TMS QUICK LINKS:
  • TMS ONLINE
  • MEMBERS ONLY
  • MEMBERSHIP INFO
  • MEETINGS CALENDAR
  • PUBLICATIONS

JOM QUICK LINKS:
  • JOM HOME PAGE
  • MORE HTML ARTICLES
  • COMPLETE ISSUES
  • BOOKS REVIEWS
  • SUBSCRIBE

In connection with the 125th anniversary of the Hall–Héroult process this year, we will review the most important progress that has been made in the twentieth century. What were the most significant improvements in this period, and which scientists and engineers came up with the ideas for these improvements? In this paper we will try to answer these questions. We will highlight the major technological breakthroughs and mention those people who played important roles in the development of these improvements.

This year is the 125th anniversary of the invention of the industrial aluminum electrolysis process. The first 20 to 30 years after 1886 were characterized by many technological improvements in the process, but we will start our review from 1914, the year when both the inventors Charles Hall and Paul Héroult passed away (Figure 1).

First, we must conclude that this is a very good method, because it has survived the many attempts that have been made to develop a viable alternative method for production of aluminum. Throughout these years aluminum production has developed from "art" to "science." A steadily increased understanding of the process has been achieved thanks to extensive research and development work, both in aluminum plants and in universities and academic institutions around the world.

Breakthrough inventions in aluminum smelting during the past 125 years have resulted in amazing gains in cell performance, as demonstrated by the jump in amperage and aluminum production in modern cells, (+2,249 kg Al per cell day) compared with the production per cell in 1914, as well as a 60% decrease in the specific energy consumption (Table I).

The Søderberg Anode
The first big improvement, named after the Norwegian inventor, is the Søderberg anode. Carl Wilhelm Søderberg was born in Sweden, but moved to Norway with his parents as a small child (Figure 2).

The Søderberg anode was patented in 1918 and it has been used in the aluminum industry since 1923. By this time prebaked anode pots had been in service for almost 40 years. What makes the Søderberg electrode unique is that it is continuous, self-baking, and monolithic. The lower part of the anode reacts during the electrolysis process, and in the upper part addition of "green" anode paste briquettes gradually replaces the anode material that is consumed at the bottom surface. The heat from the electrolyte gives the bottom of the anode the right baked consistency.

The studs in the Søderberg anode are usually placed vertically, but also horizontal stud Søderberg pots have been developed. Even now there are still some horizontal stub Søderberg potlines in operation. The so-called "Erftwerk" pots were developed in the 1950s by Vereinigte Aluminium Werke (VAW). The Elbewerk smelter in Germany started up in 1972 and was operating until 2006 at about 130 kA. These pots had continuous prebaked anodes, which still seems to be a good idea. However, these potlines are now closed for good.

The main advantages of vertical stud Søderberg pots are that they save the capital, labor, and energy required to manufacture prebaked a nodes. These pots have some inherent disadvantages, however, compared to prebake pots. Pot voltage and energy consumption are higher for Søderberg pots, current efficiency is lower, anode quality is lower, and emissions of fluorides and polycyclic aromatic hydrocarbons (PAH) are higher. Also the pot size is smaller, especially compared to modern prebake pots.

A breakthrough improvement of Søderberg pots came in the late 1970s, when the Sumitomo aluminum company marketed and sold their Søderberg pot technology. This mainly consisted of "dry" anode paste with lower pitch content and introduction of bar breakers or point feeders for alumina addition. These improvements lowered the PAH emissions and made feeding the pot easier. However, in the beginning there were a lot of operational problems for those smelters that chose to implement this Sumitomo technology.

More recently, successful and valuable improvements have been reported in Søderberg pot design and operation in some countries, mainly in Norway and Russia. Measures were taken at Elkem's plant in Lista, Norway (A.K. Syrdal and T.B. Pedersen) in the 1990s to improve current efficiency and energy consumption, as well as the environmental performance by better alumina feeding technology (point feeding) to reduce the frequency of anode effects and the greenhouse gas emissions, and by introducing a closed anode top to nearly eliminate the PAH emissions (Figure 3).¹ In the Krasnoyarsk plant in Russia different methods for hooding and sealing of the pots have been developed, and a colloidal anode with pitch content close to that of a prebaked anode, has been tested.² This "breakthrough" technology was directed by Victor Mann and Vladimir Frizorger (project creators and leaders) and conducted by Mikhail Krak, Nikolai Tonkih and Matei Golubev (carbon technology managers) at UC RUSAL's Engineering and Technology Center (ETC) in Krasnoyarsk.

