are magnified where glass production is measured in a variety of units.
In the chemical and metal industries, reliable production data are available for most countries, so the emission factors present the greatest source of uncertainty, particularly for iron and steel production. For ammonia production, CO2 emissions can be estimated using national-level data on ammonia production (or, less preferably, ammonia production capacity) and default values for the quantity of fuel (typically natural gas) required as feedstock per unit of output, the carbon content of the fuel, and the carbon oxidation factor. Any CO2 recovered for purposes of urea production is also accounted for. For iron and steel production, the CO2 emissions are estimated by applying the appropriate emission factors to national statistics on the amount of steel produced by each method and the total amount of pig iron produced that is not processed into steel. Similarly, the estimation of CO2 emissions in aluminum production requires national-level production data by process type (i.e., Søderberg or Prebake) to which the appropriate default emission factor can then be applied.
Hydrofluorocarbons (HFCs) comprise about 18 percent of total emissions (in terms of CO2 equivalents) from the IPPU sector for Annex I countries.1 The use of HFCs as substitutes for ozone-depleting substances in a variety of industrial applications is by far the largest source of HFC emissions, accounting for about 86 percent of total emissions from the sector, and their usage is growing rapidly. A smaller, but significant source of HFC emissions is the generation of trifluoromethane (HFC-23) as a by-product during the production of chlorodifluoromethane (HCFC-22).
Actual emissions of HFCs are estimated using either an emission-factor or a mass-balance approach (IPCC, 2006). Both methods can use activity data collected at either the application level (e.g., refrigeration) or the subapplication level (e.g., equipment or product type); the latter is expected to yield higher-accuracy estimates. For the emission-factor approach, HFC emissions are calculated by determining the net consumption of a chemical in a specific application or subapplication (production plus imports minus exports minus destruction of the chemical) and then applying an emission factor(s) to the net consumption. For the mass-balance approach, emissions are estimated as the sum of the sales of a chemical and, for equipment containing this chemical, the total charge of retired equipment minus the total charge of new equipment. The major source of uncertainty in national estimates of HFC emissions is the lack of activity data on chemical production or sales in countries where suppliers treat the information as confidential. This barrier to the production of reliable national estimates is being reduced with the development of regional and global databases of ozone-depleting substances. For example, databases that track the phase-out of ozone-depleting substances are directly relevant for estimating the phase-in of HFC substitutes (IPCC, 2006).
Emissions of HFC-23 can be calculated by applying a default emission factor to the quantity of HCFC-22 produced. Given the known variability in emissions from different HCFC-22 manufacturing facilities, the uncertainty in the emission factor far outweighs the uncertainty in the activity data (IPCC, 2006).
Emissions of nitrous oxide (N2O) from nitric acid and adipic acid production comprise about 7 percent of total emissions from the IPPU sector in Annex I countries.2 Nitric acid production emits N2O as a byproduct during the catalytic oxidation of ammonia, and adipic acid production (most of which takes place in a few plants in the United States and Europe) generates N2O as a by-product during a process involving the oxidation of nitric acid. Emissions of N2O from both sources can be estimated by multiplying production by a default emission factor. For nitric acid production, the major source of uncertainty in N2O emissions is the activity data. Nitric acid production is often underestimated because nitric acid is formed as part of a larger production process and is never sold on the market. For adipic acid production, neither the default emission factor nor the activity data are significant sources of uncertainty (IPCC, 2006). The default emission factor