Ammonia Reductions and Costs Implied By the Three Ambition Levels Proposed in the Draft Annex IX to the Gothenburg Protocol

The Convention on Long-range Transboundary Air Pollution has started negotiations on the revision of its Gothenburg multi-pollutant/multi-effect protocol. Among other topics, emissions of ammonia have been subject of specific scrutiny. To inform negotiations, the Task Force on Reactive Nitrogen (TFRN) has compiled a list of potential ammonia emission control measures that could form an annex to the protocol. This "Draft Annex IX" defines three ambition levels, ranging from the most stringent level "A" to the least ambitious level "C". This report presents estimates of the ammonia emission reductions and costs of the measures that are included in the Draft Annex IX prepared for the revision of the Gothenburg Protocol. The estimates for the three different ambition levels have been derived with the GAINS model, and it is assumed that these measures are equally implemented for all countries, even if they are not cost-effective. The specific assumptions in translating the specifications provided in the Draft Annex IX into GAINS input data are described in this paper. Costs for implementing the ambition levels vary across countries. Taking into account recent information on ammonia emission control costs that has been compiled by the Task Force on Reactive Nitrogen it is found that overall costs of the proposed measures are generally low. Per unit of emissions abated, costs are typically only up to 1 Euro per kg NH3-N abated with a few outliers which, however, stay below 5 Euro per kg NH3-N abated. Total abatement costs are modest, for the most ambitious scenario A they amount to 8/1000 of 1% of the GDP in 2020. In relative terms (e.g., expressed as percentage of GDP) costs are higher in the non-EU countries than in the EU-27. It is found that the specific measures, if uniformly implemented in all countries, would be less cost-effective than the (country-specific) least-cost set of measures that are derived from the GAINS optimization, e.g., for the MID and LOW cases in CIAM 4/2011 report, where the model can choose from a wider set of options.