JPH0310715B2 - - Google Patents

Description

Claim 1: A fluidized cathode furnace for electrolytically reducing alumina to aluminum in a cryolite bath containing alumina, the frame having an inner surface coated with a refractory and a carbon material and forming a cathode hole. and; a cathode having at least a top surface containing an aluminum-wettable refractory hard material; and at least two anodes depending within the cathode bore, each about 1 inch from the top surface of the cathode.
having lower surfaces spaced apart by an anode-cathode spacing (ACD) of ~5 cm and a flow path between the anode and the cathode; said at least two anodes being inclined at an angle of .degree., thereby forming lower and upper ends of said flow path between said anode and cathode; in order to ensure the supply of alumina for the electrolytic reaction; , a device for providing a minimum bath flow rate Q in the cathode bore around at least one substantially horizontal bath circulation loop; and providing at least one return flow path to complete the circulation loop. The formula Q (furnace) (cm ³ /sec) = (0.008) [total furnace current (A)]
/Δwt%Al ₂ O ₃ , and the loop is connected to (i) the channel between the anode and the cathode; (ii) the lower end of the channel between the anode and the cathode. (iii) at least one upper flow path communicating with an upper end of the flow path between the anode and the cathode; (iii) the upper flow path and the lower flow path; at least one return channel in communication with the channel, the at least one return channel located between adjacent anodes having dimensions h, W, L, with the proviso that h
is the bath depth (cm) at any point, W is the width of the return channel (cm), L is the length of the return channel (cm), and these are the formula R _f (furnace channel) = (2h+W)L/h ³ w ³ , and the maximum value of R _f (furnace flow path) is expressed as Equation 1/[R _f (furnace)] ^1/2 = _o 〓 ⁱ⁼¹ (1/[R _f (furnace flow path)] ^1/2 ) and the relationship of the formula R _f (furnace) = (X) (890 ² )/Q (furnace) ² , where (X) is approximately 0.2×
10 ^-3 /cm ⁴ to 230×10 ^-3 /cm ⁴ A falling cathode furnace. 2. The alumina concentration in the bath that enters the flow path between the anode and the cathode from the lower flow path and is supplied to the upper flow path is about 0.2% to 5.0%. The falling cathode furnace according to item 1. 3. The net total bath flow rate per unit anode area in the sulfur path between the anode and the cathode is approximately
1.2×10 ⁻¹ to 8×10 ⁻⁴ cm ³ /sec/cm ² . The falling cathode furnace according to claim 1. 4 The furnace has an anode current density (A/
cm ² ), the falling cathode furnace according to claim 1. 5 The furnace has an anode current density (A/
cm ² ), the falling cathode furnace according to claim 1. 6 The furnace has an anode current density (A/
cm ² ), the falling cathode furnace according to claim 1. 7 The furnace has an anode current density (A/
cm ² ), the falling cathode furnace according to claim 1. 8. The falling cathode furnace of claim 1, wherein the ACD is approximately 1.5 cm to 4.0 cm. 9. The falling cathode furnace of claim 1, wherein the ACD is approximately 2.0 cm to 3.0 cm. 10. The falling cathode furnace of claim 1, wherein the ACD is approximately 2.5 cm. 11 The upper surface of the cathode and the lower surface of the anode are inclined at about 5 to 10 degrees from the horizontal line.
Flowing cathode furnace described in Section 1. 12. Claim 1, wherein the upper surface of the cathode and the lower surface of the anode are inclined at about 6 to 8 degrees from the horizontal line.
Flowing cathode furnace described in Section 1. 13. The falling cathode furnace according to claim 1, wherein the upper surface of the cathode and the lower surface of the anode are inclined at about 8 degrees from the horizontal. 14. The falling cathode furnace according to claim 1, wherein the L (cm) is about 15 to 300. 15. The falling cathode furnace according to claim 1, wherein the L (cm) is about 30 to 250. 16. The falling cathode furnace according to claim 1, wherein the L (cm) is about 60 to 200. 17. The falling cathode furnace according to claim 1, wherein said L (cm) is about 122. 18. The falling cathode furnace according to claim 1, wherein the X (10 ³ /cm ⁴ ) is about 1.0 to 20. 19. The falling cathode furnace according to claim 1, wherein the X (10 ³ /cm ⁴ ) is about 5.0 to 10. 20. The falling cathode furnace according to claim 1, wherein the X (10 ³ /cm ⁴ ) is about 6.2. 21 The above Q (10 ³ /sec/cm ² ) is approximately 1.1×10 ^-3 ~ 6.0
×10 ^-2 The falling cathode furnace according to claim 1. 22. The falling cathode furnace according to claim 1, wherein the furnace has a plurality of return passages = n, where n is represented by the formula R _f (furnace passage) = n ² R _f (furnace). 23. The falling cathode furnace of claim 1, wherein said slope and said ACD are dimensioned to limit the thickness of bubbles created by said electrolytic reaction to less than 50% of said ACD. 24. Claims 1 to 2, wherein the return flow path is located between the inner wall of the cathode hole and the anode.
The falling cathode furnace according to any of Item 2. 25 The aluminum wettable fireproof hard material is 2
Titanium boride, titanium carbide, zirconium diboride,
25. A falling cathode furnace according to any one of claims 1 to 24, which is selected from zirconium carbide and mixtures thereof. BACKGROUND OF THE INVENTION The present invention relates to a novel electrolysis furnace for producing aluminum from alumina and a method of operating the same.
In particular, the present invention provides an electrolytic furnace for the production of aluminum using a cryolite-based bath electrolyte that eliminates the problems caused by the narrow anode-cathode spacing obtained in conventional falling cathode furnaces within the ACD gap. has been demonstrated to be overcome by inducing specific aspects of the bath flow to facilitate alumina supply and raw gas removal and to improve flow of the fabricated metal. Operation of the experimental electrolysis furnace demonstrated the ability to deliver large amounts of molten alumina to the electrolysis zone even at very narrow anode-cathode spacing. Electrolytic furnaces commonly used in the production of aluminum are of the classic Hall-Heroult design, with a carbon anode and a generally flat carbon-lined bottom serving as part of the cathode system. The production of aluminum by electrolytic reduction of alumina uses an electrolyte, which consists primarily of molten cryolite to which molten aluminum has been added, and may also contain fluorite, aluminum fluoride, and other metal fluoride salts. . The molten aluminum resulting from the reduction of alumina is often allowed to accumulate as a molten metal bed or pool on a carbon-lined bottom at the bottom of the vessel forming the electrolytic furnace, forming a liquid metal cathode. A carbon anode extending into the vessel and in contact with the molten electrolyte is aligned with the liquid metal cathode. A current collector bar, such as steel, is often embedded in the carbon-lined furnace bottom to complete the connection to the cathode system. Hall-Heraurt furnace designs and sizes vary, but all have relatively low energy efficiencies of about 35-45%, depending on the furnace geometry and mode of operation. In other words, the theoretical amount of electricity required to produce 1lb (2.2Kg) of aluminum is approximately
2.85Kwh/lb, but the actual power usage range is 6-8.5Kwh/lb, and the industry average is approximately
It is 7.5Kwh/lb. This deviation from the theoretical energy consumption is largely a result of the voltage drop in the electrolyte between the anode and cathode. Therefore, much research has been focused on reducing the anode-cathode distance (ACD).
However, the molten aluminum bed that serves as the cathode in the furnace can be irregular and vary in thickness due to electromagnetic effects and bath circulation, and past practice has shown that it is difficult to ensure a relatively high current efficiency and direct short circuit between the anode and the metal bed. In order to avoid this, it was necessary to keep the ACD at a safe distance of 3.5 to 6 cm. Such gap spacing is 1.4~2.7V
, which is in addition to the energy required for the electrochemical reaction itself (2.1 V based on enthalpy and free energy calculations). Therefore, great efforts have been made to develop more stable aluminum floors in order to reduce the ACD to less than 3.5 cm and obtain energy savings. The use of refractory hard materials (RHM) such as titanium diboride in the form of tiles as the cathode surface has been studied for a long time, but until very recently, RHM with good adhesive properties
No tile or surface coating was obtained. Titanium diboride is known to have unique properties in that it is electrically conductive and can be wetted with molten aluminum to form an extremely thin aluminum film. Instead of the traditional unstable molten aluminum bed,
The use of a very thin aluminum film covered by the RHM surface and flowing down the sloped cathode face has been proposed as a means of reducing ACD, increasing efficiency and reducing voltage drop. However, attempts to achieve such a goal in the past have failed due to the incompleteness of the resulting RHM surface and the inability to overcome the difficulty of supplying sufficient molten alumina to the narrow ACD (as narrow as 1.5 cm). It was attributed to Thus, narrow ACDs, including excessive persistent anodic effects, create alumina depletion problems. If too much alumina is supplied to prevent these problems, sludge adhesion (matting) may occur.
mucking) occurred, clogging the furnace and interfering with its operation. Another approach to reducing energy consumption was to smelt aluminum from aluminum chloride instead of alumina. This method requires less electricity to produce aluminum than traditional electrolytic methods.
