US3531394A - Antifoulant additive for steam-cracking process - Google Patents

Description

United States Patent 'ice 3,531,394 ANTIFOULANT ADDITIVE FOR STEAM-CRACKING PROCESS Ihor Koszman, Parsippany, N.J., assignor to Esso Research and Engineering Company, a corporation of Delaware No Drawing. Filed Apr. 25, 1968, Ser. No. 724,264 Int. Cl. Cg 9/16, 9/36; C07c 11/04 US. Cl. 20848 27 Claims ABSTRACT OF THE DISCLOSURE In the thermal cracking of petroleum fractions in admixture with steam, severe process problems due to carbon formation, such as coke deposition, carbon monoxide formation, metal dusting, and erosion are eliminated or substantially reduced by providing for the presence of phosphorus and/ or bismuth-containing compounds in the cracking zone.

FIELD OF THE INVENTION This invention relates to a process for the thermal cracking of petroleum feed fractions in admixture with steam. More particularly, this invention relates to an im proved steam cracking process wherein the formation of carbon and its attendant difficulties are substantially reduced. Still more particularly, this invention relates to an improved steam cracking process wherein phosphorus and/or bismuthcontaining materials are present in the cracking Zone.

PRIOR ART The thermal cracking of petroleum feed streams, with or without the presence of steam, is well known to the art and is widely used as a source of desirable compounds, i.e., unsaturates such as ethylene and butadiene. Generally, when noncatalytic processes are conducted, it is preferable to use steam in admixture with the petroleum feed to regulate the reaction. While the process is technically and economically quite successful, several major problems exist which militate against the development of the full potential of the steam cracking process. Essentially, these problems center around the carbonforming tendency of the process gas, i.e., vaporized petroleum feed at reaction (cracking) temperatures. The various problems related to this carbon-forming tendency will now be briefly discussed.

Perhaps the most objectionable problem relating to carbon formation is the deposition of coke on the interior of the tube walls through which the cracking mixture flows. The deposition of coke is believed to be due to formation of free radicals, e.g., when ethane is cracked, methylene radicals are formed, which then polymerize into long chain compounds and dehydrogenate to form carbon or coke on the tube walls. The coke tends to build up and reduces the effective cross-sectional area of the cracking tube, thereby necessitating higher pressures to maintain a constant throughput. Additionally, however, the coke is an excellent thermal insulator and higher furnace temperatures are required. Such higher temperatures drastically shorten tube metal life. Consequently, frequent shutdown periods for decoking and/or tube replacement are necessary and overall product yield is lost.

A second problem associated with carbon formation is erosion of furnace tubes. Furnace tube erosion is generally believed to be caused by carbon particles entering into the gas phase at high velocities. The particles strike the tubes, particularly the return bends, and cause severe erosion. Even when a small amuont of carbon enters the gas phase, erosion can be quite severe, since the erosion rate increases with the third power of the gas velocity.

3,531,394 Patented Sept. 29, 1970 This phenomenon is known as carburization, and the car-' burized material loses its original oxidation resistance, thereby becoming susceptible to chemical attack. The mechanical properties of the tube also degrade and the material becomes brittle at ambient temperatures, exhibits high creep rates at operating temperatures, and becomes nonweldable if repairs are required.

A less frequent but still catastrophic effect of carbon formation results from the interaction of the furnace tube alloy with carbon-containing gas and is known as metal dusting. The effect is one of extremely rapid metal loss and, ultimately, tube failure.

Still another, but by no means the least, severe disadvantage of carbon formation is the tendency towards carbon monoxide formation due to the reaction of carbon with the water in the feed. Obviously, the formation of large amounts of carbon monoxide reduces the product selectivity respecting the desired steam cracking products. It is generally necessary, therefore, to endeavor to keep carbon monoxide make quite low, since separation costs are generally high and the purity of some products is disadvantageously affected, e.g., 0.5 to 5.0 ppm. by weight of carbon monoxide in ethylene will poi-son an ethylene polymerization catalyst. Therefore, it is generally believed that carbon monoxide make should be below about 1 wt. percent, preferably below about 0.5 wt. percent, and more preferably below about 0.1 Wt. percent. Thus, in order to reduce the expense in separating ethylene from carbon monoxide and to prepare chemically acceptable products, the carbon monoxide make must be substantially eliminated.