Thus, it is too early now to conclude that "The Søderberg Era" is completely over. As a curiosity it can be mentioned that in spite of his success with the Søderberg electrode, it was building of violins that was Carl Søderberg's great interest in life! He built about 30 of these instruments.

Point Feeding of Alumina
The next big development to occur was the use of point feeders in the early 1960s at Alcoa. The point feeder development did not start as a process improvement effort but instead it was strictly a labor saving opportunity. Prior to this development all the alumina was fed manually in rather large quantities (~100 kg) several times per day. In the end the process improvement advantages greatly outweighed the labor savings.

In 1961 one of the Alcoa smelters in the United States (Rockdale, Texas) used point feeders that are still in common use without any design changes today.³ This feeder was developed in the Alcoa Equipment Development Division (AEDD) at Alcoa Laboratories in New Kensington, Pennsylvania as a collaborative effort and resulted from the unsuccessful effort by several individual plants to make a labor saving feeder. The first of the early feeder designs dates back to 1958 and some are still running at the Alcoa Wenatchee plant (Figure 4).

These early feeders were not as reliable as the AEDD feeder (Figure 5). The key person in the point feeder development was Dick Taylor. However, nobody outside of Alcoa has ever heard of him until now!

All modern pots now have point feeders. The method consists of punching small holes in the crust at two to six positions (usually) along the center line of the pot. The feeding is done with single piercing rods, between six and ten centimeters in diameter. These rods are mounted at the end of fast-acting pressurized air cylinders. The great advantage is that small quantities of alumina are added to the electrolyte at each break-and-feed. This method generates minimal sludge formation in the center area of the pot, and there are minimal emissions of dust and fluorides during the break-and-feed operations.

The point feeder represented a real breakthrough, also literally speaking, in the alumina feeding technology. One may safely say that the point feeding technique is one of the most important inventions in the Hall–Héroult pot technology in the last century. Although there are some detail differences in how the various companies designed each of their point feeders (just as there are design differences in the pots), the basic principle of this technology has remained unchanged since its inception, much the same as the Hall–Héroult process.

Gas Cleaning by Dry Scrubbing
The third main improvement of the last century was treatment of the pot fume with dry scrubbers. The development was motivated by the need to protect the environment to a greater extent than could be obtained by the wet scrubbing technique that was being used. The dry scrubbing process is now used in almost all aluminum smelters in the world. One great advantage of dry scrubbers compared to the older wet scrubbers is that they use the raw material alumina as the sorbent for removal of gaseous and particulate fluorides from the anode gases. The fluorides are chemisorbed on the surface of the alumina particles, which are then called secondary alumina. This material is stored in large silos and is later used as feed material to the cells. This means recycling of the captured fluorides and it thereby reduces the overall fluoride consumption significantly. Environmentally, the dry scrubber process has been instrumental in reducing the fluoride emissions from aluminum plants.

The process was developed by Alcoa (alumina fluid bed technology) and Alcan- ÅSV (alumina injection scrubbers) in Norway in the late 1960s. The earliest recorded report on this development at Alcoa was February 23, 1965—exactly 79 years after Hall's invention. This was a collaborative effort done in Alcoa's Physical Chemistry Division at Alcoa Technical Center. In terms of the development of the dry scrubbers there is a dual credit. Lester Knapp and Norman Cochran were the key persons in this development for Alcoa. The first commercial injection type dry scrubber system was installed by ABB Flakt at the VSS smelter Granges Aluminium, Sundsvall, Sweden in 1972 with separate reactors followed by cyclones and bag filter. The first prebake injection type (alumina injection into the branch duct leading into each filter compartment) dry scrubber system was installed in 1973 by ABB Fläkt (Erik Monkerud was project manager) at the HAW smelter in Hamburg, Germany. Figure 6 shows the Alcoa 398 reactor at Badin from 1971. They were mixing pot gas with alumina, there were integrated filters and they recycled the material back to the electrolysis cells. Nearly all new aluminum smelters today are built with dry scrubbers using alumina injection technology.