Reduced by 30-40%. The process uses the traditional Bayer process to convert bauxite to alumina, which is converted to aluminum chloride in a chemical plant and then smelted in an electrolytic furnace. In the furnace, aluminum chloride is extracted from aluminum and
It is further decomposed into chlorine, which is recycled back into the chemical plant to produce aluminum chloride. Such techniques use flow-through reactors with non-consumable anodes. However, existing aluminum plants cannot be modified for this process because the aluminum production furnace for this process is not compatible with Hall-Heraurt type furnaces. Therefore, the production method using aluminum chloride requires completely new capital investment. The present invention enables significant energy savings in retrofitted plants and eliminates the need for entirely new equipment. This is the background for developing the present invention. DISCLOSURE OF THE INVENTION It is an object of the present invention to provide an improved aluminum electrolytic furnace using a fluorite-based bath electrolyte and having high electrical efficiency. It is thus an object of the present invention to provide a basic furnace for producing aluminum from alumina, which can be used in a single furnace system or in a multiple furnace system for saving construction and operating costs. It is an object of the present invention to provide a furnace design that increases current efficiency and reduces voltage drop by narrowing the ACD over prior art, and that can be constructed by retrofitting current Hall furnace equipment. Therefore, less than 3 cm
It is an object of the present invention to provide an electrolysis furnace having an ACD that ensures proper alumina supply to the anode-cathode gap without the need for overfeeding. Another object of the present invention is to provide an aluminum reduction furnace design that provides a flow-through configuration and controlled atmosphere. These and other objects, which will become clear from the disclosure set out below, are achieved by providing a furnace in which the individual steps are separated as far as possible. Collection of product aluminum using a sloped solid cathode forming anode, either Soderberg or pre-baked, coupled with the correct selection of other parameters defined herein. A bath flow is induced in the anode-cathode gap to provide an adequate supply of alumina to the anode without impeding flow into the reservoir. In one form, pumping of the gas below the anode causes the bath to flow through the furnace to the furnace supply chamber and back out the furnace supply chamber on the opposite side. Individual furnaces of this type can be combined in multiple configurations to create a system that meets the physical and economic constraints of any given plant site. Thus, the present invention provides a method for creating a condensate-rich material between the anode and cathode without preventing the molten aluminum from flowing down into the collection sump and without causing fragmentation of the falling aluminum surface which would result in a reduction in current efficiency. so that the anode gas induces sufficient flow.
The present invention relates to an inclined electrolytic furnace in which a cathode is arranged. The bath is then directed from the enrichment bath area to the area where some of the alumina content has been consumed. Alumina is automatically supplied to the bath in a quantity adjusted to prevent either matting or anodic effects. According to one aspect of the invention, a fluidized cathode furnace for electrolytically reducing alumina to aluminum in a cryolite bath containing alumina, the cathode having an inner surface coated with a refractory and a carbon material; a frame defining a hole; a cathode having at least a top surface containing an aluminum-wettable refractory hard material; at least two anodes depending within the cathode hole, each about 1 inch from the top surface of the cathode;
having lower surfaces spaced apart by an anode-cathode spacing (ACD) of ~5 cm and a flow path between the anode and the cathode; said at least two anodes being inclined at an angle of .degree., thereby forming lower and upper ends of said flow path between said anode and cathode; in order to ensure the supply of alumina for the electrolytic reaction; , a device for providing a minimum bath flow rate Q in the cathode bore around at least one substantially horizontal bath circulation loop; and providing at least one return flow path to complete the circulation loop. The formula Q (furnace) (cm ³ /sec) = (0.008) [total furnace current (A)]
/Δwt%Al ₂ O ₃ , and the loop is connected to (i) the channel between the anode and the cathode; (ii) the lower end of the channel between the anode and the cathode. (iii) at least one upper flow path communicating with an upper end of the flow path between the anode and the cathode; (iv) the upper flow path and the lower flow path; at least one return passageway communicating with the channel; the at least one return passageway located between adjacent anodes having dimensions h, W, L;
h is the bath depth (cm) at any point, W is the width of the return channel (cm), L is the length of the return channel (cm), and these are the formula R _f (furnace channel) = (2h+w)L/h ³ w ³ , and the maximum value of R _f (furnace flow path) is expressed as Equation 1/[R _f (furnace)] ^1/2 = _o 〓 ⁱ⁼¹ (1/[ R _f (furnace flow path)] ^1/2 and the relationship of the formula R _f (furnace) = (X) (890 ² )/Q (furnace) ² , where (X) is approximately 0.2×
A falling cathode furnace is provided, characterized in that the irradiance is 10 ^-3 /cm ⁴ to 230×10 ^-3 /cm ⁴ .

[Brief explanation of the drawing]

FIG. 1 is an end view of an electrolytic furnace using an inverted V-shaped cathode, i.e., double tilted cathode configuration; FIG. 2 is an end view of a furnace of the present concept using a single tilted cathode; FIG. Figure 4: End view of a single tilted cathode, flow-through aluminum reduction furnace with external replenishment. A plan view of a falling cathode furnace. Figure 6 is a plan view of a multiple furnace system using multiple single inclined falling cathode furnaces in tandem. Figure 7 is a plan view of a single inclined falling cathode furnace using partial baffles to limit the bath flow rate. Figure 8 is a plan view of a furnace with two oppositely angled cathodes; Figure 9a is a side view of a hydraulic analogue model designed to simulate gas generation in a falling cathode furnace. Figure 9b is an explanatory diagram of the normal bath flow in ACD, and Figure 9c is the flow near the flowing aluminum.
Figure 9d is an illustration of the area of turbulent and fragmented bath flow with no net flow of bath material through the ACD; Figure 10 is an illustration of the area of turbulent and fragmented bath flow in conditions of high flow resistance. Figure 11 is a graphical representation of the bath flow to cathode slope for the furnace simulation model; Figure 11 is a graphical representation of the bath flow to cathode slope for the furnace simulation model at medium flow resistance; Graphical representation of the bath flow to-cathode slope of the furnace simulation model at low conditions; Figure 13 is a conceptual diagram of pump efficiency vs. flow resistance in a falling cathode furnace; Figure 14 is a graphical representation of the required flow resistance to cathode slope in the return path. , Fig. 15 is a diagram of the furnace design parameters, Fig. 16 is a cross-sectional view of the inclined cathode electrolysis experimental reactor, Fig. 17 is a graphical representation of the furnace operation stability obtained in three experimental reactors, and Fig. 18 is a diagram of the 2 Illustration of the relationship between anode-cathode polarization and cathode tilt at two current densities. DETAILED DESCRIPTION OF THE INVENTION The present invention comprises a sloped falling cathode having a surface of titanium diboride or similar refractory hard material;
The angle of inclination from the horizontal is selected to contribute to the control of bus motion, which is hydrodynamically dependent on certain key parameters. The bath motion is controlled so that a sufficient flow of the alumina-rich bath is induced in the interelectrode gap in the space just below the anode surface, without disturbing the aluminum flow and to avoid anode effects. Thus, problems such as muck formation resulting from excessive alumina supply, persistent anodic effects resulting from insufficient bath addition, and retting (protrusion formation) that impede bath circulation can be avoided without a conventional metal bed. It is possible to overcome the alumina delivery problems that arise with the extremely narrow anode-cathode spacing (ACD) made possible by the development of operational refractory hard metal surface cathodes. The success or failure of the energy-saving flowing cathode aluminum reduction furnace concept depends on two key factors: (1) durable RHM cathode material, and (2) narrow anode-cathode spacing (ACD).