Now, it has been reported that the cracking of sulfurfree naphthas eliminates coke formation. Nevertheless, such a process leads to excessive carbon monoxide formation. It has also been reported that small amounts of sulfur in the feed may mitigate coke formation, erosion, and carbon monoxide formation; see White, US. Pat. 2,621,216. However, this process necessitates a rather complex back blending scheme in order to achieve the proper sulfur level. If the sulfur level rises, however, coking becomes excessive and the advantages of sulfur addition or removal (as from high sulfur feeds) are lost. By the practice of the present invention, it has now been discovered that all of the foregoing problems can be substantially reduced or eliminated by providing for the presence of phosphorus or bismuth-containing compounds in the cracking zone. It is even possible to eliminate excessive coking due to high sulfur levels, e.g., above 400 ppm, up to and including 1000 ppm. sulfur, by the presence of small amounts of these elements or their compounds in the cracking zone.

SUMMARY OF THE INVENTION In accordance with this invention, therefore, the tendency to form carbon, and the difiiculties attendant to carbon formation, in the thermal cracking of petroleum fractions in admixture with steam, is substantially reduced and the steam cracking operation is improved by providing for the presence of a material selected from the group consisting of materials containing phosphorus, bismuth, and mixtures thereof in the cracking zone. While it is not known with certainty just how these materials act to eliminate or reduce the carbon-forming tendency, it is believed that carbon formation is promoted by the furnace tube walls and these inhibiting materials kill or poison the catalytic action of the furnace tube materials,

3 i.e., by forming a protective coating on the furnace tube Walls.

The tendency to inhibit carbon formation shown by phosphorus and bismuth is particularly surprising when viewed in the context of neighboring metals in the Periodic Chart. Thus, both phosphorus and bismuth are members of Group V-A of the Period Chart of the Elements. Consequently, it would be normal to believe that the other metallic members of Group V-A would exhibit, to a greater or lesser extent, rather similar properties. However, arsenic and antimony, the other metals of Group V-A exhibit directly opposite roperties and tend to greatly increase carbon forming tendencies. Further, it has been found that by providing for the presence of phosphorous and/or bismuth in the cracking zone the carbon-forming tendencies of materials such as arsenic or sulfur are substantially reduced or eliminted and expensive feed pretreatments, e.g., purification, are no longer necessary.

It should also be noted that the present invention is unexpected even in view of the long standing practice of phosphating steels to reduce corrosion. Thus, phosphating of steels has been generally limited to steels that are not designated as stainless steels, i.e., austenitic steels, or steels that are not specially prepared for corrosion resistance. In addition, steam cracking temperatures are much higher than the temperatures generally employed where phosphates are found helpful and the phosphates are not generally recommended for use at such high temperatures.