The major sources of fluoride emissions into the working atmosphere and to the potroom roofline are from pots with open hoods and from hot anode butts pulled from pots during the anode change operation. Two recent breakthrough inventions have to a large degree solved these environmental problems. The first was the development of on-demand dual duct suction systems to nearly double extraction flow rate during active pot work, and the second was the development of anode cooling boxes that collect the HF emissions from hot bath and anode butts. These two inventions have resulted in a sharp reduction in HF emissions to the potrooms roofline.

The normal duct suction velocity is inadequate to contain the fluoride gases from pots when the hoods are open during anode changing. The solution to this problem was to increase the duct suction velocity by a factor greater than two using a dual duct system. The first two industrial applications of the dual duct system were at the Alcoa Deschambault smelter in the fall of 2002, which increased the duct flow from 864 to 1,584 Nm3/h during anode change operations,⁴ and the Hydro Sunndal smelter in Norway also in 2002, which increased the normal duct suction of 5,000 Nm3/h to 15,000 Nm3/h during anode change operations.⁵

Emissions from hot anode butts and crust account for 35% of the total fluoride emission in potrooms. The majority of HF emissions from hot anodes occur during the first 20 minutes of cooling. The solution to this environmental problem was to put the hot anodes inside cooling boxes. In 1999 Gilles Dufour began the design and development of prototype anode cooling boxes and crust bins and were implemented in 2000 at the Alcoa Deschambault smelter in Quebec, Canada.⁶ The use of cooling boxes that contain the fluoride gases resulted in a 35% reduction in HF emissions to the potroom rooflines at Deschambault. Stig Lægreid of Hydro Aluminium developed anode butt cooling boxes that are connected to the plant fume duct system to collect the fluoride gases from hot bath and anodes taken out of pots. The first section of HAL250 cells were started in October 2002 at the Sunndal smelter in Norway.⁵ By the use of the special dual duct gas collection system and anode cooling boxes, the fluoride emissions are very low at Sunndal, less than 0.35 kg F/t Al, to meet environmental regulations (both OSPAR and local conditions). The dual duct system is now standard installation at new modern aluminum smelters.

Introduction of Computers for Cell Control
The underfeed-overfeed alumina control is another key breakthrough in the operation of aluminum pots. The advent of process control for manufacturing process was in the 1960s but only involved resistance control procedures. The initial concept of underfeed-overfeed alumina in aluminum pots was first perfected by Dr. Warren Goodnow⁷ at Kaiser Aluminum in 1974. The demand feed strategy determines the rate of alumina addition to a pot, based on line amperage and pot voltage signals. The notion of a demand to feed a higher rate of alumina is based on the observation that the cell resistance rises as the bath is depleted of alumina prior to an anode effect (Figure 7).

Kaiser Aluminum was also the first company to develop and commercialize a distributed microprocessor computer automation system, called Celtrol, for controlling the pot operating voltage and alumina feed to aluminum pots using the demand feed strategy. Celtrol was invented by Steve Price and Charlie Nemeyer (software), Mark Kafel (hardware) and later Terrel Wright at ASG in Spokane. It was very successful in upgrading older aluminum smelters to improve pot performance and minimize environmental emissions by reducing the occurrences of anode effects. The Celtrol computer system was used extensively in all Kaiser aluminum smelters as well as in many international smelters.

Bath Chemistry Changes
Baths with High AlF₃ Contents Alcoa was probably the first company to realize that higher AlF₃ concentrations in the bath could give higher current efficiency. In the Wenatchee smelter in 1954 the AlF₃ concentration was increased from 1.5 to 6% excess AlF₃, and the current efficiency went up 3%, from 84.8 to 87.9%. Thomas Holmes did the first industrial tests with high bath acidity in the 1950s.³ Further work was done in Badin in 1965, where a test at 11% AlF₃ was done in P-155 pots. However, the test was a failure, which has been vividly described by Holmes himself.⁶ The AlF₃ concentration was then increased much too fast, the pots lost their protective side ledge and "they tapped out faster that we could patch them," to use his own words. This was a serious setback and it was 11 years later, in 1976, that operation at 11.5% AlF₃ in the Badin cells gave 91% CE.

