US3108057A - Electrodes containing petroleum coke - Google Patents

Description

1963 J. F. NELSON 3,108,057

ELECTRODES CONTAINING PETROLEUM COKE Filed June 10, 1960 0 TOTAL FLUID COKE AS BASE 0 +48 MESH FLUID COKE AS BASE Q 4 I I I I I C (D 2 LL] D X .I D m I I I I I O IO 20 3O 4O 5O 6O 7O W177, FINES IN BASE STOCK SHOWN FIG-I FINES I V FINES 2 :1 FINES 3 U U a t (D Z l-LI Q 5 D 03 WT. FINES FIG-2 Joseph F. Nelson lnve nror By Patent Attorney United States Patent 3,108,057 ELECTRODES CONTAINING PETROLEUM COKE Joseph F. Nelson, Westlield, N.J., assignor to Esso Research and Engineering Company, a corporation of Delaware Filed June 10, 1960, Ser. No. 35,262 8 Claims. (Cl. 20424) This invention relates to carbon electrodes and methods of making carbon electrodes. More particularly the invention relates to carbon electrodes made from a mixture of petroleum coke made by the delayed coking process and petroleum coke made by the fluid coking process.

The carbon electrodes of the present invention are especially useful in the manufacture of aluminum by the electrolysis reduction of alumina in a fused bath. The carbon electrodes have heretofore been manufactured from petroleum coke obtained from the delayed coking process. The coke must be of relatively high purity to avoid introduction of metal impurities into the aluminum product. Mixtures of petroleum coke from. the delayed coking process and from the fluid coking process are known in the manufacture of electrodes but such electrodes have not proved entirely satisfactory. Electrodes made from delayed coke are too reactive and a large part of the electrode is lost by reaction with oxygen in the air and by reaction with carbon dioxide formed during the electrolysis of the alumina, which is carried out at about 950 C.

One of the problems in using electrodes made heretofore is the dusting of the electrode due to different rates of consumption of the coke particles and the coke binder resulting from the carbonaceous deposit from the pitch binder during baking of the electrode. The coke components that are consumed the most slowly fall off the electrode giving rise to what is called the dusting problem. This carbon dust short circuits the electrolysis bath and represents lost coke, i.e. coke that does not reduce alumina. Another problem is cracking and shrinkage of the electrode in use.

Electrodes are usually made by selecting coke particles which have been calcined and partially ground and mixin the coke particles of selected particle size distribution with a suitable carbonaceous binder. The resulting mixture is molded into the desired shape and baked to carbonize or coke the binder to form what is known as a prebaked electrode. The pitch-coke mixture can alternat-ely be used by feeding it continuously to a Soderberg electrode which continuously bakes the mixture in situ by utilizing the heat from the electrolysis bath.

The petroleum coke from the fluidized coking process which is now a well known process and is disclosed in many patents, as for example, Pfeiffer et 211., 2,881,130, issued April 7, 1959. The coke particles from the fluid coking process normally or usually have a size in the range between about 10 and 200 standard mesh or about 74 microns to 2000 microns with most or about 80 wt. percent of the coke particles having a diameter in the range between about 20 and 140 standard mesh or about 105 microns and 840 microns.

Some typical fluid coke distributions are as follows:

I. Fine fluid cokes (wt. percent retained on):

ice

2 II. Coarse fiuid coke (wt. percent retained on):

10 mesh 2.4 20 mesh 5.4 30 mesh 26 40 mesh 28 60 mesh 20 mesh 15 Through 80 mesh 3.2

Other size distributions can be obtained depending on how the fluid coker is operated, but in general fluid coke much larger than that in Example 11 above cannot be obtained due to bogging of the fluid bed when attempts were made to increase particle size. Fluid coke particles in general are too small to be used alone and coarser delayed coke particles are used to supply this coarse material.

The petroleum coke particles made in the fluid coking process are laminar in structure and each particle may comprise or be made up of between about 10 and 100 superimposed layers of coke. The real density of these coke particles from the fluid coking process after calcining is in the range between about 1.83 and 2.00.