Generally, the amount of phosphorus and/or bismuth that is present in the cracking zone is not critical and may vary over a relatively wide range. Nevertheless, it is preferred that these inhibiting materials be present in an amount of at least about 0.001% by weight of metal based on the feed, i.e., water plus petroleum fraction, to the cracking zone. More preferably, however, the material can be present in the range of 0.001 to 0.1 wt. percent, more preferably 0.01 to 0.1 wt. percent. Also, the form in which the material is added is not critical and it is only necessary that the phosphorus or bismuth be in a vaporous form, e.g., P at cracking temperatures. Suitable material which may be added to the feed to provide for the presence of the metal in the cracking zone are phosphorous and/or bismuth-containing com ounds (including the metals themselves which react with water to give vaporous P 0 or Bi O phosphoric acids, e.g., metaphosphoric acid, pyrophosphoric acid, orthophosphoric acid, phosphorous acids; hydrides, e.g., phosphine; bismuth hydroxide; organic compounds such as those having the formula R PO wherein R may be hydrogen or a C C alkyl, aryl, aralkyl, cycloalkyl, or alkaryl radicals, e.g., trimethyl phosphate, triethyl phosphate, triphenyl propyl phosphate, dimethyl propyl phosphate, cyclohexyl methyl phosphate, and the like; phosphine deriva tives, e.g., methyl phosphine, ethyl phosphine, phenyl phosphine, dimethyl phosphine, trimethyl phosphine and compounds of the formula R P wherein R may be as described above; quaternary phosphines of the formula R POH; organic phosphites; and the like. Bismuth does not readily form such compounds but, again, any compound that will allow bismuth to be in the vapor phase at cracking temperatures will be satisfactory.

Generally, organophosphorus compounds (which are generally unstable at high temperatures and decompose to P 0 are preferred, and phosphine, PH and phosphoric acids, particularly H PO are the most preferred phosphorus compounds, phosphoric acids being still more preferred. Inorganic phosphorus compounds such as the halides are generally not desirable because of their highly corrosive effect and compounds such as the inorganic, phosphates, e.g., trisodium phosphate sodium dihydrogen phosphate, and sodium tripolyphosphates, are not particu larly effective reagents.

While not wishing to be bound by any particular theory, it is believed that the phosphorus, for example, compounds react with the steam to yield an oxide of phosphorus, e.g., P 0 which is vaporous at cracking temperatures. The P 0 vapor then chemisorbs on the walls of the furnace tubes taking up the active sites. Thus, the carbon-promoting catalytic activity of the furnace tube alloy is inhibited.

Generally, it is believed that water is necessary to the proper functioning of this inventive process. Thus, pyrolytic cracking of hydrocarbons, without the presence of steam, in accordance with this invention, would not allow the advantageous results achieved herein. Consequently, the presence of steam is necessary to reduce the partial pressure of the hydrocarbon or petroleum feed fraction, thereby giving higher yields of desired olefins and lower yields of tars; to improve the heat transfer coeflicient; to maintain phosphorus at P 0 which otherwise would make the cracking gas too reducing and the P 0 would decompose to elemental phosphorus. It is apparent then, that for many reasons, steam/water is a necessary ingredient to the feed. However, its advantages are countered by the tendency towards carbon monoxide formation. This tendency has now been virtually eliminated by this inventive process.

In one embodiment of this invention, the phosphorus and/or bismuth-containing compounds mentioned hereinabove can be added to feed stocks containing relatively high sulfur levels, i.e., sulfur levels that would ordinarily cause excessive coking. While sulfur is normally present in almost all petroleum fractions, and, therefore, in many steam cracking feeds, it has been normal practice to treat the entire feed to adjust sulfur levels within close limits to achieve beneficial results in steam cracking. However, by practicing this invention, it is now possible to disregard the sulfur concentration of the feed and treat, e.g., desulfurize, only those product streams that require sulfur removal rather than the entire feed stream. Thus, advantageous results may be had in this invention when sulfur levels are suflicient to cause substantial coke formation (deleterious effects of sulfur will show at different concentrations depending on feed, with higher molecular weight feeds being able to tolerate more sulfur), and ranging from about l0-450 p.p.m. by weight. For example, in ethane cracking, sulfur should be maintained at about 10 ppm. At higher levels of sulfur, coking becomes excessive and continues to increase until the sulfur level is about 200 ppm. At lower sulfur levels, carburization, metal dusting, and CO formation become very high. All of these problems can be overcome with phosphorus addition and sulfur content can de disregarded. With heavier feeds, i.e., gas oils, sulfur content of 1.5 wt. percent is quite normal and up to 3 wt. percent sulfur feed has been cracked successfully.

