US2526875A

US2526875A - Method of collecting gases in aluminum furnaces

Info

Publication number: US2526875A
Application number: US33955A
Authority: US
Inventors: Jouannet Robert
Original assignee: Elektrokemisk AS
Current assignee: Elektrokemisk AS
Priority date: 1944-01-21
Filing date: 1948-06-19
Publication date: 1950-10-24
Anticipated expiration: 1967-10-24

Description

Oct. 24, 1950 R. JOUANNET 2,526,875

I METHOD OF COLLECTING GASES IN ALUMI-NUM FURNACES Filed June 19, 1948 2 Sheets-Sheet 1" k I INVENTOR Fofiar/ ffluanm'f ATTORNEY R. JOUANNET 2,526,875

METHOD OF COLLECTING GASES IN ALUMINUM FURNACES Oct. 24, 1950 2 Sheets-Sheet 2 Filed June 19, 1948 V INVENTOR. 170%? jbuamwf BY Patented Oct. 24, 1950 METHOD OF COLLECTING GASES IN ALUMINUM FURNACES Robert J ouannet, La Praz, near Modane,jFrance, assignorto Elektrokemisk A/S, a. corporation of Norway Application June 19, 1948, Serial No. 33,955 In France January 21, 1944 Section 1, Public Law 690, August 8, 1946 Patent expires January 21, 1964 4 Claims. (01. 204-67) In the reduction of alumina to produce aluminum, a low voltage current is passed through a bath of cryolite in which the alumina is dissolved. A carbon or graphite electrode is employed and the electric current both supplies the power for electrolysis and generates heat to maintain the bath in a molten condition. The oxygen driven off from the alumina goes to the carbon electrode where it combines to form a mixture of CO andCOz. In addition, there is always some decomposition of the oryolite so that the gases given off contain fluorine compounds.

The method of handling this gas has for many years been recognized as one of the problems in the production of aluminum. If an open furnace is employed, the pot room becomes fouled by the gases given off, andworking conditions are bad. Enclosing the furnace only gives partial benefit, as it is necessary for the workmen to get at the surface of the bath both for the introduction of additional quantities of alumnia and in order to break up the crust which forms on the surface of the bath. In such case it is impossible to make the enclosure for the furnace air-tight and the gases that are drawn off are highly diluted with air, so that handling them is cumbersome and expensive. The difiiculty of this problem is increased by the fact that great care must be taken to prevent any iron entering the bath (iron is a highly undesirable impurity in aluminum), and the temperature adjacent the surface of the bath is so high that it would bar the use of aluminum shields.

I have discovered that this problem can be solved as follows: One of the standard methods of producing aluminum involves the use of continuous electrodes which are fed slowly into the bath as they are consumed. Such electrodes, being fluid at the top, are provided with a sheath which either moves with the electrode (in such case it is ordinarily made of aluminum) or the sheath is fixed so that the electrode moves through it. In some instances the two types of sheath are both employed. According to my invention, a gas-collecting ring member is supplied which surrounds the electrode; a gas-tight connection is made between its upper portion and the sheath and the lowerportion is spaced a substantial distance away from the electrode, This forms a pocket between such member and the electrode in which gas can be collected. I have also discovered that a substantially gas-tight connection can be made between this ring member and the bath without short-circuiting the pot and without causing undue contamination by bringing the lower edge of this member down close to the crust of the bath but slightly above it. The crust on the bath as well as the bath itself is an electrical conductor and it is important that this member should not make contact with the crust. V

The final seal between the crust and the lower edge of the apron of this ring member is made by packing powdered alumina against the edge of such apron and piling it on top of the crust of the bath. Powdered alumina is a very poor conductor of heat and is substantially a nonconductor of electricity.

The connection between the ring member and the sheath may be made in various manners. For example, if the sheath is a fixed one, the ring member may be attached to it directly as by welding. Ordinarily, however, it is better to make a joint between the ring member and the sheath which permits relative movement between them. This permits the use, of a movable sheath and even when a fiXed sheath is employed, it permits the ring member to be lifted so that access can be had to the lowerpart of the electrode.

