This invention relates to methods of inhibiting coke or carbon formation on the metal surfaces of processing equipment during high temperature processing or cracking of hydrocarbons by the addition of additives to the hydrocarbon feedstream to be reacted. More particularly, this invention relates to the addition of relatively small amounts of a mixture of polar-solvent-soluble components, including Group IIA (alkaline earth metal) salts, phosphorous-containing compounds and, optionally. Group IA (alkali metal) salts and/or sulfur-containing compounds to the feedstream to be reacted.

BACKGROUND OF THE INVENTION

In conventional pyrolysis processes, a reaction mixture of feed hydrocarbons and steam flows through long coils which are heated by combustion gases to produce ethylene and other olefins, as well as other valuable gaseous by-products. The combustion gases are outside and surrounding the coils and are produced by burning natural gas or fuel oils. Heat is transferred from the combustion gases through the wall of the coil to the reaction mixture, heating it from about 700°C to higher temperatures, typically in the range of about 800 to 950°C. In the last few years, there has been a trend to heat the reaction mixture to the higher temperatures discussed above in order to obtain increased amounts of ethylene production per given amount of feed.

Unfortunately, although coke is always produced as a reaction by-product which is formed and/or collects on the coil walls, the higher temperatures tend to promote or increase this phenomenon. Coke formation has several deleterious effects including the following:

(a) Coke formation on the inner walls of the coils results in increased resistance to heat transfer to the reaction mixture. Thus, a smaller fraction of the heat of combustion is transferred to the reaction mixture and a larger fraction of the heat is lost to the surroundings.

(b) Due to the increased resistance to heat transfer, the temperature of the wall of the coil must be heated to even higher temperatures. This results in increased corrosion or erosion of the coil walls and a shorter life for the expensive high-alloy coils. (c) The coke in the coils results in larger pressure drops for hydrocarbon-stream mixtures flowing through the coils, since the flow paths are more restricted. As a consequence, more energy is required to compress the hydrocarbon product stream in a downstream portion of the process, (d) The coke in the reaction coils restricts the volume in the reaction zone, thereby decreasing the yield of ethylene and other valuable by- products. Hence, more feed hydrocarbons are needed to produce the required amounts of product. Coke formation is also a problem in transferline exchangers (often referred to as TLX's, TLE's, or quench coolers) . The objective of a TLX is to recover much of the sensible heat from the hot product stream leaving the pyrolysis unit. This product stream contains steam, unreacted hydrocarbons and the desired products and by¬ products. High-pressure steam is produced as a valuable

by-product in the TLX, and the product mixture is cooled appreciably. As in the coils of the pyrolysis furnace, coke formation and/or collection in the TLX results in poorer heat transfer resulting in decreased production of high-pressure steam. Coke formation in the TLX also results in additional pressure drop for the product stream.

In current pyrolysis units, coke formation in the pyrolysis coils and/or in the TLX eventually becomes so great that the coils and/or the TLX must be cleaned.

Although various cleaning techniques have been suggested or tried, the pyrolysis unit is usually shut down, i.e., the feedstream flows are suspended. The flow of steam, however, is generally continued since steam reacts slowly with the deposited coke to form gaseous carbon oxides and hydrogen. Moreover, air is often added to the steam. At the high temperatures in the coils, the coke in the coils reacts quite rapidly with the oxygen in the air to form carbon oxides. After several hours, the coke in the coils has generally been completely removed. This cleaning step is frequently referred to as "decoking." The coke in the TLX is not as easily removed or gasified, however, due to the lower temperatures in the TLX as compared to the coil. Cleaning or decoking of the TLX is, thus, often accomplished by mechanical means.

Certain mechanical decoking means have also been used or can be used for cleaning the coil. Decokings frequently require at least one day and sometimes two days. In conventional units, decokings are made approximately every 30 to 60 days. Decokings obviously result in increased downtime relative to ethylene production time, frequently amounting to several percent during the course of a year. Decokings are also relatively expensive and

require appreciable labor. There is hence much incentive to extend the time between decokings.