The steam cracking operation is old and well known (see, for example, Chemical Week, Nov. 13, 1965, p. 72 et seq.), and will only be briefly described hereinbelow. Generally, the petroleum feed fraction is admixed with steam, i.e., in amounts ranging from about 20-80 mol percent steam, preferably 20-60 mol percent steam, more preferably 30-60 mol percent, prior to entry into the steam cracking furnace which may be heated by any suitable means, e.g., gas firing, etc. The furnace itself normally contains two sections, a convection section wherein the feed is vaporized, if not already in that form, and a radiant or cracking section, the feed being passed in admixture with steam through one or more furnace tubes located within the furnace. The convection section is normally employed to increase heating efficiency and the petroleum-steam mixture is heated therein to tempera tures of about 1000-1100 F. However, these temperatures are below that at which the feed cracks since cracking is undesirable in the convection section. The heated feed then passes into the radiant section, i.e., the cracking zone, where the temperature of the reactants is quickly raised to about 1200-1700 F., preferably 1500-1700 F., or higher, as tube metal materials permit, and the feed is cracked. (Generally, raising the temperature of the reactants to the mentioned ranges requires heating the tubes to about 1200-2000 F., preferably 1600-2000 F. and higher as tube materials permit.) Residence times in the radiant section are carefully controlled to minimize polymerization and other undesirable reactions. Thus, residence times in the cracking Zone will range from about 0.1-1O seconds, preferably 0.1-1 second. Pressures within the tubes may range from about 0-50 p.s.i.g. but are not critical, and higher pressures, e.g., up to about 100 p.s.i.g. can be tolerated. Upon exiting the cracking zone, the reaction products are immediately quenched to stop further reaction and/or minimize loss of primary conversion products.

The petroleum fractions which may be converted by this process can vary widely and one skilled in the art will readily determine optimum conditions for different petroleum feeds. Generally, however, the process is most applicable to hydrocarbon feeds consisting essentially of cyclic or acyclic saturated hydrocarbons. Thus, hydrocarbons that may be utilized herein include such cyclic hydrocarbons as cyclopropane, cyclobutane, cyclopentane, cyclohexane, cyclooctane, cyclododecane, etc., and mixtures thereof. Acyclic hydrocarbon feeds include any alkane, namely, aliphatic hydrocarbons of the methane series or mixtures of alkanes with cycloalkanes. Preferred feeds are those saturated hydrocarbons containing from 2 to about 24 carbon atoms, more preferably alkanes containing 2 to about 12 carbon atoms, e.g., ethane, propane, butane, isobutane, hexanes, heptanes, etc., n-hexadecane, eicosane, and light naphthas boiling in the range of 90430 F., gas oils of 450800 F. boiling points, and kerosenes of 430-550 F. boiling points can also be effectively cracked in this process. Because coking can be virtually eliminated by practicing the invention described herein, higher temperatrues in the cracking zone (radiant section) can be employed, the credit for these higher temperatures being taken as increased yields or in the cracking of poor quality feed stocks, i.e., those that would normally give excessive coking.

The phosphorus or bismuth compounds can be provided in the cracking zone by a variety of methods. Thus, it has been found that these compounds exhibit a somewhat lasting effect and continuous addition is not necessary. Nevertheless, depending upon the particular application at hand, it may be desirable to continuously add these inhibiting materials with the feed, or they may be added intermittently, the period being measured by monitoring the carbon monoxide make. Alternatively, the inhibiting materials can be used as a pretreat, with steam, prior to introducing the petroleum feed into the furnace, followed by either intermittent or continuous additions.

Having now described the invention, the following examples will further serve to illustrate the process. However, no limitations, other than those in the appended claims, are to be drawn from these examples, since one skilled in the art will know of many variations and modifications.

Experiments on the effect of sulfur, phosphorus, arenic, antimony, and bisumth as regards their carbon-forming tendencies were conducted.

