Patent Application
20060001199
1. Field of Invention
This invention refers to methods of aluminum production, specifically an improved method for aluminum smelter organization, leading to smaller size and competitive operating costs.
2. Prior Art
Aluminum is a metal growing in use, with a world consumption of around 28 million tonnes per year (2003) and growing. However one notices that this industry is currently concentrated in only about 250 smelters in the world. This is the case because until now, economies of scale have favored very large production units.
Aluminum production by electrolysis of alumina is very energy intensive whatever the process used. All smelters consume from 14 to 15.5 kWh per kg of metal produced according to the age of the smelter, exact technology used, quality of maintenance, and efficiency of operation. Most of that energy consumed is the reduction energy itself.
Primary aluminum production takes place in around 250 smelters in the world. All use the Hall-Heroult Process originating in 1880, consisting of electrolysis of alumina, or aluminum oxide, mixed with cryolite and other minerals forming an eutectic bath which melts at a lower temperature than pure alumina. In this process, electrolysis of alumina reacting with a carbon anode converts alumina and carbon into liquid aluminum and carbon dioxide. Energy supply is a critical factor since the reaction is very energy intensive: roughly 14,000 kWh is needed to produce one metric tonne of molten metal today. Technology has progressed generally to improve energy efficiency, at the cost of bigger and bigger smelters in which the current intensity and the number of cells has grown from levels of 10,000 Amperes in 1900 to 300,000 Amperes today, and the capacity has grown from a few thousand tonnes/year per plant to 250,000 and even 500,000 tonnes/year today. New investments (called “greenfield plants” in engineering lingo) are rare (one per year to one every two years in average worldwide) due to the growing difficulty in concentrating on one site an energy source allowing for sufficiently low cost per kWh procured, for a period allowing depreciation of the huge capital costs (typically $4,500 to 5,000 per tonne/year capacity in 2000-2004). It is generally recognized that between 2000 and 2004 the world aluminum industry spent on average 2 US cents per kWh or roughly $280 per tonne of aluminum produced.
The need to secure low cost energy has led to favoring of energy sources which allow easiest power concentration in one site: hydro-electric, coal-fired, gas-fired, or sometimes nuclear plants. Because of sheer size of projects, such sites are under more and more scrutiny regarding their pollution potential, meaning that overall costs, including political costs, of environmental impact studies to get a project approved have been escalating. Again because of combination of size of project and of high relative depreciation costs, the costs and delays of financial engineering have been also escalating. It is common that the various site evaluations, viability assessments, preliminary impact studies, pre-feasibility studies and bankable feasibility studies for a greenfield smelter planned in a country which doesn't produce aluminum yet can encompass a period of 10, 15 and even 20 years, sometimes to end in a decision not to pursue. Thus, tens of millions of dollars of such costs have been lost in projects in Chile, Libya, Malaysia, British Columbia, etc. No new smelter has been built in the US since 1969.
Two configurations of smelter pot lines exist:
The first configuration is the so-called Pre-Baked Anode (PBA) (illustrated in Figure A): Up to 40 rodded anodes (1) per cell, produced separately (Anode Plant with Baking Oven and Rodding Shop) are consumed as carbon of the anode and alumina (2) fed from the hopper are reduced in the bath of the cell (3) into carbon dioxide and aluminum. Rodded Anodes are changed in staggered order. Carbon anode blocks are produced through mixing petroleum coke and pitch into an anode paste, then blocks are molded and baked separately, then rodded on a metal rod and stem arrangement, the rodded anodes being connected on a busbar (4) above each cell, and replaced by a new rodded anode when they are consumed to around 80% of their volume and reduced to a minimal anode butt, the anode butt being then recycled.
The second configuration is the so-called Soderberg Smelter (described in Figure B) invented by Dr. C. J. Soderberg in 1927: Anode paste is directly fed in a “sleeve” or “shell” (5) above each cell forming a single Anode per cell (6). The paste, as the anode consumes continuously, is continually baked in (7). Anode paste is simply loaded on top (8), held by staggered studs (9) which are added, and pulled out, one after the other. There is no disruption of the process caused by anode change as in the PBA process.
Note that the heat generated by the cell progressively bakes the anode, which is continuously consumed and renewed and doesn't undergo a reheating as in the PBA process.
The Soderberg process gained market share and recognition because of its simplicity, while progress was made in energy efficiency in both processes between 1927 and 1972.