It depends on the ability to control and optimize the bus flow through. Through experiments underway in production furnaces,
It is shown that a material has been developed that meets the first requirement. No. 1, filed July 9, 1982, all commonly assigned to the present assignee and pending.
No. 395343, No. 395344, No. 395345 and U.S. Patent Application No. 400762, No. 400763 dated July 22, 1982;
No. 4000764, No. 400765, No. 400772, No.
No. 400,773, et al., relates to a set of refractory hard material coatings, methods of application, and cathode structures that provide a durable, adhesive cathode surface that exhibits aluminum wettability and has excellent electrical properties. Research using a water model to simulate the hydrodynamic conditions in the ACD of a tilted cathode furnace determined the furnace design criteria necessary to meet another requirement. The results of the water model were verified in a laboratory-scale electrolysis furnace. Contract number DE−AC03−
Numerous furnace designs used in the Kaiser-DOE graded TiB ₂ cathode test reported by 76CS40215 and other published reports and patents deviate from the design requirements of the present invention and have significant cannot handle major fluid dynamics problems. These design criteria include the slope of the cathode surface,
ACD spacing, resistance to bus flow in the return flow path,
and anodic current density. While the first two of these criteria have been considered in existing designs, reports and patents, the latter two criteria, along with their importance, are pointed out for the first time in this paper. It is known in aluminum electrolysis that the anode surface is shaped to be substantially parallel to the cathode by the electrolysis process. The tilted cathode thus tilts the anode in a similar manner. As is well known, a large amount of gas is generated at the anode surface during the electrolytic process. In conventionally designed aluminum reduction reactors with horizontal anode and cathode faces, gases generated at the anode face move laterally from the reactor below the anode face and outward to the nearest vertically oriented vent; It is also known that the mouth generally consists of a vertical side of the anode. A tilted anode can either enhance or reduce this gas movement, depending on the direction and magnitude of its tilt. It will be appreciated that proper cathode and therefore anode tilting will cause the anode gas to move in the desired direction. This anode gas then drives the bath near the anode surface in the same direction by a known action also used in gas lift and gas jet pump and bubble column designs. This drive is supplemented and enhanced (if necessary) by other pumping effects, for example by appropriately designed pumps. that flow in a fluid system is established by an equilibrium between the fluid driving force and the resistance to flow within the elements within the system;
It is also well known in fluid mechanics that, depending on the configuration, the local flow velocity is in the same direction as the fluid drive direction, but sometimes in the opposite direction. The net bath achieves an equilibrium between, on the one hand, the buoyant force due to the tilt, the bubble force and the force that drives the bubble sideways below the horizontal anode, and, on the other hand, the resistance to flow, resulting in the required alumina supply. The principle of the invention is to arrange the slope in such a way that it imparts movement. The local bus velocity near the anode surface is in the desired direction, ie the direction in which the driving bubble moves. At the same time, the configuration is such that the bath velocity near the cathode surface is equal to the bath velocity near the anode, but is not so great as to impede aluminum flow over the cathode, which is in the opposite direction to the flow of bubbles that drive the bath. will be done. If an alumina rich bath is placed on a particular side of the furnace, bath flow into the space directly below the anode surface occurs when this side of the anode is lower and slopes upwardly away from this side. This slope is large enough to overcome and reverse the gas flow that would otherwise be directed from the central portion of the anode to this side of the anode. The precise configuration and placement of the slope will vary with the location of the alumina concentrate bath source. This location can be on one side, both sides, or even in the middle of the oven, depending on the size of the oven and the type of anode used (prebake or Soderberg). Alternative embodiments of the invention appear in which the slope of the cathode (and thus the anode) is uniform across the width of the furnace or in which the slope varies in various ways. These are, for example, those with unequal but constant slopes in the same direction on both halves of the furnace cross-section, and those with equal but opposite slopes on both halves (e.g. double slopes,
(inverted V shape). The concepts of the present invention apply to both pre-bake and Soderberg anode design furnaces. For example, an inverted V shape may be most suitable for a furnace whose width is covered by two prebake anode blocks, but this configuration can also be applied to vertically studded Soderberg anodes with central ventilation. . The inverted V shape is suitable for supply from both sides,
A V-shape would be suitable for central feeding. Similarly,
Tilt in only one direction (constant or variable)
Desirably applied to designs having
Not limited to the anode. The selection of cathode tilt for a particular application must be compatible with other governing parameters to achieve the desired hydrodynamic properties. These parameters are the ACD clearance (i.e. ACD spacing), the anode current density (current and anode surface dimension), and the bus return resistance (i.e. the length of the bus return channel or passage,
depth, width and material). The controlled bath flow will provide sufficient alumina-rich electrolyte to the anode surface to prevent excessive anodic effects.
Other benefits include prevention of excessive gas accumulation in the ACD;
Reducing the overvoltage of the anode and cathode, as well as reducing the fragmentation of the product aluminum flow. Figure 1 shows a vertical stud with side feed, inverted V cathode design with central gas outlet and collector bar.
FIG. 2 is an end view of the Soberberg reduction furnace. The cathode surface 6 slopes upward from both sides of the furnace toward the center of the anode 1;
The anode gas is vented through a vent 2 in the center of the anode to a collection system (not shown). collector bar 4
are arranged tightly parallel to the cathode surface 6 and offset to avoid the aluminum sump 5. The angle of the cathode plane 6 from the horizontal should be large enough to induce the required bath flow, but the anode
It must not be so large as to cause excessive bubble deformation or bath turbulence in the cathode space 7. You must also comply with the restrictions described below. For example, tilt angle has a significant effect on bubble morphology and motion, although this has not been previously noted. When the anode surface is tilted from the horizontal, the bubbles exhibit a pronounced tendency to assume an oblong shape with the long axis perpendicular to the direction of bubble motion, rising along the inclination of the anode. As the relative velocity between the bubble and liquid increases, the leading edge of the bubble becomes thicker. and over a certain range of slope,
A progressively thicker leading edge results in a reduction in the driving effect of the bubble due to increased resistance due to interface distortion and abrasion effects. For example, at an anode tilt of about 15° from horizontal, large bubbles at the anode surface were seen to experience significant strain and resistance. A suitable angle of inclination has been found to be in the range of about 2 DEG to 15 DEG, although the angle of inclination can be increased or decreased by more or less depending on furnace conditions. The preferred slope is about 5-10 degrees from horizontal, more preferably about 6-8 degrees. The most desirable cathode tilt was found to be about 8°. The cathode surface 6 is made of
The aluminum is coated with a conductive aluminum material such as _TiB2 . Although a surface comprising titanium diboride is preferred, the use of other aluminum-wettable refractory hard materials such as titanium carbide, zircon carbide, zircon diboride, or mixtures thereof is also contemplated. Suitable cathode surfaces and coating compositions providing such cathode surfaces are described in the patent application Ser.
U.S. Patent Application No. 395343, No. 395344, and No. 395344 dated May 9
No. 395345 and U.S. Patent Application No. 22 July 1982
No. 400762, No. 400772, No. 4000773, and others. The aluminum flows down from the inclined surface of the cathode into the lateral collecting sump 5, thereby reducing the possibility of reverse reactions reducing current efficiency, speeding up the extraction operation, and removing it from other furnace operations. Make your work independent. This falling thin film of aluminum is not sensitive to induced magnetic fields, which is a major improvement over the use of conventional molten aluminum beds. Furthermore, the controlled bath movement does not impede the aluminum flow. The bubbles 8 moving upward along the lower sloped edge of the anode are evacuated as necessary to prevent the bubbles from becoming too large, i.e. covering large parts of the anode surface or extending too much into the anode-cathode space. He will be prevented from leaving. However, excessive evacuation should be avoided so as not to significantly reduce the required driving force for the bath flow into the narrow anode-cathode space.