Numerous suggestions have been made as to how to eliminate or minimize coke formation in ethylene units. For example, improved control of the operating conditions or improved feedstock quality has resulted in small decreases in the rate of coke formation. The cost of making such changes, however, is often high so that these changes are frequently not cost effective. Several processes have been reported in which various additives claimed to be either inhibitors or catalysts are added to the hydrocarbon-steam feedstream. If the additive is an inhibitor, coke (or carbon) formation is inhibited, or minimized. If the additive is a catalyst, reactions between the coke and steam are presumably promoted, or catalyzed. In such a case, the formation of carbon oxides (CO or C0 ₂ ) and hydrogen are promoted. In either case, the net rate of coke that collects on the metal surfaces is decreased. Sulfur is an additive that has been proposed to reduce coke formation in Great Britain Patent No. 1,090,933, German Patent No. 1,234,205, and French Patent No. 1,497,055. At least part of the beneficial effect of sulfur is generally considered to be caused by conversion of metal oxides on the inner surfaces of the walls to metal sulfides. The metal sulfides tend to destroy the catalytic effect of metal oxides which promote coke formation. Although sulfur likely acts as an inhibitor, it also frequently promotes the destruction of the metal walls because the metal sulfides tend to flake off or be lost from the surface. Moreover, one nickel sulfide is a liquid at high temperatures.

Other additives reported include phosphorous pentoxide (See, L.M. Aserizzi, J. Hydrocarbon Process.,

1967, vol. 46, p. 4) and ammonia and ammonium nitrate (See, U.S.S.R. Patent No. 191,726). These latter compounds obviously break down at the high temperatures and oxides of nitrogen are likely to form. Potassium carbonate has also been proposed as a feedstream additive in U.S. Patent No. 2,893,941 to Kohfeldt and Herbert. In using such an additive, provisions must be made to introduce a relatively small but equal amount of the salt to each of several coils in a pyrolysis furnace. One method is to add an aqueous solution of the salt in measured amounts into the feedstream of each pyrolysis unit. As the potassium carbonate is heated in the coil to the pyrolysis temperatures, part or all of it apparently decomposes, perhaps forming K ₂ 0, and part deposits on the coke present on the walls. Such deposits apparently catalyze the gasification reactions between coke and steam so that at typical pyrolysis conditions the net formation of coke on the surfaces of the coils is low if not essentially zero. Corrosion of the inner surface of the coil has, however, been found to be a problem in the process described in U.S. Patent No. 2,893,941. Although details on what causes corrosion in this process are not known, solid deposits resulting from the potassium carbonate are known to sometimes occur, especially if the quantity of the carbonate added is not correctly controlled. Such deposits may cause intercrystalline cracking on the metal surface. Tests have been made in a commercial unit to find operating conditions in which corrosion is not a problem. Various levels of adding the potassium carbonate and different concentrations of aqueous solutions were, for example, investigated, but no suitable set of operating conditions was found. No

conditions were found which resulted in both coke-free surfaces and minimal corrosion.

U.S. Patent No. 4,889,614 to Forester has reported a somewhat similar method for reducing coke formation. He has proposed that either magnesium acetate, magnesium nitrate, magnesium sulfate, calcium acetate, calcium nitrate, or calcium chloride be used as an additive. He investigated all six salts and found that the rate of coke formation on stainless steel surfaces was reduced in the range of 1400 to 2050°F. Such a temperature range is used in all, or at least most, commercial units. He reported the percent reduction in the rates of coke formation or deposition based on numerous runs made with and without the use of one of the salts. He found, however, that corrosion of stainless steel was a major problem. Small, but significant, amounts of Fe ₃ 0 ₄ , Ni0 ₂ , Cr ₂ 0 ₃ , S0 ₂ , and Mn0 ₂ were present in the coke. The laboratory coil had to be replaced after 20-30 laboratory runs, which were normally 160 minutes in length. The process described in U.S. Patent No. 4,889,614 is apparently considerably less effective in removing or minimizing coke deposition as compared to the process of U.S. Patent No. 2,893,941. For example, calcium acetate resulted in a coke reduction of only 24% (see Table II of the '614 patent), although somewhat higher reductions occurred with magnesium nitrate and magnesium sulfate. Moreover, based on the results reported, corrosion would be so severe that the process would likely be of no commercial interest. There is also no indication that the process would be effective to minimize coke formation in the TLX, which operates at much lower temperatures than the coils.

In conclusion, no satisfactory method has to date been reported using additives for controlling coking

problems. Those processes that did control the coking problems resulted in major disadvantages that rendered the process economically unfeasible. SUMMARY OF THE INVENTION In view of the foregoing, it is readily apparent that the prior art has various undesirable drawbacks. In contrast the present invention has resulted in major improvements without any key disadvantages. Advantages of the present invention include all of the following: (a) Increased levels of production of lower olefins, including both ethylene and propylene, are provided.

(b) Time of operation between decokings is substantially lengthened and maintenance problems reduced.

