US2922755A - Manufacture of graphitizable petroleum coke - Google Patents

Description

Jan. 26, 1960 R. c. HACKLEY, JR 2,922,755

MANUFACTURE OF GRAPHITIZABLE PETROLEUM coma Filed Oct. 14, 1957 s Sheets-Sheet 1 PHOTOMICROGRAPH OF FRACTURED SURFACE OF COMMERCIAL ELECTRODE GRADE PETROLEUM COKE.

F5. 2. PHOTOMICROGRAPH OF INVENTOR. FRACTURED SURFACE OF REGULAR PETROLEUM COKE. ROY C. HACKLEY, Jr.

ATTORNEY Jan. 26, 1960 R. c. HACKLEY, JR 2,922,755

MANUFACTURE OF GRAPHITIZABLE PETROLEUM COKE Filed oct. 14, 1957 S SheetS-Shee'c 2 u. FRACTURNED PEEPENDICULATTTb I LANAR SHEETS.

b. FRACTURED APPROXIMATELY PARALLEL T0 PLANAR SHEETS.

""f INVENTOR. FIG. 3. PHOTOMICROGRAPHS 0F ROY A FRACTURED SURFACES OF BY GRAPHITIZABLE PETROLEUM COKE WJ/m/ FROM BLENDED FEEDSTOCK.

ATTORNEY Jan. 2,. 1960 R. c. HACKLEY, JR 2,922,755

MANUFACTURE OF GRAPHITIZABLE PETROLEUM COKE Filed Oct. 14, 1957 s Sheets-Sheer. s

PHITIZABLE PETROLEUM COKE FROM BLENDED FEEDSTOCK.

b. GRA

'5- 4. ELECTRONMICROGRAPHS OF INVENTOR- FLAKES OF POWDERED PREMIUM ROY HACKLEY' COKES. (WIDTH OF SHADOW IS BY TWICE THE THICKNESS OF WM PARTICLE.)

ATTORNEY United States Patent MANUFACTURE OF GRAPHITIZABLE PETROLEUM COKE Roy C. Hackley, Jr., Ponca City, Okla.

Application October 14, 1957, Serial No. 690,064 5 Claims. (Cl. 208-39) This invention relates to the manufacture of premium grade petroleum coke, primarily for use in producing metallurgical electrodes. More particularly it relates to an improved process for the manufacture of such coke in higher yields than heretofore achieved and under conditions permitting utilization in part of feed stocks lower in cost and in greater supply than materials previously found acceptable for this purpose. In addition to providing a product of at least equal quality in greater yields at lower cost in terms of starting materials, this invention provides peculiar flexibility in refinery operations in making available for premium coke production a wide range of crude and partially refined or treated petroleum fractions heretofore considered antithetical to such operation.

Premium quality petroleum coke has been manufactured for over twenty years and has been called variously No. l coke, electrode grade coke, Kendall coke, needle or needle-like coke, premium coke, and by other designations which indicate the source of the raw material, physical appearance or the intended use of the product. Primarily premium coke is ground, calcined and then converted into multicrystalline synthetic graphite, and formed into electrodes for use in metallurgical electric furnaces. I prefer either the terms graphitizable or electrode grade to describe such a premium quality petroleum coke. This type of coke can be distwo solvents, one of which is tinguished superficially from ordinary or regular pechemical composition. Such petroleum crudes are rare in natural occurrence and hence cannot provide more than a small fraction of the raw material needed for the manufacture of widely demanded premium coke. The principal raw materials now used for this purpose are carefully selected highly aromatic residues obtained from catalytic and thermal cracking prepared by various techniques to eliminate the so-called normal contaminant coke formers as discussed in the US. Patent No. 2,775,549 to Shea.