Exhaust is carried out through grooves in the anode, or through vent pipes, passages, holes, etc. Vent holes can also be created by inserting pipes into the paste or plastic area of the Soderberg anode or by sintering them into the pre-baked carbon anode. It has been found that a non-uniform distribution of vents on the anode redistributes the bath flow within the furnace to correct for the non-uniformity of the bath flow within the furnace. Alumina is added to the bath from hopper 19 to replenish the spent alumina bath. Of course, other bath constituents are also added at the same time. Arrows 10 indicate the flow of the alumina-rich bath through the narrow anode-cathode space 7 guided by gas bubbles 8;
The falling arrow 13 shows the flow of aluminum metal as a thin film falling into the sump 5. FIG. 2 is an end view of an aluminum reduction furnace using a single tilted cathode. The single sloped cathode face 6 allows the use of straight collector bars 4;
The distance between the cathode top surface and the collector bar can be kept constant. Although there is some voltage reduction advantage if the collector bars exit from both sides of the furnace 22, there is no need to extend the collector bars 4 out if it is more advantageous to deepen the aluminum sump 5. . This furnace configuration allows the use of more or less conventional construction and mounting techniques for collector bar installation. The slope of the cathode surface 6 must be sufficient to allow efficient movement of gas bubbles along the formed anode surface and to cause aluminum to flow down the cathode surface. As shown, excess anode gas bubbles 8, which could unduly distort the metal film flowing over the cathode surface, are exhausted to a gas collection system 3 through a vent 2 at the anode. The use of a mechanical bath cover 18 ensures complete control of the bath surroundings and reduces air pollution. Alumina is added to the bath by a hopper 19 at the interelectrode gap outlet 21. An automatic controller 20 can also be used to accurately refill the bus. This addition of alumina and other bath components has several advantages. First, after leaving the electrolytic zone, the bath partially consumes the molten alumina, providing the most ideal environment for rapid alumina melting. If some alumina settles to the bottom of the side storage tank 24, the absence of a metal pad or sump covering the precipitated alumina mat helps the alumina mat dissolve into the flow path. moreover,
The return channel of the bath provides extra melting time and acts as a trap for unmelted alumina to ensure that only the unmasked bath is circulated to the ACD gap. The addition of alumina to the side of the metal sump 5 of the furnace creates the possibility of the formation of mats on the underside of the metal in the sump, causing the sump to overflow and create a direct metal electrical short between the anode and cathode. Since this may occur, this arrangement avoids this. FIG. 3 is an end view of a flow-through single inclined cathode aluminum reduction furnace. As shown, the cathode angle is sufficient to efficiently move the anode gas bubble 8 upward along the surface of the formed anode, and to allow aluminum to flow down the cathode surface as indicated by arrow 13 into the aluminum sink 5. It is. Gas vents 2, which may be necessary, are provided in the anode and can also take the form of grooves or vents. The anode itself
A variation of Soderbrg or pre-baked carbon. Preferably, the aluminum sump 5 is led to a continuous extraction system to reduce disturbances to the furnace, such as changes in bed thickness. Advantageously, the vent 2 and the empty bath cavity are connected to a gas collection system. In the illustrated flow-through reactor, the electrolytic bath is replenished externally and flows from the bath inlet 9 through the passage indicated by the arrow 10, through the anode/cathode space 7 and to the bath outlet 11, where the electrolyte is electrolyzed with alumina. is partially consumed by FIG. 4 shows a dual inclined flow cathode or V-shaped cathode furnace. In this configuration, the illustrated pre-baked anode 1 is replenished by a point feeder 27 illustrated at the higher end of the cathode slope.
Replenishment can also be done in between. The aluminum metal stream 13 flows down the cathode surface 6 and enters a sump 5, which is advantageously connected to a continuous extraction system. Anode gas bubbles 8 are formed on the bottom of the anode so that the bath material flows 10.
induce. FIG. 5 is a plan view of a single furnace incorporating a furnace feed mixing section. In FIG. 5 it is illustrated that the anode 1 has a groove-shaped vent 2. In FIG. Along with the cathode surface 6, a sump 5 is illustrated which supplies aluminum to a continuous metal extraction system 14. The direction of net flow of the aluminum metal is shown by arrow 13 and the direction of net flow of the cryolite bath is shown by arrow 10. After passing between the anode 1 and the cathode surface 6, the partially spent bath flows through the outlet 11 to the furnace feed mixing zone 16 where it is replenished with alumina 1
2. After alumina replenishment, the cryolite bath is recirculated through the electrolytic furnace in the direction indicated by arrow 10. Although an optional magnetic bath circulation pump 17 is shown in the drawing adjacent to the bath inlet 9, the use of such a pump is considered unnecessary under most circumstances. A baffle plate 15 may be placed in the return passageway 28 to prevent backflow of the bath and circulation disruption in designs incorporating the furnace feed mixing zone 16. FIG. 5 is a plan view of a multiple furnace system consisting of four individual electrolysis furnaces operated in tandem. As shown, the main system includes a continuous aluminum extraction system 14,
Keeps the aluminum water level constant, thus making the heat balance and bath water level more constant. Conventional discontinuous metal extraction could be used, but arrangements would be necessary to compensate for changes in bath water level. It includes a bath surge tank, a lower alumina feed mixing tank, and a heat balancer. As shown in FIG.
6 and circulates through the anode-cathode gap in the direction indicated by arrow 10. Magnetic circulation pump 17
Use of is optional. The generated aluminum flows down and enters the aluminum sink 5 as shown by the arrow 13.
to the continuous metal extraction system 14 in the direction shown. The baffle plate 15 prevents the bath in which the alumina has been consumed from flowing back. It is easy to incorporate an alumina feed melting system into either a single furnace system or a more economical multiple furnace system. One feature of the system that allows it to operate under a controlled atmosphere is the sealed furnace top, thereby reducing reactant problems, air pollution problems, and heat loss. A sealed furnace top is not essential. Adding a non-consumable anode surface to this design allows for a self-contained and unattended multi-electrolysis unit within a single container. FIG. 7 shows an aluminum reduction furnace using a single tilted cathode. As shown, a continuous aluminum extraction system 14 is shown which removes aluminum from the furnace via a sump 5. This serves to keep the aluminum water level in the furnace constant and to make the heat balance and bath water level in the furnace more constant. If necessary,
An adjustable partial baffle 25 is used to limit the overall bath circulation speed within the return flow path 28.
Baffles made of non-corrosive insulating material, such as silicon nitride, provide the necessary restraint. Although not shown, the necessary replenishment equipment may be of any suitable type as described above. As mentioned above, instead of a single tilt cathode furnace, a dual tilt cathode furnace may also be utilized within the scope of the present invention. For example, the inverted V-shape of FIG. 1 or the V-shape of FIG. 4 allow metal aluminum to flow downwards into an aluminum sump. However, the single-inclined cathode furnace shown in Figure 2, which slopes upward toward the alumina feed side of the furnace, not only simplifies the construction of the furnace itself, but also simplifies cleaning and other maintenance of the crucible chamber of the furnace. This is considered the preferred embodiment. FIG. 8 is a plan view of an aluminum reduction furnace using double row single tilted cathodes, which is another embodiment of the present invention. The flow of electrolyte is indicated by arrow 10,
The circulation is shown from between the anode 1 and the cathode surface 6, through the area where replenishment 12 takes place, and through the ACD of the adjacent anode-cathode pair tilted in opposite directions. The furnace centerline 26 effectively demarcates the two oppositely angled cathodes. To ensure smooth and continuous circulation, there are baffles 15 and partial baffles 25, respectively, to prevent backflow of the bath and control the net bath circulation flow rate. All other elements in the figure are the same as described above. It should be noted that this configuration is an alternative to the V and inverted V configurations described above and is more suitable for use with pre-baked anodes than Soderberg.
Another configuration particularly suitable for pre-baked anodes is one in which adjacent cathodes (and anode surfaces) are inclined in the same direction but at different angles. It should be noted that although the figures do not show the presence of side protrusions and/or skins of the solidifying bath material, the invention also applies to furnaces in which they are present. The pumping action of gas bubbles working upward under a slightly inclined surface was first demonstrated in hydraulic analog experiments, with the liquid confined in a thin layer between the upper and parallel lower surfaces. In this experiment, the kinematic viscosity coefficient is very close to that of molten cryolite.