Mathematical Models for Magneto Hydrodynamic Calculations
Many companies have developed their own mathematical models for calculations of the magnetic fields in their pots. Undoubtedly, MHD design has to rank high on the list of great inventions in the 20th century. Robert F. Robl of Alcoa was doing this by hand and with a physical model prior to computers. Alcoa started to understand this in the early 1950s with design of pots with side-by-side orientation and quarter point anode risers, instead of the usual end-to-end pots with end risers. Big pots (>100 kA) probably would not be very practical without this.

Important advancements were made in the development of MHD models in the twentieth century. These led to retrofitable changes in the pot bus-bar designs in existing aluminum potlines of older pot technologies that substantially improved the MHD behavior and consequently the pot performance effi ciencies. First, magnetic compensation technology was developed by Vinko Potocnik (Alcan), Wolfgang Schmidt-Hatting and Jacques Antille (Alusuisse), Marc F.G. Jouget and Jean P. Givry (Pechiney) in the 1950s and 1960s, and Thorleif Sele and Hans Georg Nebell (Hydro) to convert endto- end prebake and Soderberg pots with compensating three-riser asymmetric bus to reduce the high Bz fields associated with the closeness of the adjacent row of pots in the same potroom. This made it possible to dramatically increase the potline current without a loss in current efficiency.

Later, Nobuo Urata (Kaiser), Detlef Vogelsang, Christian Droste, and Martin Segatz (VAW), and Jean–Pierre Dugois and Paul Morel (Pechiney) developed the MHD modeling capability to retrofit the end-riser Kaiser P69, Pechiney AP13 and Reynolds P19 prebake cells with magnetically compensated bus-bars under the pots in order to reduce the high Bz associated with the high current flow around the ends of the pot, subsequently making it possible to increase the potline current.

Large Pot Development (High-Amperage Potlines)
The size and amperage of Hall-Héroult pots have steadily increased in the twentieth century. Typical pot size was about 50 kA in 1940, compared to 10 to 20 kA pots in 1914. In 1963 Alcoa had a potline in North Carolina running at 155 kA and in 1969 Alcoa had a 225 kA potline in operation in Tennessee. These 225 kA pots had individual elevation adjustment of anode pairs, because it was then believed that this was necessary for operation of so large pots. Later experience has shown that this was unnecessary. The Høyanger 220 kA potline was started in 1981 with a fixed anode bridge.

The AP18 prebake cell is well-known as being the first modern prebake cell. The project was led by Eric Barrillon and Gerard Hudault. The cell was originally designed by Jean–Pierre Dugois and Pierre Homsi (busbar modeling), Paul Bonny (computer control system), Jean-Louis Gerphagnon (cell hardware and construction), using extensive state-of-the-art magnetic and thermoelectric modeling. In 1976, Aluminium Pechiney's Laboratoire de Recherches des Fabrications (LRF) started up the first four prototype AP18 prebake cells operating at 175 kA, and in 1979 they installed 60 AP18 cells in potline F at St. Jean-de-Maurienne, France. The first commercial potline of AP-18 cells was started at Fort William, UK in 1981. Later this type of cell achieved a record operating performance of 95% current efficiency and 13.3 DC kWh/kg Al at 180 kA.⁸

The next major advancement in cell technology was the development of the +300 kA superpots. The first cells to operate above 300 kA were the Alcoa A817 and the Pechiney AP30 cells. In 1978 Alcoa was running a pilot cell at its Massena, NY smelter at 280 kA. This was the basis for the Alcoa A817 pots that were installed at the Portland, Australia plant. Construction of the Portland plant started in 1980 but due to an economic downturn the construction was delayed and the two potlines with a total of 404 pots operating at 300 kA did not start until 1986. There were initially extensive operating problems with the pot that did not get solved until a magnetic retrofit in 2002. No other potlines of this type were ever built due to the operating problems (Figure 8).

At nearly the same time the Pechiney AP30 pot was developed at St. Jean-de- Maurienne which also operated above 300 kA. The project was led by Maurice Keinborg, Jean-Louis Gerphagnon and Bernard Langon. The cell was originally designed by Jean Pierre Dugois and Pierre Homsi (bus-bar modeling), Benoit Sulmont (computer control system), Christian Duval (cell hardware and construction) and Bernard Langon (operations). New pot technology inventions developed for the AP30 cell include forced-air cooling of the cathode shell using localized jets, and detection of the bath level via chisel stroke.