The calcined delayed coke particles have a higher real density than those from the fluid coking process and this density is about 2.0 to 2.10. After moderate grinding, the delayed coke particles as an example have a size between about 30 mesh (590 microns) and /2 inch in diameter with most of the particles or about 80-90 wt. percent being between about /2 inch and 20 standard mesh (840 microns) with the balance being finer than 20 mesh. By varying the degree and method of grinding other particle size distributions can readily be obtained. By severe grinding particles below 100 mesh or 200 mesh or even 325 mesh in size can be obtained. For the purposes of this invention, it is not necessary to grind the delayed coke to these very fine particle sizes.

The calcining of the coke particles made from the delayed coking process and from the fluid coking process is carried out in the conventional manner at a temperature between about 1800 F. and 3000 F. and for about 0.5 hour to 1-3 weeks, the longer times being used with the lower temperatures. The preferred calcining temperature is about 2300 F. to 2800 F. The calcination treatment increases the density of the coke particles and removes volatile materials.

In general two types of electrodes are .used by the industry, namely, (1) prebaked electrodes, and (2) Soderberg self-baking electrodes. The present invention may be used in both forms of electrodes.

In the drawing;

FIGS. 1 and 2 represent graphs showing how maximum packing or bulk density varies with different mix tures of calcined fluid coke fractions.

According to the present invention a carbon electrode is made from a mixture of petroleum coke from the delayed coking process, unground petroleum coke from the fluid coking process and ground petroleum coke from the fluid coking process. The coke mixture used is made up of about 60 to by weight of the petroleum coke from the fluid coking process, a part of which is ground, and about 40 to 10% by weight of delayed coke of a size between about 20 mesh and /2 inch diameter particles. The coke from the fluid coking process that is ground is made up of particles of a size smaller than about 200 mesh (74 microns) preferably containing some material smaller than 300 mesh (40 microns).

The ratio of ground to unground petroleum coke from the fluid coking process is between about 0.1 :and 1.0 or preferably between about 0.15 and 0.75 to give a high packing density and low electrical resistivity in the electrode. Or, in other words, of the entire fluid coke mixp ture about 90% to 50% by weight is the coarse or unground coke from the fluid coking process and 10% to 50% by weight is the finely ground coke or coke flour from the fluid coking process.

The petroleum cokes are first calcined and after grinding where necessary to obtain the desired particle size and sieving in some cases are then mixed at an elevated temperature ie at a temperature above the softening point of the pitch in their proper ratio with about 12 to 40 wt. percent of pitch binder, that is about 12 to 40 parts by weight per 100 parts by weight of the coke-binder mixture. The higher amounts of pitch are used when Soderberg electrodes are made. The quantities of pitch used are about 13 to 20% for prebaked electrodes and 25 to 35 for Soderberg. The mixture is then molded in the case of prebaked electrodes under pressure between about 3000 and 10,000 psi. or extruded, and then baked for periods up to 30 days at about 1800 F. to 2400 F. The prebaked electrodes thus made are then ready for use. In the case of Soderberg electrodes, the pitch coke mixture is fed directly to the Soderberg electrode.

The electrodes of this invention have electrical resistivities in the range of 1.9 to 3.0 X 10- ohm-in. and crushing strengths of about 4500 to 9000 psi. They are resistant to cracking and dusting. The use of fluid coke fines instead of delayed coke fines leads to a much cleaner operation in preparing the coke for making electrodes. Delayed coke fines are very dusty, whereas fluid coke fines are much cleaner to handle.

The coke from the delayed coking process reduces cracking and shrinkage of the carbon electrode. The coke from the fluid coking process fills the voids between the coke from the delayed coking process and gives high electrical conductivity. The density of the fluid coke is closer to the density of the coke resulting from decomposition of the pitch binder, than is the density of the delayed coke. As a result, the fluid coke and the coke residue from the pitch have more nearly the same reactivity, which results in less dusting of the electrode during the electrolysis of the alumina.

The pitch binder is usually an aromatic coal tar pitch such as coal tar pitches having a melting point within the range between about 70 C. and 120 C. and has about 5% or less of hydrogen. The concentration of benzene insoluble material in the binder is preferably between about 20 and 35% by weight of the binder and the nitrobenzene insoluble material in the binder is between about and 20% by weight of the binder. A good pitch binder will leave greater than 50% of carbon when it is coked. Other pitch binders such as those from petroleum having the desirable and necessary characteristics may be used instead of the conventional coal tar pitch binder.