I have found that the best way to make a connection between the ring member and the sheath which permits relative movement is to bring the ring member either against or adjacent the sheath and arrange a pocket so that alumina can be packed to .fill the joint between thetwo. If the space between the sheath and the ring memberis relatively large, small chunks of alumina can first be inserted andthen fine material can be put on top; or if the space is small, the whole packing may be formed of finely-powdered alumnia. Of course, other sand-like material can be employed but as there, isa possibility of this material getting into the bath, it is preferable to use alumina, which can do no harm.

While this invention is particularly useful in connection with pots for the production of aluminum, it may also be employed with equipment for any other similar process where like conditions are found.

This invention can readily be understood by reference to the accompanying drawings in which Fig. 1 is a transverse section through such a furnace embodying my invention; Fig. 2 is a detailed view showing the connection between the ring member and the fixed sheath and also showing the manner in which the gases are withdrawn from the pocket between the ring member and the electrode; Fig. 3 is a section on line 33 of Fig. 2 and Figs. 4, 5 and 6 are detailed views indicating alternative forms of construction.

In the drawings, the pot I0 is of usual construction. The electrode [2 is of the continuous self-baking type, sometimes termed the Soederw berg type and goes down into the pot as indicated in Fig. 2. The electrode moves in a fixed casing !4 reinforced by channels l5. The electrode is suspended by contact rods 18 which go down into the electrode mass. The detailed manner in which these rods are handled is explained more fully in my co-pending application Serial No. 25,174, filed May 5, 1948, now United States Patent No. 2,475,452, issued July 5, 1949. The contact members [8 are attached to support bars 20 and these in turn are clamped to bus bars 22 which also serve to carry the weight of the electrode. The bus bars 22 can be raised and lowered by jacks 24. The broken line through the mass of the electrode l2 indicates approximately the zone where the electrode material has become baked and solidified.

Coming now to my present invention, which is shown in detail in Figs. 2 and 3, a member 3!] is bolted to the bottom channel [6 and entirely surrounds the electrode so that in effect it is part of the fixed casing or sheath. It is to be noted I that this member is shaped to form a continuous pocket or groove around the electrode. The ring member in this instance comprises an apron of corrugated or fluted metal 32 (see Fig. 3) to which is welded a top plate 34. The inner edge of the plate 34 has a downwardly depending flange 35 which can go down into the groove or pocket in the member 30. At spaced intervals flanges 38 are welded to the plate 34 for suspension of the ring member.

A shaft 40 runs down each side of the furnace and each such shaft is carried by brackets 42 attached to the upper channel l6. These shafts are rotatable and carry drums 44 for winding up suspension chains 46 which are attached to the flanges 38 of the ring members. If desired, the strips 32 can be interrupted on each side of the electrode to provide a sight box 48 which has a removable cover 48. Such boxes may also be used for introducing alumina within the line of the ring member.

In the present embodiment of my invention, an exhaust pipe 50 is welded into the ring member 32 at some convenient point. This pipe 50 is in turn welded into an inverted cup 52 which fits down into a circular recess formed in the upper surface of the gas-receiving chamber 54. The gas which passes down into the gas-receiving chamber 54 may then be burned as under the hood 56 to destroy tar vapors from the electrode, and the residual gas withdrawn through the pipe 85 to any convenient point where it may either be disposed of or treated to recover the fluorine products according to known processes. The groove in the member 32 and the recess in the member 54 are packed with finely-powdered alumina to prevent gas leakage.

During the operation of the furnace, the aluminum goes to the bottom of the pot as indicated at 51 in Figs. 1 and 2. Above this is the molten bath of cryolite containing dissolved alumina indicated at 58. The upper surface of this bath (and to some extent its edge portions) tends to solidify to form a crust indicated at 50. The upper layer of this crust is somewhat irregular and tends to merge into the loose alumina 6! above it but there is a distinct difference between the two as the fused crust is an electrical conductor, whereas the loose alumina is not. This crust may rest on the surface of the bath or may have sufiicient strength to stay above this surface when molten aluminum is drawn off, lowering the surface level of the bath. Apron 32 is so proportioned that its bottom edge comes down adjacent to the crust on the bath but does not enter it sufiiciently to make electrical contact. The loose alumina BI is then packed in between the edge of the apron 32 and the side of the pot 10 to form a substantially gas-tight seal.