SULFUR Hydrogen sulfide was dissolved in water, the amount being determined by titration with iodine. The solution was then heated to convert to steam plus hydrogen sulfide and the vapors mixed with preheated ethane (2025 wt. percent steam on total feed) at about 1000 F. The vaporous mixture was fed into a 1'' ID. x 3' long reactor tube of 310 stainless steel. The tube wall temperature was maintained at 1580" F. to obtain the desired conversion. A similar experiment was conducted without the addition of hydrogen sulfide.

In the run without hydrogen sulfide, no coking was observed but the carbon monoxide formation and metal dusting rate were both high. At 10 p.p.m. by weight sulfur on total feed, metal dusting was eliminated and carbon monoxide make was reduced to 0.1 wt. percent. Coking was slight but not excessive. As the level of sulfur addition was increased, carbon monoxide make was no longer measurable. However, the coking rate increased linearly with sulfur addition (as H 8) until a level of about 200 p.p.m. weight of sulfur on total feed was reached. At this level, 12 grams of coke was produced in a two hour test. At higher sulfur levels, the coking rate increased but at a declining rate and a 400 p.p.m. weight of sulfur on total feed, 18 grams of coke were produced in a two hour test run.

ANTIMONY In similar tests to that described above, arsenic in the form of arsenious acid was sprinkled inside a new cracking tube and the sulfur-free ethane-steam test feed mixture was cracked. After two hours, 17.9 grams of coke had formed. In yet another test, arsenic in the form of arsine (AsH was added to the ethane-steam test feed. The coking rate was 2-3 grams/ 2 hours, which is equivalent to a high coking rate in normal commercial steam crackers, e.g., when the feed contains excessive sulfur. Thus, arsenic, too, promotes coke formation in steam cracking operations.

PHOSPHORUS In similar tests as described above, a small amount of phosphoric acid (syrupy, was poured into a new cracking tube and brought to temperature, i.e., 1580 F., with a nitrogen purge. At that temperature, the phosphoric acid evaporated and was swept out with the nitrogen leaving the tube coated with a layer of phosphorus, believed to be a phosphate. In a subsequent cracking test with the ethane-steam test feed, no metal dusting or carbon monoxide formation was observed. This result favorably compares with low sulfur levels. However, the coking rate was surprisingly low, i.e., 0.1 gram/2 hours and this factor offers a drastic difference from the result obtained with low sulfur concentrations.

In subsequent tests, arsenious acid was added to a tube previously treated with phosphoric acid. The coking rate remained at the same low level, i.e., 0.1 gram/2 hours. A similar result was obtained when a high sulfur level ethane-steam, i.e., 200 p.p.m. sulfur, was cracked in the treated tube.

In another experiment, some additional phosphoric acid along with 1000 p.p.m. by weight of sulfur, as hydrogen sulfide, Was added to the tube. The cracking rate after the run was 0.46 gram/2 hours, still a very low rate.

In another experiment, trimethyl phosphate and phosphine, PH in two separate runs, were continuously added at a level of p.p.m. (as phosphorus based on total feed) to the pure ethane-steam test feed. Initially, the phosphorus seemed to have little effect on coking rate or carbon monoxide rate. After 5 hours, however, carbon monoxide make decreased and continued to decrease with continued operation. This factor tends to support the belief that the tube, or at least active catalytic sites, in the tube, become coated with an inhibiting phosphorus compound and that a finite period is required for such coating to form. Subsequently, but during the same continuous run, a high sulfur ethane-steam feed, i.e., 200 p.p.m. weight sulfur, was cracked and a very low coking rate, comparable to that obtained above, was observed; also, the carbon monoxide make was too low to measure. Similar efiects as those reported above for phosphorus are observed for bismuth in an experiment using bismuth oxide.