It is reported by Shea (in US. Patent No. 2,775,549), and is generally accepted as true by those skilled in the art as reflected in the literature, that the incorporation or presence of reduced crude or any of various extraction and distillation residues in the thus selected or prepared coker charge for production of premium grade coke will prevent formation of the desired electrode grade coke. The result, instead, will be production of regular petroleum coke, even when very small percentages (less than 1% by weight) of the so-called contaminants are present. Surprisingly, and to the contrary, I have found 2,922,755 Patented Jan. 26, 19 60 (1) 100% catalytic or thermal cracker cycle oils or bottoms (2) A mixture of catalytic or thermal cracker cycle oils or bottoms and conventional coker gas oil and from about 10% to about 30% of Component B, which may be made up of one or more refinery residues such as the following:

(1) Virgin reduced crude (2) :Duo Sol extracts (3) Furfural extracts "(4) Hydroformer bottoms A Duo-Sol extract is the extract obtained in the Duo- Sol process of'refining lubricating oils. In this system of extraction two partially miscible solvents of different specific gravities flow counter current to each other through a series of extraction stages. Since the solvents are immiscible they produce two layers of sufficient specific gravity difference to allow gravity separation in each of the extraction stages, the term DuoSol means propane and the other is a blend of phenol and cresol.

The residues which may be used for Component B in my process are considered and have been taught to be unsuitable as charge stocks for premium coke manufacture, either individually, or when mixed with suitable stocks. In view of known facts and the teachings of the prior art, the operability of my process is unexpected and indeed contra-indicated and draws its economic merit and special advantage from producing much higher yields of premium coke from lower average cost feeds with consequent refinery flexibility.

The following discussion of premium coke manufacture should serve to make the functioning of my process more understandable.

The charge stock customarily used for coking is a hydrocarbon residue of high carbon to hydrogen ratio. When coke is desired for the manufacture of synthetic graphite, a hydrocarbon residue such as a thermal tar 'or pressure tar is selected or is specially prepared, in

which the high carbon to hydrogen ratio is primarily attributable to the presence of higher molecular weight fused ring aromatic hydrocarbons. 'In these compounds the carbon atoms are already arranged in a regular hexagonal pattern. On heating at around 840 F., the planar aromatic systems polymerize and become'rigid; that is, coke forms. In the delayed coking process, which is one of the most commonly used and most economical at the present time, the charge stock is preheated to about 910 F. and is then pumped at about 70 p.s.i. into a vertical coking drum through an inlet at the base. The pressure in the coking drum is maintained at about 65 psi. and the drum is well insulated to minimize heat loss, so that the reaction temperature lies between about 830 F. and 900 F. The hot charge stock decomposes over a period of several hours, liberating hydrocarbon vapors which rise through the mass continuously, stirring one direction. This type of stirring by unidirectional flow in one dimension that they are somewhat needlelike in appearance. X-ray examination of premium coke 'at this point shows a' general layering of the aromatic planes. The'planes are stacked approximately parallel in sheets, of sufficient thickness to be plainly visible in the surfaces of fracture coke particles under dark field illumination in a metallographic microscope. Both the fiat surfaces and the broken edges of sheets can be seen in the photomicrograph-Figure 1.

Although individual layers of carbon atoms in premium coke are spaced for the most part in the hexagonal pattern 5 1 typical of graphite, the layers still contain some hydrogen, are only approximately parallel to each other, are not equally spacedwith regard to each other, and are in random orientation around an axis normal toparallel layers.

When the coke is heated at about 1800! F. to 3600 F., 7

there is a pronounced evolution of gas accompanied by a growth of the planes." At this stage, although the particle size may be considerable, the properties of completed graphites are not observed. The generally accepted model .is one in which, although there are deep stacks of parallel planes, there is little ordering of these planes with respect to each other around an axis perpendicular to the planes. The product of this step is usually called calcined coke.

Upon heating to about 4500 F. to 55-00 F., the stacks of parallel planes become sufliciently oriented so as to form aggregates of many small graphite crystals. Although these small crystals show a tendency to have nearly parallel layer planes within each aggregate they are disoriented with respect to each other and there is evidence of the presence of many stacking faults within aggregates. The faults and imperfections are actually desirable because they are believed to contribute to the semiconductor behavior of multicrystalline synthetic graphite. This attribute is responsible for the low electrical resistance of synthetic graphite electrodes in the range of 750 to 2750" I. This property is essential for eflicient operation of electric furnaces. Natural graphite has what one would call typical metallic electrical behavior, that is, resistance increases continuously with temperature, and is therefore not nearly as suitable a material for furnace electrodes ,as synthetic graphite.

