with less capital per unit capacity, with less energy consumption and emissions, in facilities of lesser size and complexity such that "smaller operators" than large aluminum companies can and will enter the business, and with a higher output per employee. Expanding on employment, a reverbatory furnace operation has an output of 595 t Al/employee year "without casting facilities", an output of 560 t/employee year "with casting" and a rotary barrel operation 1250 t/employee year. On a weighted average this yields about 800 t/employee year, for the secondary industry. In Table 9 and 10 this indicates the following expansions in employment amongst regions over the period 1996 to 2012 (rounded off to the nearest 100). There are no revolutionary processes at a stage of development that could displace the Hall-Heroult process before 2015. One process, carbothermic reduction of Al2O,, which would use less electrical energy, with cogeneration, 10.2 AC kwh/kg Al (34% less than the component for present USA, and 24% less than the best electrolytic smelters in the world) has been demonstrated at a 150 kVA stage only. It is the direct carbothermic reduction of Bayer alumina in a staged electric arc furnace at 1650 to 1850°C. Further development requires the resolution of several factors including the continual management of the descending charge of briquettes (coke and alumina) against a condensing flux of vapors of aluminum and aluminum suboxides, and the longevity of the materials of construction in such demanding conditions. Preliminary designs of a plant showed reactors not unlike blast furnaces rated at 40-50 megawatts per unit from which liquid Al was tapped. There is a downside to this potential process in terms of mitigation of global warming gases. The overall chemical reaction is Thus, for every tonne of Al produced at 100% efficiency, 2.45 tonnes of CO2 are produced. which is double that for the Hall-Heroult process. Conceding the cogeneration there would be an extra lt of CO2/t Al in this process. No company is expressing interest in this process now, partly because of satisfaction and familiarity with the prevailing technology, partly due to the suspected $250 million it would take to demonstrate the direct carbothermic process and the subsequent uncertainty that it would be adopted over the Hall-Heroult process. There are two emerging semirevolutionary modifications to the Hall-Heroult process that promise to reduce unit energy consumption. The first, would benefit older (50 to 160 kA prebaked anode cells) by reducing unit energy by 10 to 20%. The technical and economic feasibility of Refractory Hard Metal (RHM), wetted cathodes into prebaked cells is under way, partially funded by the Metals Initiative Act by Reynolds Metals, Great Lakes Carbon Research (GLCR) and Kaiser Aluminum Corporation. Reports are that the evaluation of the material, a solid planar surface of titanium diboride-graphite (TiB,) composite which becomes the active cathode, (replacing the mobile, liquid aluminum) is progressing in industrial 68 kA cells. Conceding satisfactory results from this first industrial trial, favorable economics from the projected lifetimes, and cost of retrofitting TiB,, the elapsed time for an expanded plant demonstration, actual conversion of a hotline and beginning acceptance by the industry will be at least 8 years. This cathode retrofit which has the potential for decreasing the interelectrode distance from about 4.5 cm to 3 cm, is compatible with conventional carbon anodes. The voltage across a cell could be reduced from 4.5 to 3.8 volts. Laboratory and some tentative pilot scale trials have been made of a novel anode material-a composite of metal and metal oxides with special properties-that would be formed into shapes that would not be consumed (as is carbon) during the electrolytic decomposition of Al2O3. These materials are dubbed "dimensionally stable anodes." Oxygen is evolved rather than CO2 so if transitioned, their use would reduce global warming gas emissions (CO2) by 1.5 t/t Al. However, the decomposition potential is increased from 1.2 to 2.2 volts, with some slight decrease in overvoltage component so that overall, unless coupled with the RHM cathodes above, only minor change in the electrical power might be expected. However, all the thermal energy for making carbon anodes with its attendant emissions would not be needed. So far, these anodes have not proved adequately robust and have contaminated the Al product. All this devolves to the practical situation that there is no substitute for the present electrolytic process. By the ongoing combination of applying increased knowledge, advanced material, automated controls and enhanced human skills, there will be a gradual improvement in unit energy and production to the extent where thermal efficiencies may maximize at 50% compared with best available technology at 47%. Conclusions Under scenarios for fuel increases with power prorated specifically for the primary Al industry, the increased costs in production of primary aluminum are likely to cause shutdowns of some capacities in USA and OECD-Europe. The prevailing advantages in costs for the USA over OECD-Europe would continue but to a lesser extent. Cumulative lost profits would be about $12 billion for both regions. Japan, with a capability for producing value added aluminum product comparable to that of USA and OECD, would be unaffected by the scenarios proposed, because, for some decades, that nation has engaged in joint venture production of aluminum offshore in developing countries. Under scenarios for fuel increases with adder for power fully applied the projected economics made the entire USA capacity non-competitive by 2010 and it is possible the 4000Kt would shut down leading to the loss of 23,000 jobs. OECD-Europe, would be similarly impacted, but that capacity in Canada, Norway and Iceland could well remain operational. The likely closure of 4300Kt capacity would mean loss of 18,000 jobs. These projections assume the future supply/demand balance allows the prevailing LME 1994$ to persist. It is unlikely under any scenario, with or without any greenhouse gas limitation policies, that any new primary aluminum plants will be built in the USA. This also applies to Western Europe, except Norway and Iceland. Other countries within OECD where plants may be built include: Canada, Australia and New Zealand where hydroelectric or moderately priced thermal power is/could be available. The major proportion of new aluminum smelters will be distributed amongst Latin America, Africa, OPEC countries, and Asia, conceding power stations coming onstream there. If prevailing projections for the growth of market for aluminum products around the world are realized and new plants are not in production by 2005, then there is the potential for a shortage of primary Al which would adversely affect the competitiveness of aluminum and its derivative products world wide and in particular, the developed nations. Even with additional primary aluminum capacity around the world, there will be an increasing proportion of aluminum obtained from recycling. This sector of the industry is likely to grow at a much greater rate than primary aluminum; will consume less energy per unit production, and with modest capital, create new employment in all regions. While there are semirevolutionary and technological improvements envisioned and under development for decreasing the unit energy required for primary aluminum extracted by the HallHeroult process, radical changes are not likely to be transitioned by 2010. Retrofits and modernizations will lead to small decreases in energy and increases in capacity. |