The use of 10 to 40% of delayed coke and 90 to 60% of fluid coke provides an optimum mix which takes advantage of the delayed cokes property to prevent cracking and undue shrinking of the electrode and fluid cokes property to reduce dusting of the electrode. Dusting is a serious problem with 100% delayed coke electrodes. It is permissible to allow a small amount of delayed coke fines smaller than 20 mesh or even as small as 200 or 325 mesh to be incorporated in the carbon mix for making the electrodes, but these should be kept at a minimum in order to obtain the best electrode.

The optimum ratio of ground to unground fluid coke can be determined by making various mixtures of the two to determine what mixture gives the maximum packing density. The greater the packing density, the lower the resistivity of the electrode, other factors being kept constant. This ratio of the ground to unground coke will depend on the particle size and distribution of the unground coke, which varies with different modes of the coking operation and with different cokers, and the degree of fineness to which the ground portion is pulverized. It has to be determined for each type of operation that may come up. Examples of maximum packing density are given in FIGS. 1 and 2.

FIG. 1 represents graphically the results of work carried out to determine maximum bulk density of mixtures of fluid coke and to show the effect of particle size and size range on the relative settled bulk densities of fluid cokes.

In FIG. 1 the curve with the open circles represents the mixture with total fluid coke as a base or base stock and the curve with the solid circles represents the mixture with +48 mesh fluid coke.

FIG. 1 shows that adding fine coke particles (92% through 325 mesh of calcined fluid coke) to total original fluid calcined coke as a base or to a coarse fraction (larger than 48 mesh) of the total original fluid calcined coke as a base resulted in essentially the same maximum density at less than about 30 wt. percent of the fine coke particles. The maximum bulk density for the two mixtures shown in FIG. 1 is between about 20 and 30 wt. percent fine coke particles. The abscissa in FIG. 1 represents weight percent fine coke particles in fines plus the base or base stock.

FIG. 2 represents graphically the results of work carried out to determine maximum settled or bulk density of calcined fluid coke mixtures obtained by addition of calcined fine coke particle fractions to a base or base stock. Variations in maximum bulk density are a function of the particle sizes in the fine coke particle fractions.

In FIG. 2 the curves show the results of adding three types of calcined fines to the base of 48 mesh calcined fluid coke. The following Table I shows the size distribution of three fines fractions of coke.

TABLE I Cumulative, Wt. Percent Fines 1 Fines 2 Fines 3 In FIG. 2 the top curve with the solid circles represents the mixture containing fines 1, the next lower curve with the inverted open triagles represents the mixture containing fines 2 and the lowest curve with the open squares represents the mixture containing fines 3.

The abscissa in FIG. 2 represents the weight percent fine coke particles in fines plus the base or base stock.

From an inspection of FIG. 2, it will be seen that maximum bulk densities are obtained at about 25 wt. percent fines with a density improvement approaching 20%. It will also be seen that the greatest increase in maximum bulk density was obtained in these experiments with the finest coke fraction or fines 1.

In mixtures of unground and ground fluid coke with the pitch binder, an adjustment has to be made in some cases for the fluidity of the pitch, since the pitch, which is still quite viscous at the mixing temperature, interferes with the movement of the coke particles into the position of the maximum packing density. This interference can be compensated for by increasing the ground portion of the fluid coke by 5 to about 50% of the amout of ground fluid coke present, at the expense of the unground coke. The size of this adjustment can only be determined by making electrodes with various mixtures of the partcular coke at hand and measuring the electrical resistivity of the electrode. The lower the resistivity, the better the mix, but this mix will fall within the amounts specified above. The exact amounts will have to be determined since there is a high interdependency of coke particles size and distribution for both the ground and unground fluid coke and pitch softening point, the amount of the carbon residue from the pitch when the electrode is baked as Well as the many other properties of the coke and the pitch. In the case of pitches which are relatively fluid at the mixing temperatures, the adjustment for fluidity is small and may even be neglected. The higher the mixing temperature, the less the adjustment for pitch fluidity. This invention teaches the limits within which the best electrode properties are obtainable and the mechanism for determining the best electrode that can be made from the raw materials of this invention.

Table II illustrates some coke mixes falling within this invention. Many other mixes can be used that are within the scope of this invention.