In this case the seal is between the edge of the ring member and the side of the pot. In other examples it is between the edge of the ring member and the crust. This distinction is relah tively unimportant. The essential point is that there must be a substantial gas-impervious cover for the bath (which cover may include the crust) and the apron must not make electrical contact with the bath either through the crust or the pot as this may tend to short-circuit the furnace. Where the seal is between the ring member and the pot, the ring member must be raised (at rather infrequent intervals) to introduce alumina unless special means for supplying the alumina is provided as described above.

In Fig. 4 a simple form of ring member 62 is indicated suspended by a chain or cable 54. In this case instead of packing the joint between the sheath 5t and the ring member 62 with alumina, a V-shaped metal piece 68 is dropped into the pocket between the members 62 and 66. This view also illustrates the manner in which solidlfied materials (perhaps solidified drops from the bath carried up with the escaping gases) tend to build up on the inside of the ring member if its sides are vertical as shown in this figure. Because of that it is preferable to have the ring member flared away from the electrode as shown in Figs. 1 and 2.

Fig. 5 shows a structure having a movable sheath l0 and a ring member 12 suspended by chain 14. The permanent sheath is indicated at T6. In this case the joint between the member 72 and the movable sheath 10 is packed with powdered material as indicated at 18. As an appreciable clearance between members 12 and 1D is here indicated, this packing may be made up of pieces of appreciable size made by breaking up a part of the crust or using the alumina where it is partly sintered adjacent the crust. This may be supplemented with finely-powdered alumina.

Fig. 6 shows a ring member welded directly to the permanent sheath 82 so that no such suspension member is needed. The pipe 34 is used for withdrawing gases. Since this ring member cannot be raised, ordinarily the clearance between its lower edge and the crust should be greater than is otherwise necessary so that by inserting a bar at an angle, the crust underneath this ring member may be broken and alumina in-- troduced. It is understood that with any of the forms of device shown it is customary (as in the usual operation of an aluminum pot) to break the crust periodically. This causes the loose alumina on the crust (with additional alumina if desired) to enter the bath to be dissolved in the molten cryolite. In carrying out my invention conditions are then once more maintained to cause the crust to reform and again loose alumina is packed against thebottom'of the gas-collecting ring to; form the desiredsubstantially gas-tight joint between the ringandthe'crust.

I have previously stated. that the apparatus of this invention permits me to withdraw the gases in a muchmore concentrated. state than was'previously possible. For. example, I am; familiar with the operations carried out under United States Patent No. 2,031,554 having to do with the recovery of gases from an aluminum reduction plant. The operation of that process was dependent upon using an enclosed furnace, but nevertheless the average gases removed contained less than 1% of gases which came from the electrochemical process and there was a large dilution with air. According to the present invention during normal operation there is virtually no air dilution.

One direct result of this ability to withdraw the gases in undiluted form is that they are sufliciently concentrated so that the combustile elements in them may be burned. This is important since the hydrocarbon tar vapors coming out from the lower part of the electrode mass (resulting from the decomposition of the binder used in the electrode) were extremely difficult to remove from the gases previously obtained and would go through the scrubbing apparatus and have an injurious action on surrounding vegetation.

Experience has shown that the gases actually developed by the chemical reaction contain a major percentage of carbon dioxide. For example, the ratio of CO to CO2 may equal approximately 40/60. In this ratio there might be a tendency for the gases moving up around the electrode to corrode the sides of the electrode under the thermal conditions obtaining, following the reaction CO2+C=2CO. It may be noted that the temperature of the bath is about 950 C. and while the tar vapors will be much cooler, the temperature in the pocket between the ring member and the electrode will range between 950 and about 500 C. It has been found that when one entrains the tar vapors directly with the gases evolved from the reaction, some of the carbon from the tar appears to react with the gases generated below the surface of the bath, so that the resulting proportion of CO to CO2 may be increased to about 60/40. This tends to protect the side of the electrode against corrosion and of course at the same time the side of the electrode is protected against direct exposure and combustion in the air. Also this makes a gas mixture which can readily be burned for the destruction of tars, as previously stated.

