Non-CO2 gases from manufacturing (HFCs, PFCs, SF6, and N2O) are increasingand. Furthermore, PFCs and SF6 have extremely long atmospheric lifetimes (thousands of years) and GWP values (thousands of times those of CO2) resulting in virtually irreversible atmospheric impacts. Fortunately, there are technically-feasible, low cost emission reduction options available for a number of applications. Since the SAR, implementation of major technological advances have led to significant emission reductions of N2O and the fluorinated greenhouse gases produced as unintended by-products. For the case of fluorinated gases being used as working fluids or process gases, process changes, improved containment and recovery, and use of alternative compounds and technologies have been adopted. On-going research and development efforts are expected to further expand emission reduction options. Energy efficiency improvements are also being achieved in some refrigeration and foam insulation applications, which use fluorinated gases. Emission reduction options by sector are highlighted below. The Chapter 3 Appendix reviews use and emissions of HFCs and PFCs being used as substitutes for ozone-depleting substances.
Adipic acid production. Various techniques, like thermal and catalytic destruction, are available to reduce emissions of N2O by 90% 98% (Reimer et al., 2000). Reimer et al. (2000) report costs of catalytic destruction to be between US$20 and US$60/tN2O, which is less than US$1/tCeq. Costs of thermal destruction in boilers are even lower. The inter-industry group of five major adipic acid manufacturers worldwide in 1991 to 1993 have agreed on information exchange and on a substantial emission cut before the year 2000. These major producers probably will have reduced their joint emissions by 91%. It is estimated that emissions from the 24 plants producing adipic acid worldwide will be reduced by 62% in the year 2000 compared to 1990 (Reimer et al., 2000).
Nitric acid production. Concentrations of N2O in nitric acid production off-gases are lower than in the case of adipic acid production. Catalytic destruction seems to be the most promising option for emission reduction. Catalysts for this purpose are under development in a few places in the world. Oonk and Schöffel (1999) estimate that emissions can be reduced to a large extent at costs between US$2 and US$10/tCeq.
The smelting process entails electrolytic reduction of alumina (Al2O3) to produce aluminium (Al). The smelter pot contains alumina dissolved in an electrolyte, which mainly consists of molten cryolite (Na3AlF6). Normal smelting is interrupted by an anode effect that is triggered when alumina concentrations drop; excess voltages between the anode and alumina bath result in the formation of PFCs (CF4 and C2F6) from carbon in the anode and fluorine in the cryolite (Huglen and Kvande, 1994; Cook, 1995; Kimmerle and Potvin, 1997). Several processes for primary aluminium production are in use, with specific emissions ranging from typically 0.15 to 1.34 kg CF4 per tonne Al16 depending on type of technology (determined by anode type and alumina feeding technology) (IAI, 2000). Measurements made at smelters with the best available technology (point feed prebake) indicate an emissions rate as low as 0.006 kg CF4 per tonne Al (Marks et al., 2000). Worldwide average emissions for 1995 are estimated to range from 0.26 to 0.77 kg CF4 per tonne Al (Harnisch et al., 1998; IEA, 2000). Manufacturers have carried out two surveys on the occurrence of anode effects and associated PFC-emissions (IPAI, 1996; IAI, 2000). Based on 60% coverage of world production (no data on Russia and China) they estimated a mean emission value of 0.3 kg CF4 per tonne Al in 1997. Emission reductions were achieved from 1990 to 1995 by conversion to newer technologies, retrofitting existing plants, and improved plant operation. Industry-government partnerships also played a significant role in reducing PFC emissions. As of November 1998, 10 countries (which accounted for 50% of global aluminium production in 1998) have undertaken industry-government initiatives to reduce PFC emissions from primary aluminium production (US EPA, 1999d). It has been estimated that emissions could be further reduced via equipment retrofits, such as the addition or improvement of computer control systems (a minor retrofit) and the conversion to point-feed systems (a major retrofit). One study estimated 1995 emissions could be reduced an additional 10%50% (depending on technology type and region) with maximum costs ranging from US$110/tCO2eq for a minor retrofit to nearly US$1100/tCO2eq for a major retrofit (IEA, 2000). A second study estimates that 1995 emissions could be reduced by 40% at costs lower than US$30/tCeq, by 65% at costs lower than US$100/tCeq and by 85% at costs lower than US$300/tCeq (Harnisch et al., 1998; 15% discount rate, 10 year amortization).
The development of an inert, non-carbon anode is being pursued through governmental and industrial research and development efforts. A non-carbon anode would remove the source of carbon for PFC generation, thereby eliminating PFC emissions (AA, 1998). A commercially viable design is expected by 2020.
The semiconductor industry uses HFC-23, CF4, C2F6, C3F8, c-C4F8, SF6 and NF3 in two production processes: plasma etching thin films (etch) and plasma cleaning chemical vapour deposition (CVD) tool chambers. These chemicals are critical to current manufacturing methods because they possess unique characteristics when used in a plasma that currently cannot be duplicated by alternatives. The industrys technical reliance on high GWP chemicals is increasing as a consequence of growing demand for semiconductor devices (15% average annual growth), and ever-increasing complexity of semiconductor devices.
Baseline processes consume from 15%-60% of influent PFCs depending on the chemical used and the process application (etch or CVD). PFC emissions, however, vary depending on a number of factors: gas used, type/brand of equipment used, company-specific process parameters, number of PFC-using steps in a production process, generation of PFC by-product chemicals, and whether abatement equipment has been implemented. Semiconductor product types, manufacturing processes, and, consequently, emissions vary significantly across worldwide semiconductor fabrication facilities.
PFC use by the semiconductor industry began in the early 1990s. Global emissions from semiconductor manufacturing have been estimated at 4 MtCeq in 1995 (Harnisch et al., 1998). Options for reducing PFC emissions from semiconductor manufacture include process optimization, alternative chemicals, recovery and/or recycling, and effluent abatement. A number of emission reduction options are now commercially available. For plasma-enhanced CVD chamber cleans, switching to PFCs that are more fully dissociated in the plasma or installing reactive fluorine generators upstream of the chamber is favoured. For etch tools, PFC abatement is currently available (Worth, 2000). However, the size of wafers being processed and the design and age of the fabrication facility have a major impact on the applicability of PFC emission reduction technology. A recent study for the EU (Harnisch and Hendriks, 2000) estimated that 60% of projected emissions from this sector could be abated through the use of NF3 in chamber cleaning at US$110/tCeq. According to the same study another 10% are available through alternative etch chemistry at no costs and about 20% through oxidation of exhausts from etch chambers at US$330/tCeq. The remaining emissions from existing systems are assumed to be currently virtually unabatable.
Through the World Semiconductor Council, semiconductor manufacturers in the EU, Japan, Korea, Taiwan (China), and the USA have set a voluntary emission reduction target to lower PFC emissions by at least 10% by 2010 from 1995 (1997 for Korea and 1997/1999 average for Taiwan (China) baselines (World Semiconductor Council, 1999). Members of the World Semiconductor Council represent over 90% of global semiconductor manufacture.
Other reports in this collection