What is claimed is:

1. In a process for thermally cracking, in a cracking zone, a petroleum fraction by passing the same in admixture with steam through one or more tubes in a cracking furnace, the improvement which comprises providing, in the cracking zone, a material containing phosphorus or bismuth selected from the group consisting of phosphoruscontaining compounds, bismuth-containing compounds, and mixtures thereof, thereby substantially reducing the carbon formation tendencies of the cracking process, the phosphorus or bismuth being contained in the vapor phase in the cracking zone.

2. The process of claim 1 wherein the phosphorus or bismuth is provided in the cracking zone at an amount of at least about 0.001% by weight, based on total feed.

3. The process of claim 1 wherein the bismuth is provided in a bismuth-containing compound.

4. The process of claim 1 wherein the phosphorus is provided as a phosphorus-containing compound.

5. The process of claim 4 wherein an organic phosphorus compound is employed.

6. The process of claim 4 wherein the phosphorus is provided from a material selected from the group consisting of phosphates, phosphites, phosphines, and phosphorus acids.

7. The process of claim 1 wherein the cracking zone is heated to temperatures of about 12001700 F.

8. The process of claim 1 wherein sulfur in amounts sufficient to cause excessive coking is present in the cracking zone.

9. The process of claim 8 wherein sulfur in an amount of at least about 10 p.p.m. by weight, based on total feed, is present.

10. The process of claim 8 wherein sulfur in an amount of at least about 200 ppm. by weight, based on total feed, is present.

11. The process of claim 1 wherein the steam comprises about 20-80 mol percent of the steam-petroleum feed.

12. The process of claim 1 wherein arsenic in amounts suflicient to cause excessive coking is present in the cracking zone.

13. The process of claim 1 wherein the petroleum fraction is a C C saturated hydrocarbon.

14. In a process for thermally cracking, in a cracking zone, a saturated hydrocarbon by passing the same in admixture with from about 2080 mol percent steam,

through one or more tubes in a cracking furnace and subjecting the tubes to heat suflicient to raise the temperature of the reactants in the cracking zone to about 12001700 R, the improvement which comprises providing, in th cracking zone, a phosphorus-containing compound, the phosphorus being contained in the vapor phase in the cracking zone.

15. The process of claim 14 wherein the phosphorus is present in an amount of at least about 0.001 wt. percent based on total feed.

16. The process of claim 15 wherein the amount ranges from about 0.001 to 0.1 wt. percent.

17. The process of claim 14 wherein the phosphorus is provided as an oxide of phosphorus.

18. The process of claim 14 wherein the phosphorus is provided as a phosphoric acid.

19. The process of claim 14 wherein the phosphorus is provided as phosphine.

20. The process of claim 14 wherein the phosphoruscontaining compound is added to the feed continuously.

21. The process of claim 14 wherein the phosphorus compound is added to the feed intermittently, said phosphorus compound introduced to said cracking zone when the carbon monoxide content of the products from said cracking operation is greater than 1.0 wt. percent.

22. The process of claim 14 wherein sulfur is present in amounts above about 10 ppm. by Weight based on total feed.

23. The process of claim 22 wherein the sulfur is present in amounts of at least 200 ppm. by weight.

24. The process of claim 2 wherein said material is P205 Or H3PO4.

25. The process of claim 15 wherein said material is P205 or H3PO4. ,n

26. The process of claim 25 wherein said saturated hydrocarbon has from 2 to 24 carbon atoms.

27. The process of claim 14 wherein said saturated hydrocarbon is a light naphtha boiling in the range of from to 430 F., a kerosene boiling from 430 to 550 F. or a gas oil boiling between 450 and 800 F.

References Cited UNITED STATES PATENTS 1,847,095 3/1932 Mittasch et a1 208-48 3,261,774 7/1966 Ne'wkirk et al. 208-48 3,405,054 10/1968 Arkis et al. 208-48 FOREIGN PATENTS 985,180 3/1965 United Kingdom.

DELBERT E. GANTZ, Primary Examiner G. E. SCHMITKONS, Assistant Examiner US. Cl. X.R. 208; 260683