCs according to their individual ozone creation potential. For practical reasons, this could be done for a reduced number of highly potential VOCs. E.g., at present solvent using industries will have the same POCP factor (0.4-0.5 as shown in Table 6-2) regardless of the potency of the individual compound in the present EDIP model. During the assessment of a specific product, individual factors applied to individual solvents will give a more realistic impact assessment of the specific tropospheric ozone formation.

6.6.4 Simplified POCP factor for nmVOC?

IVL has in a recent study concluded that POCP ranges might be more useful to introduce in stead of site-specific POCP values (i.e. high-NO_x vs. low NO_x) (Altensted & Pleijel 1998). Except for production processes, all POCP factors in low and high NO_x areas vary between 0,4-0,6 kg C₂H₄/kg for nmVOC. nmVOC from production processes is allocated a POCP factor of 0,3 kg C₂H₄/kg. In the light of the uncertainties involved in the calculation of the POCP factors and the uncertainties of the emission data, it should be considered whether one weighted factor for POCP can be used for both high and low-NO_x areas. This would simplify calculations considerably and save resources in the elaboration of normalisation references for the use in LCA.

6.6.5 Exclusion of methane in future POCP calculations?

In the light of the relatively small contribution from methane to the total impact from POCP, it should be considered whether methane should be excluded in future POCP calculations.

6.6.6 Worldwide

The calculated worldwide normalisation reference for 1990 is relatively high as compared to the normalisation references for EU and Denmark. This is due to the relatively high global CO emissions per capita derived from the EDGAR database. The uncertainties of the emission data in the EDGAR database should be compared to the uncertainties of the CORINAIR database.

The POCP factors used to calculate the worldwide normalisation reference are derived from European POCP model calculations and divided into low-NO_x and high-NO_x areas. POCP factors in tropical areas might deviate considerably from the European POCP values meteorological factors (i.e. sunlight) influence the formation of ozone. This potential source of error should be investigated in the future when calculating a worldwide normalisation reference for POPC.

6.7 References

Altenstedt, J. & Pleijel, K. 1998, POCP for individual VOC under European conditions. IVL Report B-1305, Swedish Environmental Research Institute, Stockholm.

Carter, W.P.L., Pierce, J.A., Luo, D. & Malkina, I.L. 1995, Environmental Chamber Study of maximum incremental reactivities of volatile organic compounds. Atmospheric Environment, 29 (18) pp. 2499-2511.

Derwent, R.G. 1996, Photochemical ozone creation potentials for a large number of reactive hydrocarbons under European conditions. Atmospheric Environment, 30 (2) pp. 181-199.

Derwent, R.G. & Jenkin, M.E. 1991, Hydrocarbons and the long-range transport of ozone and PAN across Europe. Atmospheric Environment, 25 (8) pp. 1661-1678.

Fenger, J. 1995, Ozon som luftforurening. DMU Tema-rapport 1995/3.

Hauschild, M. & Wenzel, H. 1998, Photochemical ozone formation as a criterion in the environmental assessment of products in Environmental assessment of products. Volume 2 Scientific background eds. Hauschild. M. & Wenzel. H. London: Chapman & Hall.

McBride, S.J., Oravetz, M.A. & Russel, A.G. 1997, Cost-benefit and uncer.tainty issues in using organic reactivity to regulate urban ozone. Environm. Sci. Technol. 31 (5), pp. 138A-244A.

Olivier, J.G.J, Bouwman, A.F., van der Maas, C.W.M., Berdewski, J.J.M., Veldt, C., Bloos, J.P.J., Visschedijk, A.J.H., Zandveld, P.Y.J. & Haverlag, J.L. 1996, Description of EDGAR version 2.0: A set of global emission inventories of greenhouse gasses and ozone-depleting substances for all anthropogenic and most natural sources on a per country basis and on 1ox1o grid. RIVM report nr. 771060 002/TNO-MEP report nr. R96/119.

Ritter, M. 1997, CORINAIR 94 - Summary Report - European Emission Inventory for Air Pollutants. Copenhagen: European Environment Agency.

UN-ECE 1979, Convention on Long-range Transboundary Air Pollution. United Nations, Economic Commission for Europe. Available: http//:www.unece.org.

UN-ECE 1991, Protocol to the 1979 Convention on Long-range Transboundary Air Pollution concerning the control of emissions of volatile organic compounds or their transboundary fluxes. United Nations, Economic Commission for Europe. Available: http//:www.unece.org.

Wenzel, H., Hauschild, M. & Alting, L. 1997, Environmental Assessment of Products, Vol. 1: Methodology, tools and case studies in product develop.ment. London: Chapman & Hall.

Appendix A.1: Calculation of normalisation reference for Denmark.

Total POCP calculated from anthropogenic NMVOC-, CO and CH₄-emissions from Denmark.

Appendix A.2: Calculation of normalisation reference for EU-15

Click here to see the Table.

1 CORINAIR 94 Summary report, Report from the European Topic Centre on Air Pollution (1997)

2 Low NO_x countries include Denmark, Sweden, Finland and Ireland

3 75% of the CH₄ emissions listed arise from agriculture and forestry

4 nmVOC contribution from luwc is 21 kg nmVOC/person/year in low-NO_x countries, and in high-NO_x countries only 6,8 kg nmVOC/person/year.

Appendix A.3: Calculation of a worldwide normalisation reference

Click here to see the Table.

Footnotes

[8] The UN-ECE database is available at http://www.unece.org.

| Front page | | Contents | | Previous | | Next | | Top |

Version 1.0 April 2005, © Danish Environmental Protection Agency