(c) Coke buildup in both the pyrolysis coils and TLX's is reduced. In many cases, essentially no coke accumulates in the coil, resulting in more uniform and more stable operation during the entire pyrolysis cycle. Otherwise, as coke is deposited, small but significant changes in operation are normally required.

(d) Economies relative to energy requirements are improved, including lower fuel requirements for pyrolysis furnaces, greater steam production from TLX's and lower energy requirements for compressors.

(e) The expensive high-alloy steel coils in the pyrolysis furnace and the TLX's are replaced less frequently.

(f) Flexibility to use different hydrocarbons as feedstock is increased.

All of these advantages have been achieved by introducing a mixture of additives to the feedstream of

the pyrolysis unit in amounts effective to maintain corrosion passivation of the internal surfaces of the furnace, while reducing the coke deposition on the surfaces. These additives comprise a mixture of polar- solvent-soluble compounds including Group IIA metal salts and phosphorous-containing compounds, and, optionally, Group IA metal salts and/or sulfur-containing compounds.

Upon further study of the specification and appended claims, further advantages of this invention will become apparent to those skilled in the art. DETAILED DESCRIPTION OF THE INVENTION

Preferred Group IIA salts for the additive mixture are soluble in polar solvents and include inorganic salts and salts of alkanoic acids. More preferred Group IIA salts include alkanoic acids salts of calcium and magnesium containing up to 6 carbon atoms and having either straight or branched chain configurations (e.g., formate, acetate, propionate, iso-propionate, and so forth) and inorganic salts of calcium and magnesium including phosphates and nitrates. These compounds are readily available, cheap, and are easily dissolved in a polar solvent such as water.

Little is known about the catalytic mechanism of Group IIA salts in the process of coke gasification. Studies of the reactivity of various calcium compounds such as calcium oxide, calcium hydroxide, calcium carbonate, calcium sulfate and calcium salts of alkanoic acids exhibit the same reactivity with the same percentage ratio of calcium to coke (equal to 3.57%). At such a ratio, the rate of coke gasification changes from 5 to 5.2 mg/min for all of the above compounds, suggesting that an identical intermediate compound is formed by the primary salts. Without being held to any particular theory, this intermediate is likely to be CaO.

Moreover, calcium salts of alkanoic acids typically break down at a temperature of 500°C into CaO and other compounds, which again suggests that CaO initiates the process. Preferred Group IA salts for the mixture are soluble in polar solvents and include inorganic salts and salts of alkanoic acids. More preferred Group IA salts include alkanoic acid salts of sodium and potassium containing up to 6 carbon atoms and having either straight or branched chain configurations and inorganic salts of sodium and potassium including carbonates, nitrates, phosphates and sulfates. These salts are optionally added to the additive mixture for the pyrolysis of heavy feed materials such as heavy naphtha and gas oils. The reactivity of the Group IA salts during coke gasification is substantially greater than that of the Group IIA salts, permitting a reduction in coke formation during pyrolysis of heavy hydrocarbon feed material with relatively small additions of these salts to the additive mixture. The addition of these salts also apparently reduces the formation of coke in heat exchangers, which considerably increases the operational time of the entire furnace system.

Preferred sulfur-containing compounds are soluble in polar solvents and include both organic and inorganic compounds. More preferred sulfur-containing compounds include Group IA and Group IIA metal sulfides, ammonium sulfide, mercaptans and other organic sulfides. Water soluble sulfides are optionally added to the additive mixture if the hydrocarbon feed to be pyrolyzed does not contain sufficient amounts of sulfur-containing compounds. Sulfur-containing compounds are needed because the use of additives such as alkali metals and/or alkaline earth metals (i.e., Groups IA and IIA from the

Periodic Table of the Elements) can increase the content of carbon monoxide and dioxide (CO, C0 ₂ ) by a factor of several times, necessitating additional measures to effect their removal from the pyrolitic gas mixture. In order to achieve this removal, small concentrations of sulfur containing compounds are introduced into the aforementioned compound if sufficient sulfur is not available in the feedstock.

Preferred phosphorous-containing compounds for the additive mixture include mercaptans, mono-ammonium acid phosphate, phosphorous acids, the phosphates of Group IA and IIA metals and organophosphorus compounds. The preferred phosphorous-containing compounds serve to promote more rapid attack or oxidation of the coke. Although the exact role of these phosphorous-containing compounds is unknown, they are believed to loosen or minimize adhesion of coke to the metal surface, allowing the coke deposit to be carried away by the pyrolysis product. The relative amount of the above additives in the addition mixture is preferably adjusted to obtain the desired reduction in coke formation on the metal surfaces and to simultaneously maintain corrosion passivation of the metal surfaces, resulting in low corrosion levels in the coil and TLX metals.