Performed using room temperature water. The surface slope is 2.5°,
The gap between the top and bottom surfaces was set to 2.2 cm, and air bubbles were generated on the porous top surface by pressurizing with air at a flow rate matching the gas generation amount of a typical high-current aluminum electrolytic furnace. As a result, the net or average liquid velocity between the simulated anode and cathode was determined to be in the range of 5-10 cm/sec. These rates were calculated to be more than sufficient to feed alumina at the rate necessary for proper furnace operation when the alumina concentration in the cryolite is normal. Although this experiment did not require the surface slope to overcome the opposing bubble forces present in an operating horizontal anode aluminum electrolyzer, e.g. due to magnetic influences, the principle of gas-driven bath circulation was clearly demonstrated. It was done. It has been found that the resulting bath circulation is controlled by the balance between the pumping efficiency of the gas bubbles in the ACD gap and the back pressure or resistance to bath flow through the return channel. Further results and data were obtained from a 1:20 scale aluminum electrolytic furnace and hydraulic analog model. A hydraulic analogue model was created that simulates a gas-generating anode surface (approximately 1 m ² ) above a sloping empty cathode surface. This apparatus simulated a typical "falling cathode" aluminum reduction furnace design in which the operating anode is above the sloped falling TiB ₂ cathode surface. Water model studies were conducted using room temperature water. Because the observed fluid had a Reynolds number in the turbulent range (>5000), the flow patterns and velocities observed in the water model were similar to those expected in a full-scale reactor. In this region, flow is primarily controlled by the physical dimensions of the flow path and is independent of fluid properties (eg, viscosity). Alundum perforated plate (pore diameter approx.
A pressurized air chamber with a bottom plate made of 20μ) was used to simulate the gases (e.g. Co ₂ ) generated at a real anode. Gas velocity 0.1778 through the perforated plate to simulate an anode current density of 0.68 amp/ ^cm2
cm/sec was used (gas velocity was corrected for temperature and static pressure). Simulated currents up to about 1.4 amp/cm ² were experimented with this model. The design of the model whose side view is shown in FIG. 9a simulates one half of the furnace shown in FIG.
The plan view of the model simulates the furnace shown in FIG. The wall at the higher end of the anode corresponds to a vertical plane passing through the central groove or vent 2. The figure shows ACD,
The relationship between BFL, φ, h, ho, lower flow path width and upper flow path width (defined later) and anode 1, cathode surface 6 and bus 29 is illustrated. The following definitions are used to describe the water model and commercial scale aluminum reduction reactor in contrast to the required bubble flow upwardly below the sloped anode surface. BFL - Anode surface dimension in the direction of bubble flow (BFL = 122 cm for most tests). BFW - Anode surface dimension perpendicular to the direction of bubble flow (BFW = 61 cm for most tests). ACD - Vertical distance between inclined anode and cathode planes corresponding to number 7 in Figure 2 (model test
ACD was changed from 1cm to 5cm). Cathode tilt (φ) - the angle between the upwardly sloping cathode surface and the horizontal plane (the cathode tilt was varied from 0 to 15° in the model tests), h - the vertical immersion depth of the anode from the liquid surface to the anode surface ( (h varies along the BFL dimension line of the anode according to the degree of cathode tilt), ho - the minimum anode immersion depth, i.e. the anode immersion depth at the high end of the anode (most models
(ho = 10 cm), the required bath flow typically passes through four different types of channels or passageways. Namely: 1) ACD gap between anode and cathode. The introduction and flow of anode gas here generates the force necessary to maintain the required bath circulation within the furnace. 2) An upper channel into which the bath flow flows from the ACD gap, which corresponds to the liquid storage tank 24 in FIG. Here the flow is directed to one or both sides of the anode. The majority of the anode gas is discharged from the bath in this flow path, providing ideal bath disturbance for the melting of the alumina fed to the furnace. This channel can be along the sides of the furnace, as shown in FIGS. 2 and 7, or in the center of the furnace, as shown in FIG. The definition of the dimensions of the upper channel is as follows. Depth = depth of bath in channel (ho in water model)
+Vertical component of ACD. The vertical component is effectively equal to the ACD at slack angles). Width = horizontal dimension in the same direction as the bubble flow of the ACD gap. Length = horizontal dimension to BFW of anode + width of return channel. 3) Return channel. The bus flows from the upper flow path.
Channels leading to corresponding downward channels present along the anode edge entering the ACD gap. An example of a return flow path is indicated by the numeral 28 in FIGS. 5-7. In its simplest form, and the model used in this study, the return flow path is as shown in Figure 7.
The depth or width of the return channel can also be varied to vary the flow resistance (similar to the function provided by partial baffles). In the water model and similarly designed reactors, the return flow path dimensions are defined as follows: Depth = Depth of the bath in the channel (if the channel is narrow and the working anode forms one or both sides of the channel, the effective channel depth is the actual depth minus the ACD) (This is because the upward flow in the adjacent ACD gap tends to stagnate a layer of similar thickness at the bottom of the return channel. This reduced return channel depth was used in the water model study) . Because of the sloping cathode, the depth of the return channel increases as the bath flows from the upper channel to the lower channel. Width = horizontal width of the channel perpendicular to the flow within the channel. Length = horizontal distance between the upper and lower channels (in water models this is approximately equal to BFL). 4) Downward flow path. Carrying the bus from the return channel,
A flow path that completes the bus circulation loop by distributing the bus along the anode edge where the bus enters the ACD gap. The dimensions of this channel are defined similarly to the dimensions of the upper channel. Most cell designs and water models have a basin or trough at the bottom of the channel to collect the aluminum metal as it flows down the sloped cathode. Such a receiver is shown in Figures 1 and 9a. In all cases, the bath flow rate Q in the model and full-scale furnaces is defined as the total volumetric flow rate of the bath stream entering the ACD gap from the lower channel. The water model was allowed to be modified in a variety of ways to simulate a real electrolysis furnace and evaluate the effects of changes in specified dimensions or operating conditions. A device was applied to the model to simulate an adjustable partial baffle (25 in Figure 7). The slope of the cathode
To understand the effect on the bus flow in the ACD gap, the anode and cathode slopes of the model were adjusted from 0 to
The angle was changed to 15°. The ACD gap was varied from 1 to 5 cm in water model studies. The gas flow rate was variable to simulate different current densities. Colored dye was injected into the ACD gap and return passage to observe and measure the fluid flow in the model. The formation, or conformity, of the anode to the downwardly slanted cathode can be achieved even in laboratory electrolytic test furnaces, according to Kaiser Aluminum and Chemical Corporation's report to the Department of Energy (DOE Contract No.
No. EY-76-C-03-1257, Final Report, ``Energy Savings Through the Use of Improved Aluminum Reduction Furnace Cathodes,'' November 30, 1977). Figure 9b shows the required flow of gas bubbles 8 in the ACD,
and the direction 10 of the bath flow. It is clear from this figure that too high a bath velocity on the cathode surface in the same direction as the anode surface can impede the metal flow down on the cathode. Therefore, although unidirectional flow is desirable, excessive flow velocity at the cathode surface should be avoided. Water modeling studies have revealed that under certain design conditions three different undesirable phenomena can occur within an ACD. Namely: Γ reflux. Backflow occurred when the bath flowing upwardly along the anode bottom surface reversed its direction of flow and flowed downwardly along the cathode surface as shown in Figure 9c. In extreme reverse flow conditions, the bus forms multiple small vortices within the ACD, as shown in Figure 9d, resulting in no or very little net new bus entering the ACD. Insufficient supply of new baths in practical electrolysis furnaces leads to persistent anode effects. ΓAerotsuk (gas occlusion). Aerodynamics occurred when the anode gas was removed too slowly and the ACD became filled with large stagnant gas bubbles. In this state, the bus flow becomes insufficient. When this condition occurs in an operating furnace, such gas bubbles inhibit the local electrolytic reaction and increase the anodic current density in that part of the anode, eventually resulting in a loss of anode polarization voltage and It can also lead to the onset of effects. ΓExcessive bubble thickness. Under certain conditions, it occurred over a wide area of the bath in contact with the cathode surface. When this condition occurs in an operating furnace, the current efficiency drops sharply due to the rapid back reaction between the CO ₂ bubbles and the molten aluminum metal on the cathode surface. The above phenomena and the net bus flow in the ACD gap and return channel were studied as a function of cathode tilt, ACD, fluid resistance in the return channel, and simulated anode current density. Although all of the observed flow properties are controlled according to the general fluid mechanics principles described above, a quantitative description thereof has not been established to date. To illustrate a representative example of the application of the present invention to the design of an energy-saving improved aluminum reduction furnace, the general furnace specifications in the following table were used in the water model and later examples. In practice, select these general furnace specifications to suit your practical application and use the teachings of the present invention to determine important furnace design specifications. Heat balance and investment efficiency calculations have been made indicating that it is desirable to operate "falling cathode" furnaces at higher anode current densities than are typical used in the industry today.