Natural graphite exists mainly in fairly large crystals .madeup of planar layers of carbon atoms, arranged for the most part in a pattern of regular hexagons, each atom being placed opposite the open centers of hexagons in. two adjacent layers. The bonding between carbon atoms in a layer is much stronger than the bonding between adjacent layers, which serves to explain the highly anisotropic nature of natural graphite. The tendency of graphite to cleave along the layer planes is. Well known. The coeflicient of thermal expansion in the direction perpendicular to the layers is very high compared with the coefficient in the direction of the layers. For example; samples of Ceylon graphite have been found to have a 'coefficient in the c-dimension of approximately 28 X1O- deg. C. and in the a-dimension of only 0.95x10- deg. C. Multicrystalline synthetic graphite displays this anisotropic characteristic to a more limited extent because of the somewhat disoriented position of the many small crystalline aggregates with respect to each other; however, since the layers in premium coke lie approximately parallel to the longitudinal axis of each small coke particle, there is a tendency for the graphite crystals to orient in similar fashion during graphitization. In the manufacture of electrodes the long-shaped coke particles are. partially oriented in the direction of flow through the extrusion file. It is not too surprising, therefore, that even in a finished electrode the coefficient of thermal expansion in the transverse direction may be as much as two to five times as great as the coeflicient in the longitudinal direction. Below are listed typical physical properties of extruded graphitized petroleum coke in the two directions.

Parallel Perpento grain dicular to grain Specific resistance, milliohm-crm 0.86 1. 62 Coat-I; of thermal expansion/deg. O.XlO- l. 1 4. 1 Elastic modulus 1O lb./in. 1. 84 0. 78 Flexural strength lb./in. 4, 520 3, 010

The diameter of the opening in the extrusion die, grain size of the crushed coke, graphitization temperature and other factors will produce different degrees of orientation of the graphite crystals; consequently the physical properties of synthetic graphite are quite variable. v

A petroleum residue such as virgin reduced crude has a high carbon to hydrogen ratio, desirable for coking, but in most instances the proportion of aromatic hydrocarbons present is not very great. There are present some rather complex resinous substances of high molecular weight which are believed to produce a cross-linked, rigid structure during the coking process which interferes with the formation of sheets of stacked carbon planes. This may or may not be an accurate explanation of the phenomenon; however ordinary coke is usually more dimensionally stable and resistant to fracture than premium coke and contains more bubbles and voids. Upon examination in the metallograph, the fracture surfaces of regular coke present a very disordered appearance. Some sheets of carbon can be observed, but they are folded, convoluted and often appear only in isolated zones. This is quite evident in the photomicrograph of Figure 2. After calcining and graphitization, this type of coke produces a product which contains isolated patches of graphite immersed in a mass of disoriented carbon. The electrical properties of ordinary petroleum coke are poor, and consequently such coke is of little utility in the electrode field and finds its use mainly as a fuel or for blending purposes with consequent low market value.

Previous workers in this field have "suggested that in the formation of graphitizable coke thermal treatment of the charge stock merely removes asphaltene and resinformers. As a result of my investigations, however, I believe the phenomenon is more complex. It is my belief that drastic thermal treatment of hydrocarbon feed stocks accomplishes the following:

(1) Removal of 'asphalte'nes and other resin-forrning components, and

.( 2) Formation of aromatic hydrocarbon structures which influence the coking characteristics of the bottoms fractions very strongly .toward the formation of planar sheets of stacked carbon layers. It is further postulated that this tendency is so strong in the case of some specific structures that it will predominate in spite of the presence of considerable quantities of otherwise undesirable impurities.

Theexact nature of the aromatic hydrocarbons which have the desirable layer-forming tendency is unknown, as the thermal tars have not been analyzed completely; however it is known that certain aromatic hydrocarbons such as anthracene may be converted to graphite with extraordinary case. There is a possibility that the presence of more reactive carbon atoms in high molecular weight fused ring aromatic hydrocarbons (similar to the 9 and 10 positions pf anthracene) provides the desirable tendency toward polymerization into layers of hexagonally p ce car o atoms..,

Ten laboratory experiments were performed in which a: highly aromatic thermal tar was mixed with various pro.- portions of vacuum reduced crude and coked. The resultlarge percentages of other refinery residues to the thermal ing coke Was calcined, crushed and converted into extruded graphite rods on which determinations were made of the coeflicient of thermal expansion, to serve as a measure of coke quality. The results are summarized in Table l.