TABLE 11 Composition of Calcined Coke and Coke Particle Size Distribution in Formulating Electrodes Percent 4 40 50 40 A. Delayed Coke (mesh)... 1: 2 32 :8 $8 20 6 0 20 5 6 s 60 20 74 80 36 30 15 B. Fluid Coke (Mesh) +100 15 3 +140 15 15 Nil +200 8 10 o -200 1 4 0 -200 0 0 0 C. Fluid Coke Ground {+300 75 50 60 (Mesh). -300 50 40 Electrode:

Percent 41.. 20 10 40 30 15 20 Percent B. 60 65 50 3O 4O 60 55 Percent O 20 25 20 15 30 30 15 25 4 s 4 12 16 20 15 7.5 s s 6 3 9 12 16 12 6.0 9 +14 4.8 2.4 7.2 9.6 4 3 1.5 4 +20 4.2 4 4.3 4.7 1.8 2.4 3.6 4.4 60 12 1a 10 9 4.5 6 0.0 40.7 Mesh 80 21.6 23.4 18 16.2 9 12 18.0 9.3 +100 9.0 9.9 7.5 6.7 6 s 12.0 1.6

-In a further refinement of the invention the coarse deeluding ground and unground fluid coke, the ground layed coke can be incorporated with the ground and unfluid coke comprising coke flour having a size finer ground fluid coke to determine the maximum packing than 200 mesh in an amount between about 6 and 45% density. The maximum packing density is determined by weight of the total cok mixture and the unground by well mixing the coke particles and then moderately fluid coke particles being of a size larger than 200 mesh vibrating the mixture until no further setthng takes place. 4.5 with about 14 to 30 parts by weight of a carbonaceous The volume is then read. The smallest volume for various binder, molding the resulting mixture and baking it at a mixtures of coke of total constant weight represents the temperature between about 1800 F. and 2400 F. for maximum packing density. I about 8 hours to 300 hours.

2. A method of making an electrode which comprises EXAMPLE 1 providing a mixture of coke particles admixed with 12 A prebaked electrode is prepared by mixing about 70% P 40 P F of Pitch Said coke f t by weight of coke particles from the fluid coking process lncl'lldlng tlfllclned Coke havlng a F denslty m with 30% by weight of coke particles from the delayed the range of abou} 9 and calcllled y process, The delayed ok ti l h v a ize between coke :parhcles having a mmimum real density of about about 20 mesh and /2 inch in diameter. The flllld" 2.10, the delayed coke particles comprising between coke particles have about 49% by weight of coarse p about 1 0 and 40 wt. percent of the total coke mixture, ticles having a size between about 10 me h nd 200 the fluid coke particles comprising about to 90 wt. mesh and the rest 01- about 2.1% by weight of coke fines percent of the total coke mixture and including fluid or flour having a particle size finer than 200 mesh based coke fines to provide fines having a size less than about on the total coke used to prepare the electrode. These 60 200 mesh in an amount between about 10 and 50% by col-:es were calcined at about 2400 F. for about one W ight of the flu d C k mix re a d Sa d delayed coke hour. About 82 parts by weight of this calcined coke being of a size between about 20 mesh and /2 inch in mixture was admixed with about 1 8 parts by weight of coal diameter and adjusting the fluidity of said pitch binder tar pitch binder having a melting point of about 210 F. at the mixing temperature to assure the movement of the The resulting mixture was subjected to a pressure of coke particles into the position of the maximum packing about 5000 p.s.i. in amold. The molded mixture was then density by changing the amount of the ground portion baked at a temperature of about 2000 F. to coke the of the fluid coke by about 5 to 50% of the amount of binder. The baked electrode had a compression strength ground fluid coke present in the mixture at the expense of about 7000 p.s.i. and an electrical resistivity of about of th unground fluid coke, molding the mixture and 2.5 10- ohm-inch. baking the molded mixture at a temperature above about A good electrode for use in the manufacture of alu- 1800 F. minurn must have a minimum compression strength of 3. A method of making an electrode which comprises about 4200 p.s.i. (pounds per square inch), a minimum providing a mixture of coke particles admixed with 12 to bulk density of about 1.45 and a maximum resistivity of 40 Weight percent of a pitch binder, said coke mixture about 4.0 l0- ohm-inch. The electrodes made by the including calcined fluid coke having a real density in the range of about 1.83 to 2.0 and calcined delayed coke particles having a minimum real density of about 2.0, the delayed coke particles comprising between about and 40 Wt. percent of the total coke mixture, the fluid coke particles comprising about 60 to 90 wt. percent of the total coke mixture and including fluid coke fines to provide fines having a size less than about 200 mesh in an amount between about 10 and 50% by weight of the fluid coke mixture and said delayed coke being of a size between about mesh and /2 inch in diameter and adjusting the fluidity of said pitch binder at the mixing temperature to assure the movement of the coke particles into the position of the maximum packing density by changing the amount of the ground portion of the fluid coke up to about 50% by weight of the amount of ground fluid coke present at the expense of the unground fluid coke, molding the mixture and baking the molded mixture at a temperature above about 1800 F.