According to an embodiment of the present invention, the additive mixture should preferably contain about 50 to about 99 wt. % Group IIA metal salt; up to about 33 wt. % Group IA metal salt; up to about 20 wt. % sulfur containing compound and about 0.01 to about 20 wt. % phosphorous-containing compound. It should be taken into consideration that, during pyrolysis of light feed material (e.g., ethane, propane), the compound need not contain potassium and/or sodium salts of carbonic acids;

if these salts are present under such circumstances, their content is preferably minimized. Alternatively, during pyrolysis of heavy feed materials (e.g., naphtha and gas oil) the content of salts of carbonic acids in the compound is preferably provided according to the aforementioned ratio. Moreover, in the event that the hydrocarbon feed contains sufficient sulfur, the addition of ammonium sulfide is preferably minimized.

The preferred method of introducing the additive mixture into the hydrocarbon feedstream is to disperse the additive mixture in polar solvent, followed by the feedstream at an appropriate location upstream of the pyrolysis coils.

Although water is the most preferred polar solvent, methanol, ethanol and/or various other polar organic solvents can be used with good results. Water has the advantage of being readily available and cheap. Concentrations of less than about 1 gram of the additive mixture per liter of solution (or about 0.1 wt. % additives in the solution) are preferred. The compound can be prepared, for example, in a mixer, where the concentration of salts can reach as high as 10% of the total mass. Subsequently, the concentrated solution can be fed into a reservoir, where it is mixed with water until it reaches, for example, a concentration of about

500-700 mg/1 whereupon it can then be introduced into the furnace. The concentration of the solution is not of key importance except to note that significantly more concentrated solutions have been found to promote corrosion or destruction of the coils. Without being held to any specific theory, apparently dilute solutions act to distribute the salts or the residue of the salts more uniformly on the inner walls of the coil and inner walls of the TLX's.

According to an embodiment of the invention, the dilute solution is preferably introduced into the feedstock stream by injection into a coil through which the feed mixture flows. The injection site is preferably located in the preheater section of the pyrolysis furnace about 5-10 meters upstream from the entrance to the pyrolysis coil. This technique was found to be effective in introducing uniform amounts of additive to each coil in the reaction zone of the furnace which is preferably held at a temperature ranging from about 800 to 900°C. Additive mixture expenditure into the furnace is preferably regulated in a range of 0.1 to 250 parts, more preferably 0.5-40 parts, of Group IIA metal per million parts of feedstock, dependent upon the differential pressure of the pyrocoil. For example, when the differential pressure of the pyrocoil is raised about 0.1-0.2 kg/cm ² above the initial pressure, an automatic increase of additive mixture is preferably effected, although it should be mentioned that the maximum amount of the additive mixture is preferably limited because corrosion tends to become a problem at higher concentrations. This method of feeding the additive mixture into the furnace eliminates potential negative effects, such as those arising from deposition of the salts on the metal structure and from the excessive accumulation of salts on the pyrocoil, and permits control of the process.

Alternative methods of introducing the additives to the pyrolysis unit may be found such as by using different solvents, somewhat different concentrations of additives, or even eliminating the use of a solvent altogether (e.g., through the use of a gaseous carrier), but such modifications are considered to be within the skill of those currently practicing the art.

The additive mixtures described above are generally disclosed in terms of their broadest application to the method of prevention or reduction of coke formation and the preservation of surface passivation. Occasionally, the additive mixtures may not be applicable as described. The additive mixtures for which this occurs will be readily recognized by those skilled in the art. In all such cases, the present invention can be successfully practiced, for example, by conventional modifications known to those skilled in the art (e.g., by changing to alternative conventional reagents, by routine modification of reaction conditions, etc.) or by the substitution of other additive mixtures and proportions disclosed herein. In all additive mixtures, all reagents are known or readily preparable from known starting materials.

Without further elaboration, it is believed that one skilled in the art can, using the preceding description, utilize the present invention to its fullest extent. The following preferred specific embodiments are, therefore, to be construed as merely illustrative, and not limitative of the remainder of the disclosure in any way whatsoever. In the following examples, all temperatures are set forth are in degrees Celsius and all parts and percentages are by weight, unless otherwise indicated. Example 1

Comparative pyrolysis runs were made for ethane pyrolyzed in an industrial furnace having four pyrolysis coils and having a total rated capacity of 10,000 kg hydrocarbon feedstock/hr. The exit temperature from each coil was 855°C.