【table】

[Table] Furnace according to simulation parameters Furnace Number of furnace return passages 1 4 4
h _O , minimum depth of anode immersion 10 10 10
(cm)
The design parameters used in the water model study are included in the table for comparison with the actual reactor. The schematic furnace design under consideration is shown in Figure 1, with a return passage connected to a single central upper passage, and a split VSS
The furnace has two outer downward passages at each end of the anode (each half of the furnace is similar to the furnace shown in FIG. 7). Figures 10, 11, and 12 illustrate the effects of cathode tilt, ACD, and return channel flow resistance R _f on the net bus flow across the ACD gap at a simulated anode current density of 0.68 amp/cm ² . In these figures, cases where backflow was observed at the cathode surface are indicated by broken lines. In all cases, this is associated with a significant reduction in net bath flow, which becomes more excessive as the cathode tilt becomes slacker. Similar results were obtained for other simulated current densities. A furnace parameter design limit diagram created from this data is shown in FIG. The use of this diagram for the design of a furnace with controlled bath flow is detailed in the example below. The general properties of this diagram are as follows. At a particular operating condition, the flow resistance is related to the width of the return channel. Regarding the width of each return flow path, Fig. 10,
A contrast plot of the data at constant flow rates from the diagrams corresponding to FIGS. 11 and 12 achieves the flow rate selected as suitable to feed the furnace, as calculated by the method defined later. Therefore, a relationship between cathode tilt and ACD is derived that must be satisfied at a certain anode current density. The relationship between this cathode tilt and ACD is represented by the curve in FIG. Thus, for each return passage width, this curve represents the conditions for achieving proper alumina supply. These curves are constrained by boundary conditions defined by undesirable hydrodynamic conditions as described above. That is, the prohibited region due to excessive cell thickness limits the operating region in the lower ACD dimension, ie to the left of the diagram (FIG. 15). Excessive anode (and cathode) tilt resulting in excessive anode immersion depth creates a boundary upper limit at a tilt of approximately 15°. The desired operating region is the region somewhat above the corresponding curve representing the width of the chosen return path, so as to reduce the ACD without violating the region noted as ``Bubble Thickness Limit'' in Figure 15. Ru. Operation in this region ensures adequate alumina supply with a sufficiently large value of Q, while at the same time avoiding excessive bubble thickness and the associated loss of current efficiency, as well as excessive velocity of the bath at the cathode surface (and hence aluminum flow down). (inhibition of The following example illustrates the construction and use of such a parameter diagram. Example 1 - Selection of Return Channel Dimensions The bath flow rate Q (cm ³ /sec) must be sufficient to provide the necessary alumina to sustain the electrolytic reaction to prevent anode effects in the furnace during operation. Must be. At a maximum current efficiency of 100%, the minimum flow rate Q required for a practical aluminum reduction furnace is given by the following equation. IQ (cm ³ /sec) = 0.008 [Total furnace current (A)] / Δwt% Al ₂ O
₃ However, Δwt% Al ₂ O ₃ is the difference in weight % of Al ₂ O ₃ in the bus entering and exiting the ACD gap, that is, Δ alumina. The constant 0.008 is derived from the Faraday formula: . 0.008 = (Molecular weight of Al ₂ O ₃ ) (100) / (Faraday/mol) (Faraday constant) (Bath density) For convenience, Q is the volume of bath per equivalent area of the anode /
It can also be expressed in seconds. Therefore, assuming a minimum anode current density of 0.5 A/cm ² and a maximum bath consumption per ACD pass of 5%, that is, Δwt% Al ₂ O ₃ = 5,
From the formula, the minimum allowable value Q=8×10 ^-4 cm ³ /sec/cm ² (anode area) is obtained. Furnace performance and thermal stability are uniform throughout the bath.
Enhanced by keeping Al ₂ O ₃ . For the above mentioned furnace, the calculated minimum bath flow is 202,4000 for water model furnace, furnace and furnace respectively.
and 12000cm ³ /sec. Although the minimum bus flow Q is theoretically sufficient, operational problems (e.g.
Excessive anodic effect, high effective bus resistance, and ACD
In practice, a value higher than this minimum value should be used to prevent excessive voltage (due to excessive bubble volume in the gap). The bath flow will vary somewhat depending on furnace operating conditions (eg, due to bulges, crusting, etc.) and may be less than the theoretical bath flow. Therefore, this minimum bath flow rate was multiplied by a design factor of 4-5 to obtain desired bath flow rates of 890, 20,000, and 60,000 cm ³ /sec for the water model and example reactor, respectively. Approximately 450, 10000 and 30000
cm ³ /sec is considered appropriate, but water model data indicate more reliable and stable bath circulation at the above desired Q value than the minimum theoretical Q value. According to water model data, the flow resistance characteristics of the return channel are an important element of the present invention. As the return path becomes more restrictive (a result of trying to maximize the anode area within the furnace), the sensitivity of the bath flow to changes in the flow resistance characteristics of the return path increases. Since there are many furnace designs with comparable effective flow resistance characteristics in the return passages, it is beneficial to develop a simplified hydraulic model to provide general design criteria for the return passages. The head loss h _l due to flow resistance in the return channel is given by the following well-known formula: . h _l = 2f _f LV ² /D _eq g where f _f = Fanning friction coefficient V = velocity L = length (approximately equal to BFL in the water model) D _eq = equivalent hydraulic diameter g = weight constant Friability The two coefficients are composite values that reflect the differences at the bottom and sides of the channel. For a single open channel. D _eq = 4 (cross-sectional area)/wetting parameter. D _eq = 2hw/2h+w where w = Width of the return channel h = Depth of water at any point in the channel (reducing the ACD correction amount in the water model) h = h _p +xsinφ h _p = Above the anode h at the end (i.e. minimum anode immersion depth) x = length of the return channel so distance from the upper channel φ = cathode inclination angle Since the observation velocity in the return channel changes with the change in water depth, It is better to use a bath volumetric flow rate that is independent of channel dimensional changes. The bus volumetric flow rate Q is given by velocity x flow path cross-sectional area. In other words, ． Q=Vhw, and further. V=Q/hw The Q value used in these formulas is the net effective average Q of the channel, determined by measuring the time required for the injected deeply colored dye to pass through the channel. represent The water head loss can be expressed by the following formula: ． h=K _f Q ² R _f where R _f represents the flow resistance geometric term that depends on the physical dimensions of the return path, and K _f represents the flow resistance geometric term that is less sensitive to physical enlargement. It is a characteristic coefficient and can actually be determined by the following formula: . R _f (furnace flow path) = (2h + w) L/h ³ w ³ and . K _f = f _f /2g Since the value of h changes in the inclined return passage,
R _f is actually calculated as an integral function of h over the channel length L, with the width w being constant throughout. The four different return path flow resistances used in the water model study are shown in Table 2. The furnace design in the present invention is similar to a pumped fluid loop in that the gas bubbles in the ACD gap perform the pumping action to drive a bus around the ACD gap-return circuit. A similar pump efficiency diagram shown in FIG. 13 is used to explain the circulation characteristics of a "falling cathode" furnace. Pump efficiency increases with cathode tilt and/or decreases
Curves 1, 2, and 3, which reflect ACD, improve in this order.