The AP30 cell technology era began in 1981 with the development of cells operating at 280 kA, which was industrialized on potline G of 120 AP30 cells started up in the Saint-Jean-de- Maurienne, France in 1986. In 1991 the Dunkirk smelter with 264 AP30 cells was started at 293 kA, and is now reported to be operating beyond 360 kA. The most recent technology breakthough was the development of AP50 (500 kA) prebake cells that was started by LRF in 1989 at Saint Jean de Maurienne. This represents a jump of 200 kA higher than the previous generation of AP30 cells.

A recent breakthrough invention has been the development of high amperage aluminum pots, 400-500 kA, that operate at low specific energy consumption, 12.500 DC kWh/kg Al. Due to the high cost and decreasing availability of electrical power in China, Northeastern University Institute (NEUI) has developed a family of high-energy-efficiency pot (HEEP) technology. This family of 400 kA aluminum cells operates stably and effi ciently at 3.85 volts and 12.50 DC kWh/kg Al.9 The operating amperage of NEUI400 HEEP pots actually exceeded the amperage indices, for example: Henan Zhongfu Industry Co. (415 kA), Linfeng Aluminium Industry and Power Co. (440 kA), Shandong Nanshan Aluminium Co. (430 kA) and Jinning Aluminium Co. (460 kA). NEUI is also developing a family of NEUI500 (500 kA) pots that will also operate at 3.85 volts and 12.500 DC kWh/kg Al. To operate at the low anode-cathode distances and energy values the HEEP pots have low anode current density, as well as improved magnetic and thermal designs. The NEUI 300 and NEUI 400 prebake project is directed by Mr. Lu Dingxiong and Mr. Liang Xuemin. The development of the busbar arrangement, magnetic fluid stability technology and thermal design was completed by Mao Jihong, Mr. Qi Xiquan, Mao Yu, and Ban Yungang. The computer control system was developed by Wang Dequan and Qi Xiquan.

New Cathode Materials
Initially all aluminum electrolysis pots had a monolithic carbon cathode lining that was installed manually by ramming a plastic paste into place. Prebake cathodes first appeared in pots at St. Jean-de-Maurienne, France in the 1920s. From the 1950s to 1970s there was a gradual conversion by aluminum companies to use prebaked cathode blocks with rammed paste in joints and seams. Since the 1970s there has been an increase in the added graphite content (semi-graphitic) in cathode blocks in order to reduce the electrical resistance of cathode blocks and thus allow a reduction in specific energy consumption of cells. The first major manufacturers of cathode blocks for aluminum pots were Great Lakes Carbon and Union Carbide Carbon companies in North America, as well as Sigri and Carbon Savoie companies in Europe.

Sumitomo Corporation was the first company to manufacture and commercialize fully graphitized cathode blocks for use in aluminum cells. The brand name was SK-Block, where "S" was coming from Sumitomo and "K" coming from Kyowa Carbon, the original developers, and it has been known for about 30 years throughout the industry. Graphitized cathode blocks provide significant energy savings in aluminum pots due to its unique high electrical conductivity. It was initially employed in the conversion of standard wet paste VS Søderberg pots to the "Sumitomo" dry anode paste Søderberg technology in the 1980s. But in recent years graphitized cathode blocks have proven to be especially successful in reducing the cathode voltage drop in modern prebake pots, thus allowing smelters to increase the potline amperage even higher.

The introduction of silicon carbide bricks for sidewalls came about as a result of the increasing potline amperage. The plants needed to reduce the sidewall insulation to increase the heat loss through the sides of the pot and thereby maintain a protective side ledge of solid cryolite. SiC had similar thermal conductivity to carbon and was a good choice because it also provided a silicon tracer for sidewall attack. It is used in most modern pots now. The first major producers of high grade silicon-nitride bonded silicon carbide, which has a higher chemical resistance to molten cryolite, were Carborundum and Norton refractory companies in the US and later Annawerk in Germany.