4. A method of making an electrode which comprises mixing about 10%-40% by weight of calcined delayed coke of a size between about 20 mesh and /2 inch with about 60%-90% by Weight of calcined fluid coke ineluding ground and unground fluid coke, the ground fluid coke comprising coke flour having a size finer than 200 mesh in an amount between about 6 and 45% by weight of the total coke mixture and the unground fluid coke particles being of a size larger than about 200 mesh with a pitch binder, molding the resulting mixture and baking at a temperature of at least 1800 F. for at least about 8 hours.

5. A method according to claim 4 wherein the density of the calcined fluid coke is closer to the density of the coke resulting from decomposition of said pitch binder than is the density of the calcined delayed coke and dusting of the electrode is reduced.

6. A method of making an electrode which comprises mixing about 60%-90% by weight of calcined fluid coke having a real density in the range of about 1.83 to 2.0 and including fluid coke fines resulting from grind- 0 ing fluid coke to provide fines having a size less than about 200 mesh and in an amount between about 10% and 'by weight of th fluid coke mixture, about 10%-40% by weight of calcined delayed coke of a size between about 20 mesh and /2 inch and having a minimum real density of about 2.0, with about 12 to 40 parts by weight per 100 parts by weight of the cokebinder mixture by weight of a pitch binder, molding the resulting mixture and baking it at a temperature above about 1800 F. for about 8 hours to 300 hours.

7. A method of making an electrode which comprises mixing calcined fluid coke having a real density in the range of about 1.83 to 2.0 and calcined delayed coke particles having a minimum real density of about 2.0 and being of a size coarser than about 20 mesh, the delayed coke particles comprising between about 10% and 40% by weight of the total coke mixture, the fluid coke particles comprising about to by weight of the total coke mixture and including fluid coke fines having a size less than about 200 mesh in an amount between about 10% and 50% by weight of the fluid coke mixture, with about 12 to 40% by weight of the cokepitch binder mixture of a pitch binder, molding the resulting mixture and baking it at a temperature above about 1800 F. for about 8 hours to 300 hours.

8. A method of making an electrode which comprises mixing about 10%-40% by weight of calcined delayed coke of a size between about 20 mesh and /2 inch, about 60%-90% by Weight of calcined fluid coke including unground fluid coke and fluid coke fines comprising coke flour having a size finer than 200 mesh in an amount between about 6 and 45% by Weight of the total coke mixture and the unground fluid coke particles being of a size larger than about 200 mesh, with a pitch binder, molding the resulting mixture and baking at a temperature of at least 1800 F. for at least about 8 hours.

References Cited in the file of this patent FOREIGN PATENTS 574,468 Canada Apr. 21, 1959

Claims (1)

1. A METHOD OF MAKING AN ELECTRODE WHICH COMPRISES MIXING ABOUT 10%-40% BY WEIGHT OF CALCINED "DELAYED" COKE OF A SIZE BETWEEN ABOUT 20 MESH AND 1/2 INCH WITH ABOUT 60%-90% BY WEIGHT CALCINED "FLUID" COKE INCLUDING GROUND AND UNGROUND "FLUID" COKE, THE GROUND "FLUID" COKE COMPRISING COKE FLOUR HAVING A SIZE FINER THAN 200 MESH IN AN AMOUNT BETWEEN ABOUIT 6 AND 45% BY WEIGHT OF THE TOTAL COKE MIXTURE AND THE UNGROUND "FLUID" COKE PARTICLES BEING OF A SIZE LARGER THAN 200 MESH WITH ABOUT 14 TO 30 PARTS BY WEIGHT OF A CARBONACEOUS BINDER, MOULDING THE RESULTING MIXTURE AND BAKING IT AT A TEMPERATURE BETWEEN ABOUT 1800*F. AND 2400*F. FOR ABOUT 8 HOURS TO 300 HOURS.

Images