TABLE 1 Composition of Coker Calculated Propor- Charge, Percent by tion of Total Yield,

Weight 'Coke Yield percent Percent of C.T.E./deg. 0. Example No. Total of Synthetic Highly Virgin Charge From From Graphite Rods Aromatic Reduced Thermal Reduced Thermal Crude Tar Crude Tar It will be noticed that, except at the lowest levels, up to the 32% level of reduced crude in the coker charge there is very little change in the coefiicient of thermal expansion. Since the coefficient of thermal expansion of the finished graphite depends to a great extent on the method of preparing the sample, it is compared in this case to the value obtained for graphite from 100% thermal tar, when prepared by the same technique.

It is noteworthy that the coeflicient of thermal expansion of the graphite resulting from coking a charge containing 5% Duo-Sol extract indicates a slight but delete- .rious decline in product quality, but at the 10% to range and especially at or about the 20% level the product is as good or better than that made with 100% thermal tar. Theseexperiments have been repeated and the data have been analyzed by statistical methods. While I do not wish to be bound by any theory as to the reasons for these variations, the most reasonable explanation is that at low levels of contamination, disoriented particles of carbon form which are so small as to be of colloidal dimensions and may interfere with the stacking and orientation of carbon layers. At slightly higher concentration levels the particles which form may be so large that they become isolated from the rest of the carbon structure and no longer interfere with stacking and orientation. At very high levels of contamination, of course, the whole situation reverses and zones of sheets of stacked carbon layers become isolated in the mass of disoriented carbon. The anomalous effect of the addition'of small percentages (1 to 10 percent) of low grade feedstock makes its appearance in a somewhat erratic and unpredictable manner. The manufacture of premium coke from a feedstock mixed in these proportions merely results in an uncontrolled, if small, variation in product quality without any substantial economic advantages. As no anomalous effect has been observed beyond the ten per cent level, I consider this a safe lower limit of contamination. With regard to the upper limit, satisfactory cokes have been made from blended feeds containing as much as fifty percent vacuum reduced crude. The results obtained at these high levels of contamination are inconsistent. Because of the great difference in market value of graphitizable coke and regular coke, it is inadvisable to take the risk of having batches of coke downgraded by addition of very bottoms varies from to percent and that of the residue varies from 25 to 15 percent.

In addition to the experiments described in Examples 1 to 10 wherein a blended feed stock consisting of a highly aromatic thermal tar admixed with varying amounts of virgin reduced crude were used, similar experiments were run utilizing a blended feed stock consisting of the highly aromatic thermal tar admixed with varying amounts of other refinery residues. These other refinery residues were Duo-Sol extract, furfural extract, and hydroformer bottoms. In all experiments the results were similar. Some of the advantages of a process using a blended feed stock have been pointed out or are obvious to those skilled in the art. One advantage of my invention which has not been pointed out and may not be obvious is that the value of the refinery residue is increased by a factor of five to twelve. V

The new and unexpected results which I obtain-with my process depend to a great extent upon psje of athermal tar of the proper type. In Examples lthrough 1.0 the thermal tar used was one which wasknown to be of a highly aromatic nature, suitable for t-heproduction of electrode grade coke of the highest ,quality'f 'Qbyiously, if the quality of the thermal tar is not controlled,-the proportion of low grade residue which can ,betole'rated in the mixed coker charge will vary, making control of the entire process very difficult. In order to illustrate fully the operation of my process I have, therefore, included in the following example the step of manufacturing the thermal tar which made up the major portion ofthe coker charge. I

EXAMPLE 11 A mixed feed of thermal catalytic cracking bottoms and coker gas oil from manufacture of regular coke was subjected to thermal cracking. A mixture of 5,248 barrels of thermal cracker bottoms obtained by this operation and 779 barrels of low ash vacuum reduced crude was coked by the Kellogg delayed coking process. There were produced 514.6 tons of graphitizable petroleum coke, a yield of 46.6 Wt. percent based on total charge. The yields of gas, gasoline, and-coker cycle oil were 9.0%, 10.2% and 33.0%, respectively. The properties of the charge stocks are suniniarizediin Table 2. V

TABLE 2 Properties of feedstacks Feed to Thermal GrackingUnit Feed to Delayed Coker' Resulting Blendz 87% Thermal r a 'Cracker- Catalyt c Regular Coker Vacuum Thermal Bottoms, 13%

Cracker Gas 011 Reduced Crude Cracker Vac. Red. Bottoms Bottoms Crude Gravity 4.4. Vapor pressure-- Flash 320 (C). Viscosity, S.S.U., at 100 F 89 1 Viscosity, turol, at 122 F 143.3. Pour point +40. Sulfur- 1.12. Aniline point 118 Cale Bromine number Carbon residue, percent, rams Ash er o rf atin Percent carbon.