Pot Tending Machines
The cranes in the potlines have indeed become increasingly more sophisticated. In addition to include the cavity cleaning scoop, modern cranes can be equipped with a pneumatic driven punch for crust breaking and a bin and a feed spout for the alumina-bath mixture that is used to cover the newly placed anodes. Alternatively, pot tending motorized vehicles can be used for anode changing and also for metal tapping. Although aluminum production is still labor-intensive, these improvements have greatly reduced the need for heavy manual work for the operators and exposure to dust and fluoride fumes.

Large multi-purpose potroom cranes became necessary only after the construction of large modern high-amperage aluminum pots due to their use of very large and heavy anodes, as well as larger aluminum tapping crucibles for the increased aluminum production. However, potroom cranes are no longer required to add alumina to ore bins on pots, as this can be done by air slides and dense phase transport of alumina.

ECL in France has been supplying cranes to the primary aluminum industry since 1947 with 1,000 PTM in operation in potrooms, and with addition furnace tending assemblies and cranes in the carbon plant and cast house. ECL was created in 1947 in Lille by Robert DeBuire, a mechanical structure engineer, Joseph Tella, an electrical engineer and Daniel DuClaux, a mechanical engineer. NKM Noell Special Cranes in Germany has been working closely together with aluminum companies for more than 40 years and has now more than 1,000 cranes in operation worldwide.

ECL invented the pacman, or clam shell device, that is used by crane operators to clean large pieces of crust and carbon out of the bath in pots when changing anodes. It was not implemented in the AP18 technology until some plants were experiencing an excessive number of anode spikes. The pacman was first installed in the AP18 potline at the Karmøy smelter in 1985 (Figure 9).

Slotted Anodes
The smelter in Deschambault, Quebec was the first to use slots (with very small slot depths) in anodes to stop cracking anodes. The idea was that these small slots would act to stop crack formation. Ron Barclay was the Alumax carbon expert trying to solve the anode cracking problem. However, they stopped using slots once they solved the cracking problem, and they then were not realizing the value and the effect of the gas bubble removal from the underside of the anodes.

The Alouette smelter in Quebec was experimenting in making "huge" anodes for their AP30 pots to combine two anodes into one anode. They found out that this dramatically increased the voltage drop and pot instability, and thus they went back to the "regular" size anode, but put small slots in it to reduce the voltage drop.

At about 1996 Eric Lavoie and Luke Tremblay from Reynolds' Baie Comeau smelter visited the Alouette smelter and recognized the value of slots in the anodes. Reynolds then started working at optimizing these slots, how many slots that were needed per anode, and how deep the slots should be. Xiangwen Wang from Reynolds made measurements in 1998-1999 on the current distribution, etc., on anodes at Baie Comeau and developed specific recommendations on the number of slots and the slot depth to achieve the maximum benefits of slots. Baie Comeau adopted Xiangwen Wang's recommendations and implemented equipment to "cut slots with circular saws". All anodes used in all AP-18 potlines at Baie Comeau had slots (Figure 10). As a result they achieved a minimum voltage reduction of 50 mV on all pots, and they were able to increase the amperage by 10 kA.

Thus, Baie Comeau was the first plant in the world to successfully implement slotted anodes in all potlines in order to increase potline amperage. Erik Trembley was the offi cial slotted anode concept man, and Xiangwen Wang was the official person that got the science right!

Increased Amperage in Existing Pots
Previously, the pots were designed for a given amperage, and that amperage was the target, if they could reach it. Increased amperage in existing pots has obvious advantages. This is why many aluminum producers have increased the amperage in their potlines in the last 20 to 30 years.

An example is the retrofitting and modernization of older 150 kA end-to-end pots that have made it possible to operate some of these pots at above 200 kA. Some end-to-end potlines have now even reached 220 kA. These are indeed impressive results. It is really surprising how much improvement has been achieved here, and for some potlines it has been possible to increase amperage with more than 30 to 40%, and even up to 50%. Economically this has been one of the success stories for many aluminum smelters.

The potlines have indeed become a safer working place in the twentieth century. This is mainly due to increased awareness and attention about safety and risk-based management, and workers' health and safety have become key elements in modern management philosophy. Introduction of automatic alumina breakers and feeders has had a great influence on safety by reducing the manual work. The PTM equipment, and especially the scoop for cavity cleaning, has reduced the need for operators working on the floor during anode change.