Summing the matter up, the use of this apparatus permits me to carry out a process (when the furnace is in full operation and air has been displaced from the gas passages) in which the gases resulting from the primary process of reducing the aluminum are blended without admixture of substantial quantities of oxygen with the hydrocarbons resulting from the decomposition of tars in the electrode, and this blanding takes place while the gases are still hot enough to be conducive to the formation of CO in the presence of excess hydrocarbons from the tars. Th resulting gases are then burned with admixture of air to decompose any tar vapors remaining and finally the gases are subjected to a known process for scrubbing out fluorine compounds.

In the foregoing specification, I have indicated the best method known to me of carrying out my invention and have indicatedsome modifications that may be employed. However, many other modifications may be made without departing.

from the spiritoimy invention.

This application is a continuation in part of my earlier application, Serial No. 587,514, filed April 10, 1945, for. Open. Electrolytic Cell with Device for CollectingGasesj. now abandoned.

What I claim is:

1. The method of operating an aluminum furnace of a type comprising a pot for the furnace charge, a casing for guiding an electrode down towards the pot, a carbonaceous electrode in said casing, a gas-collecting ring member spaced out from the casing at the lower edge of the ring member and means for making a substantially gas-tight joint between the casing and an upper part of the ring, which method comprises the steps of maintaining within the pot a charge of alumina dissolved in molten cryolite under conditions to cause a crust to form on such'charge reaching up to a level approaching but not contacting said gas-collecting ring so that no substantial electrical discharge is made through such ring, piling substantially electrically non-conductive alumina on such crust in quantity sumcient to embed the lower edge of such ring thus forming a gas-tight joint between the charge and the gas-collecting ring, collecting gases within such ring, and withdrawing such gases from within the ring.

2. A method as specified in claim 1 which includes the further step of burning the gases collected within the ring, with air.

3. A method as specified in claim 1 in which said electrode is of the continuous self-baking type and in which the gases collected within the ring comprise tar vapors from the electrode and CO and CO2 evolved Within the bath and in which conditions are maintained so that the temperature of a mixture of such gases is within the range of between 500 C. and 900 C. so that a reaction takes place between tar vapors and CO2 to increase the proportion of CO in the mixture.

4. The method of operating an aluminum furnace of a type comprising a pot for the furnace charge, a casing for forming and guiding an electrode down towards the pot, a continuous selfbaking type of carbonaceous electrode in said casing extending down into the pot below the lower edge of the casing, a gas-collecting ring member having its lower edge spaced out from the electrode to form a gas-collecting area around the electrode and having an exhaust pipe for withdrawing gases from within such area, and means for making a substantially gas-tight joint between such ring member and the casing, which method comprises the steps of maintaining within the pot a charge of alumina dissolved in molten cryolite, maintaining such charge under conditions to cause a crust to form thereon which reaches up to a level approaching but not contacting said gas-collecting ring, piling alumina on said crust in quantity sufiicient to embed the lower edge of such ring so that a gas-collecting chamber is formed around the lower portion of the electrode above the charge which chamber is substantially gas-tight except for said exhaust pipe, collecting within such chamber gases from the bath and vapors from the electrode substantially without admixture with air, withdrawing such gases and vapors from the chamber and burning the same.

ROBERT JOUANNET.

(References on following page) REFERENCES CITED FOREIGN PATENTS The following references are of record in the Number Country Date file of this patent: 361,683 France Aug; 9, 1906 UNITED STATES PATENTS OTHER REFERENCES Number Name Date Journal of the Electrochemical Society, vol. 94,

2,031,554 Torchet Feb. 18, 1 3 No. 5, November 1948, pages 220 to 231,

2,100,927 Trematore Nov. 30, 1937