The flow resistance increases in the order of slope lines a, b, c reflecting the increasing length and/or decreasing cross-sectional area of the return flow path. The observed flow rate is determined by the intersection of the relevant pump efficiency and flow resistance curves. Thus, the pumping/flow conditions can be represented by overlapping vertical lines on the pump diagram at the minimum flow rate required to meet the Al ₂ O ₃ supply requirement for the electrolytic reaction in the ACD. A sudden change in the slope of some of the pump efficiency curves at low flow rates is associated with backflow within the ACD. Table 2 describes the effect of return channel flow resistance on ACD at a constant cathode tilt and ACD. At low flow rates (Q), the table shows that a drag coefficient of 2.7 x 10 ^-1 cm ^-4 is outside the limits of the invention at the specified conditions. This is confirmed by the observation of a predominant reflux condition within the ACD. For the same specified conditions, the R _f value of 2.7 × 10 ^-2 cm ^-4 is within the limits of the present invention, but is too large to achieve the desired flow rate of 890 cm ³ /s in the water model, and some ( (slight) regurgitation. It should be noted that in some cases, a slight backflow may be advantageous, since a small backflow tends to encourage rather than hinder the flow of aluminum down the cathode surface. Scaling up the water model to a practical aluminum reduction reactor presents a number of alternative designs within the scope of this invention. The basic requirements that must be met in a production furnace are: 1) The bath flow rate Q entering the ACD is sufficient to support the electrolytic reaction. The minimum required bus flow rate is given by the formula: . Q _nio (cm ³ /sec) = 0.008 [Furnace total current (A)] / Δwt
%Al ₂ O ₃ Δwt % Al ₂ O ₃ is 0.2 and in a 100KA furnace,
The minimum bath flow is 4000 cm ³ /sec.

[Table] 2) The head of the gas-induced pumping action (reducing flow restriction losses in the ACD) should be equal to the combined head h of the return, upper and lower channels. At a given slope, ACD, current density, and anode length, the water head h generated in the water model and full-scale reactor should be similar. 3) The ability to use single or multiple return channel designs with appropriate dimensions to obtain the required effective flow resistance R _f term. Furnace R _f
The term is estimated from R _f of the water model using the following equation: XI. R _f (furnace) = R _f (model) × [f _f (model) / f _f (
Furnace)] [Q (Model) Q (Furnace)] ^2However , changes in the abrasive coefficient and total furnace current are taken into account. If the dimensions of the upper and lower channels are equal to or smaller than those of the return channel, the varying mass flow rates in the upper and lower channels must be included in the calculations. In the large and low predicted reactor design and water model, the dimensions of the upper and lower channels are large enough to ensure that their small head loss hl
is negligible compared to the head loss in the return flow path, which is more constrained. The net pumping efficiency data reported includes the effect of ACD flow resistance. The R _f (furnace) term thus provides the necessary design basis for the combined effects of return channels or multiple return channels in full-scale furnaces. Since a "falling cathode" furnace design may include multiple (n) return channels, the effective value for each return channel is given by: XII. R _f (furnace flow path) = R _f (model) [f _f (model)/f _f (furnace)] [nQ (model)/Q (furnace)] ^2However , each Q is defined as above. Now, the value of R _f (furnace) is calculated as the sum of all (n) return channels from the following formula: . 1/[R _f (furnace)] ^1/2 = _o 〓 ⁱ⁼¹ (1/[R _f (furnace channel i)] ^1/2 ) and for n equivalent channels: . R _f (furnace flow path) = n ² R _f (furnace) where R _f (furnace) is calculated from the following formula: . R _f (furnace) = R _f (model) (6.25×10 ⁸ )
[Δwt%Al ₂ O ₃ /furnace current] In ^{the 2} equations, R _f (model) is approximately 230×10 ^-3 /cm
⁴ to approximately 0.2×10 ⁻³ /cm ⁴ . For example, assuming that the attrition coefficients (as specified in Table 1) of the water model and the example furnace are equal:
The desired R _f term for each return channel is given by Equation XII, and its width is calculated from the corresponding equation. Furnace example R _f (furnace flow path) = R _f (model) [4 (890)/20000] ² R _f (furnace flow path) = 0.032R _f (model) w (furnace flow path) = 20cm Furnace example R _f (furnace flow path) = R _f (model) [4 (890) / 60000] ² R _f (furnace flow path) = 0.0035R _f (model) w (furnace flow path) = 52cm Above value of return flow path width is the design factor 5 for the desired bus flow Q and the desired R _f (model) = 6.2×10 ^-3
is based on the value of If the operation of the example furnace is determined to be stable enough to reduce the bath flow design factor to 3, the calculated value w(furnace flow path) would be reduced to 33 cm. An alternative way to reduce the calculated value w(furnace flow path) to 3 without reducing the bath flow design factor is to increase the anode immersion depth by 8 cm (eg, h _p =18 cm). These alternatives demonstrate the utility of the teachings of the present invention in the design of "falling cathode" furnaces. Figure 14 shows the desired Q value of 890 for this model.
Represents the R _f (model) value (as a function of cathode tilt and ACD) required to obtain cm ³ /sec. R _f values for other Q values can be determined from water model data as shown in FIGS. 10, 11, and 12. Example 2 - ACD Selection Furnace energy efficiency is improved by reducing ACD (and thus bus voltage losses) without significant loss in current efficiency. If the ACD value is relatively large (eg, 4 cm or more), there is a strong tendency for backflow conditions to occur within the ACD gap. If this reversal occurs, the net Q value will be significantly reduced. When the ACD value decreases,
The required bus flow through the ACD gap is facilitated and higher flow resistance in the return flow path is tolerated. However, at very small ACD values (approximately 1 cm), secondary effects between the air bubble and the vicinity of the empty cathode surface tend to slow down the required bath flow (eg, as a result of increased flow resistance in the ACD gap). Figures 10, 11 and 12 show that the maximum bus flow is
This shows that it does not necessarily occur in ACD. Increasing the cathode tilt from 2° to 5° decreased the measured maximum bubble thickness from about 1.0 cm to about 0.5 cm, and the net liquid flow rate increased accordingly. As the slope increases beyond 5°, the bubble thickness is observed to remain fairly constant, and the bubbles are observed to move more slowly, assuming the characteristic shape described above. It was determined that changing the ACD did not significantly affect the observed bubble thickness. If air bubbles protrude over a large portion of the ACD, the CO ₂ will react adversely with the aluminum metal thin film on the cathode, significantly impairing the current efficiency and greatly increasing the electrical resistance of the bus. For these reasons and practical considerations, the preferred ACD
is approximately 2 to 4 cm, with a desirable range of approximately 2 to 3 cm. Example 3 - Cathode Selection Bath flow when the cathode tilt is equal to or greater than 5° in all cases except when the return flow path is severely restricted (Figure 10) or the ACD is greater than 4 cm. Q exceeds the desired 890 cm ³ /sec. Increasing cathode tilt can result in excessive backflow and aerots (excessive gas bubble stagnation) in the ACD gap.
Helps overcome. On the other hand, too steep an angle may impede the flow of molten metal on the cathode surface. Back pressure due to restriction of the return flow path can also be offset by increasing the cathode tilt. However, a large cathode slope creates an excessively upslope bath flow in the ACD gap, impeding the flow of aluminum from the cathode face, and creating a practical problem in that the anode immersion depth within the bath varies greatly. It's possible. When considering also the variation in cell thickness disposed in Example 2, the preferred cathode tilt is in the range of about 5-11 degrees, with about 8 degrees being the most preferred tilt. Example 4 - Furnace Parameter Design Limit Diagram The construction and use of the furnace parameter diagram described above will now be described in detail. The parameter constraints used to construct this diagram are shown in Table 3. For convenience and to help visualize the furnace design, the return channel flow resistance R _f is shown in Figure 15 in the form of the corresponding channel width (channel length 121 cm, anode current density 0.68 A/cm). ² , assuming a minimum bath flow rate of 890 cm ³ /s and a minimum flow path height h _O = 10.0 cm). The most desirable set of reactor design parameters for the water model under the specified conditions is: Cathode tilt = 7.5° ACD = 2.5 cm R _f (return channel) = 6.2 × 10 ^-3 /cm ⁴ bath flow Q = 890 cm ³ /sec The range of critical reactor hydrodynamic design parameters described below is based on: 1) measured data and observations from water model simulation studies;
2) presented data analysis; 3) scale-up to commercial scale reactor;

[Table] Listen.
Practical considerations for measuring. A set of parameter ranges is shown in Table 4 for a water model with a BFL of 122 cm and a BFW of 61 cm. Parameters for the corresponding commercial scale aluminum reduction furnace are shown in Table 5. The criteria of the present invention can be applied to several types of furnaces, such as those shown in FIGS. 1-7.