The environmental problems have shown remarkable progress. The fluoride emissions from the smelters were a huge pollution problem in the past, and the invention of dry scrubbers is perhaps the greatest contribution to improved environmental protection. The fluoride emissions are now reduced to a fraction of what they were before 1960. Other improvements came later, after 1990, with the increased awareness of perfluorocarbon emissions from anode effects as significant contributors to the greenhouse effect. Present emissions are now only about one-tenth of what they were before 1990.

Energy reduction has been huge, from 40 to 13 kWh/kg Al during the twentieth century. The main contributor has been lower pot voltage in all parts of the pots, together with improved current efficiency.

In the twentieth century the understanding of the fundamentals of the Hall-Héroult process have indeed been increased significantly. Here we remember the Boudouard reaction, where carbon reacts with carbon monoxide to form carbon dioxide, and the Pearson-Waddington equation for calculation of current efficiency from the ratio of carbon dioxide to carbon monoxide in the anode gas. However, in the technology of the process the only invention that has been given the name of its inventor is the Søderberg pot.

In spite of the fact that there are still many unsolved problems, we have indeed come a long way since the days of Hall and Héroult. Our paper pays a tribute to those who have contributed to make the Hall-Héroult process safer and environmentally cleaner, as well as making it a more energy efficient and profi table process.

The past century was full of remarkable discoveries, developments, and achievements for the industry. But what will shape our industry in the future? Will it be drained cathode, huge kA, inert anode, carbothermic, low temperature electrolysis, organic electrolytes? Will it be dominated by materials development, sensor development/ control, etc.? What impact will SO₂ and CO₂ regulations have on production methods and smelter location? What impact will the public's insatiable appetite for electric power (e.g., electric cars) have on the industry energy cost and location of smelters (e.g., stranded power locations)? Or, will aluminum smelters thrive in areas of high population by becoming the best friend of power companies as "surge capacitors" for electric grids? Indeed, with these possibilities, the future of our favorite metal is as bright as the metal itself!

During the writing of this paper the authors have received valuable information from several people. We would especially like to thank Jay Bruggeman, John Johnson, Olivier Martin, Tor Bjarne Pedersen, Michel Reverdy, Jomar Thonstad, Geir Wedde, and Siegfried Wilkening for their kind interest and help.

1. T.B. Pedersen, A.K. Syrdal, and A. Saethre, Light Metals 1995, ed. James W. Evans (Warrendale, PA: TMS, 1995), pp. 253–256.
2. V. Mann, Light Metals 2006, ed. T.J. Galloway (Warrendale, PA: TMS, 2006), pp. 181–183.
3. T. Holmes, Light Metals 1995, ed. James W. Evans (Warrendale, PA: TMS, 1995), pp. 371–373.
4. S. Broek, N.R. Dando, S.J. Lindsay, and A. Moras, Light Metals 2011, ed. Stephen Lindsay (Warrendale, PA: TMS, 2011), pp. 361–367.
5. J.A. Haugan, A.H. Husøy, and K.ø. Vee, (Presented at the TMS 2003 Annual Meeting, San Diego, CA, March 3–6, 2003).
6. J-P. Gagne, R. Boulianne, J.-F. Magnan, M.-A. Thibault, G. Dufour, and C. Gauthier, ed. T.J. Galloway (Warrendale, PA: TMS, 2006), pp. 213–217.
7. W. Goodnow, US Patent 3,812,024, 1974.
8. M. Keinborg and J.P. Cuny, Light Metals 1982, ed. J.E.Anderson (Warrendale, PA: TMS, 1982), pp. 449–460.
9. Lu Dingxiong, Mao Jihong, Ban Yungang, Qi Xiquan, Yang Qingchen and Dong Hui, Light Metals 2011, ed. Stephen Lindsay (Warrendale, PA: TMS, 2011), pp. 455–460.

Gary P. Tarcy is Manager Electrolysis and Energy, Alcoa Inc., Alcoa Center, PA; Halvor Kvande is Chief Engineer, Hydro Aluminium, Oslo, Norway; and Alton Tabereaux is currently a technical consultant to aluminum companies in both prebake and Soderberg cell technologies. Tabereaux retired in 2006 as Manager of Process Technology, Alcoa Primary Metals. Mr. Tarcy can be reached at gary.tarcy@alcoa.com.