In the sixties, the invention by Aluminium Pechiney of so-called Magnetic Compensation combined with computerized process control monitoring alumina feeding, anode change, anode height, and other parameters gave an advantage to the PBA technology as shown in Figure A. This eventually lead to Soderberg shown in Figure B being abandoned for greenfields, even though similar improvements, including Magnetic Compensation, were later introduced with success on certain Soderbergs. Soderbergs were also criticized for their higher level of pollution. However, recent improvements have brought the best Soderbergs to the same level as PBA's. In 2004 Soderbergs still represent some 20% of world capacity, and extensions (added potlines) of Soderberg plants are still occurring, notably in Brazil. Any new investment in a greenfield smelter faces the following problems which our invention addresses:
(a) Minimal investment is of some 250,000 metric tonnes per year (tpy) capacity, needing a constant power supply of around 600 MW.
(b) Capital cost is at least $4,500 to $5,500 per tpy without energy generation.
(c) If energy generation is added, investment costs go from $500-$650 per kW for natural gas (but gas prices are high and volatile), to $1,000-$1,200 for conventional pulverized coal boilers using the purest quality of high caloric content coal, to $2,000 for nuclear plants, to $2,500 and well over for geothermic and hydro electric plants, but economies of scale generally lead to minimal power of around 1,200 MW per plant. Total capital cost escalates to well over $2 billion. In practice, greenfields are today built close to existing hydro-electric capacity (Quebec, Mozambique, Iceland, Venezuela . . . ) or to oil fields generating excess gas (Middle East). Availability of sites is more and more a constraint.
(d) Separate baking ovens for anodes consume extra power, generally from burners consuming costly, highly refined hydrocarbons.
(e) Prebaked Anode material is heated twice: during baking, and again when inserted in the cell bath. Rodded anode butts need to be recycled. The whole prebaked anode process imposes a costly investment in continuous handling, butt removal, crushing, stem straightening, yoke straightening, re-rodding.
A consequence of these growing inconveniences is that the price per tonne of aluminum measured in relation to the price per tonne of carbon steel, which had been regularly getting more competitive between 1900 and the late 1960's, plateau'ed around 1970-74 and since then has lost competitiveness. See Figures C-a and C-b. Figure C-a illustrates that, since 1910 until the early 70's, aluminum became more price competitive against steel. Figure C-b illustrates that since 1972 the trend has been reversed, with steel becoming more competitive.
Production costs of aluminum being highly energy intensive, the world industry has of course always favored sites were energy costs can be low in the long term. The world industry, in 2000-04, pays an average of 2 US cents per kWh, with actual cost ranging from as low as 0.5 cents (Quebec, Mozambique) to as high as 3 cents and more (most US and European smelters). A crucial issue when deciding where and when to build a new smelter is the long term contracts guaranteeing a low energy cost for at least the years, which is the period of high depreciation costs. Obtaining such long-term contracts, favorable rates, on such high total volume of energy per year is the factor which is delaying more and more all greenfield smelter projects in the world. For instance, Malaysia has been attempting to fund a smelter since 1988; Chile and Qatar since the mid 80's; Guinea, since 1937. In a highly favorable country such as Iceland a greenfield smelter takes a decade or more to go from initial assessments to start up.
This combination of a Soderberg aluminum smelter combined with a fluidized bed boiler (FBB) power plant has several economic benefits over mainstream design of power plants using either nuclear fuel, hydro-electric energy, geothermal energy, petroleum gas, or high grade coal burned in pulverized form, and of aluminum smelters using the conventional PBA technology:
A method producing aluminum combining an aluminum smelter and a power plant which allows lower capital intensity, small total investment, and competitive operating costs.
Fig A shows a prebaked anode cell cross section.
Fig B shows a Soderberg cell cross section.
Fig C-a shows a graph of the ratio of primary aluminum to steel prices from 1910 to 2000.
Fig C-b shows a graph of the ratio of primary aluminum to steel prices from 1972 to 2000.
Fig D shows a generic vertical stud Soderberg and circulating fluidized bed boiler facility.
The invention combines a Vertical Stud Soderberg smelter without anode paste plant (the anode paste is procured on the free market) with a Fluidized Bed Boiler Power Plant burning either coal of any grade or low quality carbon fuels such as coke or even peat or lignite. The invention envisions a combination of a Soderberg smelter system with its own power plant as shown on flowchart figure D.