Percent hydrogen (by difference) Distillation I.B.P

" The total ash of the coke was 0.032%. The spectrographic analysis of the ash is tabulated below, along with analyses of samples of regular coke and premium coke from 100% thermal tar for purposes of comparison.

TABLE 3 I Spectrographzc analyses Premium gg Regular Regular Coke g Coke Coke Blended Th e a! Feed Percent Ash 0. 096 0. 21 O. 032 0. 068 Metals Analysis, Percent of S I i ;Photomicrographs ot fractured surfaces of the coke appear in Figure 3. The existence of planar sheets of stacked carbon layers is quite apparent. When the coke was pulverized and the particles were examined under the electron microscope, they were found to consist prinicipally of plates or flakes of irregular outline. This be' Electrodes were ccmpounded, molded, graphitized, and

Cracked at 670.

Cracking at 716 used in full-scale tests in the refining of aluminum. They met every test .of acceptability to which they were subjected and in electrical properties were even superior 'to electrodes made from a premium coke derived from thermal tar. Electrodes prepared as described above from a blended feedstock are being subjected to further electrical tests.

While particular embodiments of the invention have been described, it will be understood, of course, that the invention is not limited thereto since many modifications may be made, particularly in selection of acceptable contaminant blend feeds for use with thermal tar and in determining various blending ranges dependent on the quality of each and that of the thermal tar; and it is, therefore, contemplated to cover by the appended claims any such modifications as fall within the true spirit'an scope of the invention. I

The invention having 'thus been described, what is claimed and desired to be secured by Letters Patent is: .1. A process for producing graphitizable coke which when calcined and graphitized has a longitudinal coefiicient of thermal expansion/ C. not substantially above about 0.55 10 which process comprises subjecting .to coking conditions of temperature and pressure a petroleum derived charge stock comprising a highly aromatic thermal tar and from about 10% to about 30% by weight of a virgin reduced crude.

2. A process for producing graphitizable coke which when calcined and graphitized has a longitudinal coeflicient of thermal expansion/ C. not substantially above "about 0.55 X10 which process comprises subjecting to coking conditions of temperature and pressure a petroleum derived charge stock comprising a highly aromatic thermal I tar and from about 15% to about 25% by weight of a 2,922,755 9 than about 70% to 90% by Weight of a highly aromatic thermal tar and from about 10% to about 30% by weight of a virgin reduced crude.

References Cited in the file of this patent UNITED STATES PATENTS 5. A process for producing graphitizable coke which 1,966,801 Lord July when calcined and graphitized has a longitudinal coef- 5 2135JO8 Ostergaard ficient of thermal expansion per C. not substantially 23,10,922 Barron 16, above about 0.55 10 which process comprises sub- 2,745,794 AlOZeYY et May jecting to coking conditions of temperature and pressure 2,775,549 Shea 25,

a petroleum derived charge stock consisting essentially of a highly aromatic thermal tar and from about 10% 10 to about 30% by weight of a virgin reduced crude.

FOREIGN PATENTS 449,870 Great Britain July 1,

Claims (1)

1. A PROCESS FOR PRODUCING GRAPHITIZABLE COKE WHICH WHEN CALCINED AND GRAPHITIZED HAS A LONGITUDINAL COEFFICIENT OF THERMAL EXPANSION/*C. NOT SUBSTANTIALLY ABOVE ABOUT 0.55X10-**6, WHICH PROCESS COMPRISES SUBJECTING TO COKING CONDITIONS OF TEMPERATURE AND PRESSURE A PETROLEUM DERIVED CHANGE STOCK COMPRISING A HIGHLY AROMATIC THERMAL TAR AND FROM ABOUT 10% TO ABOUT 30% BY WEIGHT OF A VIRGIN REDUCED CRUDE.

Images