In particular, the furnace can also consist of an anode that spans the width of the furnace as one continuous piece. An example of this anode is the anode shown in FIG. 2 minus the central vent 2, where the bubble flow length (BFL) is approximately equal to the overall width of the anode. This type of furnace design results in a narrow furnace (approximately 122 cm anode width in the most desirable case). Another type of furnace would be that shown in FIG. 2, which included a central vent 2 in the design. A central vent evacuates air bubbles that accumulate under the cutmost half of the anode that the bath flow traverses. In this case, the BFL is defined as half the overall width of the anode. To prevent excessive air bubble accumulation within the ACD gap, which can increase voltage losses and reduce current efficiency,

【table】

[Table] It is necessary to exhaust the anode gas. In this type of furnace, the most desirable overall working anode width would be approximately 244 cm, which is the more practical furnace width. In commercial furnaces, it is desirable to operate at the highest possible anode current density without loss of current efficiency (to minimize capital and labor costs per pound of aluminum produced).
At the low ACD values used in "falling cathode" furnaces, these high current densities are advantageous in helping maintain the thermal balance of the furnace. However, at high anodic current densities, gas bubbles accumulate faster and grow larger than at lower current densities. Thus, to counteract the deleterious effects of bubble accumulation in the ACD, the desired BFL (anode dimension in the direction of bubble flow) is decreased in proportion to the increase in anode current density. For example, increasing the anode current density from 1.0 A/cm ² to 2.0 A/cm ² requires decreasing the desired BFL from 122 cm to 61 cm. The preferred furnace design for low anode current densities in an inverted V cathode furnace is shown in FIG. Higher anode current density (higher than about 1.3A/ ^cm2 )
To increase the overall working anode width desired for
The concept of one gas vent for discharging anode gas can be expanded to include multiple gas vents to further increase the width of the anode. The vent can be present as a groove between adjacent anode bodies or as a row of appropriately spaced holes through the anode body. Yet another approach is to use a furnace design typified by the design shown in FIG. In this case, the central channel serves multiple purposes, such as as a gas vent and as an upper channel that carries the bus exiting the ACD gap to the return channel. In this case, the BFL is defined as half of the total operating anode width, which yields a more practical sized commercial reactor. It will also be appreciated that a central metal collection trough or receiver can be provided to reverse the anode tilt shown in FIG. In the latter case, the upper passageway would be located along the outside of the furnace, as in FIG. These furnace design concepts can be combined to
Alternatively, it is also possible to have a multi-element furnace consisting of a number of furnaces connected by metal flow, or a single-cavity furnace comprising adjacent multiple units of the above-mentioned furnaces. In all such furnace designs, the return flow path dimensions are calculated according to the teachings presented in Example 1. The first step is to select the required bath flow Q based on the scale of the furnace being designed, and then to determine an appropriate flow resistance term R _f using the developed relational expression. The head loss formula is then applied under specified conditions to calculate a set of preferred return passage design options. Thermal balance and other furnace design and operating criteria are then used to select one of the hydraulically equivalent return path designs. The final selection of this return flow path design is made with consideration to the impact on the critical bath flow. Example 5 - Falling Cathode Experimental Reactor Test data from a laboratory scale aluminum reduction reactor with a tilted TiB ₂ cathode demonstrates the effectiveness of using a water model to simulate the hydrodynamics of a commercial aluminum reduction reactor. A 61 cm x 122 cm x 46 cm deep graphite box filled with an approximately 600 lb (273 Kg) cryolite bath was heated in a sealed electric furnace. After the bath is melted, the test cathode and anode are lowered through the filling port at the top of the furnace, and the first
It was placed in the cryolite bath shown in Figure 6.
The automatic Al ₂ O ₃ supply and anode drop device allowed continuous electrolytic testing for up to 3 weeks without interruption for anode replacement. Under all electrolytic conditions, the initial slope of the lower anode plane (usually horizontal) is between 1 and 5
After days of electrolysis, the slope of the cathode surface directly below changed.
The non-uniformity of the anode surface shown in FIG. 16 corresponds to the behavior of bubbles seen in the water model. Starting edge of the anode surface in the water model (corresponding to area A in Figure 16)
In this case, the bubbles tend to move very slowly and erratically, resulting in very uneven coverage of the anode surface. This will result in uneven consumption of anode carbon. As the bubble moves upward along the anode surface, it gains speed and creates a more uniform flow along the anode surface. Electrolysis under this flow condition results in a more uniform consumption of the anode carbon, as shown in region C of FIG. Visual observation of the flow pattern in the experimental electrolysis reactor during operation matches the pattern observed in the water model. Example 6 - Electrolytic Test Data In the experimental electrolytic furnace described in Example 5, furnace voltage data was measured as a function of current density, ACD value, and cathode tilt. The reference voltage in the conventional metal bed cathode is ₃ .
Obtained by replacing with cm deep aluminum floor. FIG. 17 shows the dramatic reduction in furnace voltage noise obtained by replacing the unstable horizontal aluminum bed with a tilted TiB ₂ cathode in an aluminum reduction furnace. In addition to reducing furnace voltage noise (eg, improving furnace stability), the falling cathode experimental reactor exhibited the following characteristics: 1) The diagram of the experimental furnace voltage for a falling cathode reactor is shown in Figure 17 due to the anode effect resulting from insufficient bath flow or from disturbance of the metal bed (which is a result of bubble thickness and excessive backflow).・No peaks caused by cathode short circuit are shown. At low ACD values (e.g. less than 2 cm), conventional furnace voltage control systems (by raising and lowering the anode) required significant operator intervention to perform metal bed cathode tests, whereas tilted TiB ₂ cathode tests required Little or no human intervention was required. 2) A stable tilted cathode test run was performed with an ACD of less than 1 cm. 3) The reactance of the furnace to changes in current density or ACD is rapid and decisive. A new steady state furnace voltage was achieved within about 1 minute compared to the 15-30 minute restabilization time required for metal bed cathodes. This allows for simpler and more reliable furnace automation techniques. 4) Air bubbles were vented from the anode surface primarily along the top edge of the inclined anode surface. When an anode with a horizontal surface was placed in a furnace with a tilted cathode, the initial evacuation of the anode gas bubbles was randomly distributed around all four edges of the anode surface. As electrolysis progressed, the anode surface began to conform to the cathode slope and the gas bubble exhaust became concentrated along the top edge of the sloped anode surface. During the test with the metal bed cathode, only gas bubble evacuation anywhere on the anode surface was observed. In addition to improving the stability of the furnace, electrolytic tests have demonstrated that a significant reduction in the anode-cathode polarization voltage can be achieved by using a "falling cathode". 18th
The figure shows the measured anodic and cathodic polarization voltages as a function of the _TiB2 cathode tilt of the experimental electrolysis furnace. At the desired cathode tilt of 8°, approximately 45% of the anode-cathode polarization voltage
Experimental data show that approximately 0.22 V of furnace voltage savings can be obtained in addition to that achieved by reducing ACD. To date, the advantages of this "falling cathode" furnace have not been reported in the literature. The observed decrease in polarization voltage with increasing bath flow rate (see Figure 12, which shows an example of flow rate as a function of cathode tilt, along with Figure 18) is consistent with electrochemical theory in stirred or fluidized systems. In the laminar flow regime, the lumped polarization voltage is proportional to the square root of the bus speed. The measured voltage in FIG. 18 does not decrease with increasing cathode tilt as expected for laminar flow. This means that the bus conditions in the ACD are in a transition or turbulent region where the lumped polarity voltage changes less sensitively to the bus flow rate. water model・
The data support this hypothesis. Electrolysis experimental data for a tilted TiB ₂ cathode furnace indicate improved furnace stability (e.g. increased current efficiency and reduced anode effect) at narrower ACD values (i.e. reduced furnace voltage values) compared to a horizontal metal bed cathode furnace. reduction),
We demonstrated improved furnace responsiveness to ACD and current changes (e.g., allowing for simpler and more reliable automation of furnace control) and reduced anode-cathode polarization voltages. 1st
The asymptotic portion of the anode-cathode polarization voltage line at cathode tilts greater than 5° in Figure 8 is consistent with the desired cathode tilt range shown in water model studies. The foregoing description of the invention may be modified in various ways by those skilled in the art.
It is evident that changes and modifications are possible and are considered to be within the scope of the invention as set forth in the appended claims.