In the plant run made without the additive mixture, sufficient steam was added to the ethane to produce a hydrocarbon/steam mixture that contained 40% by weight

steam. The differential pressure across the pyrolysis coils at an ethylene load of 2500 kg/hr/coil and a steam load of 1000 kg/hr/coil was approximately 1.5 kg/cm ² . Formation of coke was indicated by an increase in differential pressure across the pyrolysis coil as the runs progressed. After 40 days of operation, there was a need to decoke the unit. Significant levels of coke had formed on the inner surfaces of portions of the coil wall, and appreciable amounts of CO and C0 ₂ were produced when the coils were decoked.

A comparative 180 day plant run was also conducted under the same conditions as the first plant run, except that an additive mixture was introduced by means of an aqueous-based solution into the ethane-steam feed mixture. The additive mixture employed during the run was as follows: 97.0 wt. % calcium acetate, 0.1 wt. % ammonium sulfide, and 2.9 wt. % mono-ammonium acid phosphate. The salt mixture was introduced at a concentration of 5-10 ppm during startup and was maintained at this level throughout the run, since no noticeable increase in differential coil pressure was observed over the course of the run. Moreover, during the 180-day active cycle of the furnace, the quantity of steam was set such that the hydrocarbon/steam mixture consisted of 20% steam.

As a result of these changes, the ethylene output was 1.5% higher than that obtained without additives. Moreover, the presence of ammonium sulfide in the additive mixture lowered the formation of CO to a level comparable to that formed in the absence of the additive mixture. This effect can be seen in Table 1. Table 1 illustrates the composition of the pyrogas at the point of discharge from the furnace. Data to the left

represents the furnace with the additive mixture. Data to the right represents furnace without the additive mixture.

TABLE 1

Furnace w ou a t ve m xture was shut own after 40 days for coke burning

"Percentage of product field from feedstock

Finally, no significant amount of coke collected in the coils during any portion of the 180 days of continuous

operation, no substantial change in the pressure across the pyrolysis coil was observed and visual inspection of sections of the coil upon completion of the 180-day run gave no evidence of corrosion. Example 2

Comparative pyrolysis runs were made using a commercial furnace having four coils and a total rated capacity of 20,000 kg hydrocarbon feedstock/hr. The nominal temperature of operation was 835°C. Naphtha with an initial boiling point of 35°C and a final boiling point of 175°C was used as the feed hydrocarbon. The composition of the naphtha was as follows: aromatic hydrocarbons, 6.37 wt. %; cyclical paraffins, 22.03 wt. %; isoparaffins, 26.48 wt. %; n-paraffins, 45.08 wt. %; and sulfur 0.04 wt. %.

In the plant run made without the additive mixture, 5000 kg naphtha/hr/coil were mixed with 3000 kg steam/hr/coil. The pressure drop across each coil was initially 1.4 kg/cm ² . As the unit was operated, the pressure drop increased due to the buildup of coke in the coil. Eventually after about 40 days, significant coke had collected so that the unit had to be shut down and decoked.

A comparative plant run was conducted under the same conditions as the first plant run except that an aqueous- based additive mixture was added to the feed mixture.

The composition of the additive mixture was 99.0 wt. % calcium acetate, 0.99 wt. % potassium acetate and 0.01 wt. % mono-ammonium acid phosphate. No sulfur containing additives were required, because the feed material naturally contained about 0.04 wt. % sulfur. The additive mixture was injected to produce 5-10 ppm of additives in the feedstream. The additive mixture allowed a reduction in steam flow of about 30%.

Over a 180-day run, the pressure drop remained essentially constant across the pyrocoils, and ethylene and propylene production was about 2% higher than that of the first run. Since there was no need to shut down the unit for 180 days, the run extended about 3.3 times longer than the run without additives. In this case, the shutdown was necessitated by coke formation in the TLX.

Essentially, no coke was found in any of the coils of the furnace. Upon completion of the run, the coil and TLX's were inspected. No corrosion problems were noted.

Further results are presented in Table 2. Table 2 illustrates the composition of the pyrogas at the point of discharge from the furnace. Data to the left represents the furnace with the additive mixture. Data to the right represents furnace without the additive mixture.

TABLE 2

Pyrobenzine 21 . 52 /23 . 69 22 . 18/2192 19 . 58/- i9 . β -

Heavy resm initial boiling T 5.5/5.7 5.2/5.9 5.5/- 5.7/- > 200 degC

^' Furnace without the additive mixture is shut down after 40 days for coke burning. "Percentage of product yield from feedstock.

Example 3

Comparative pyrolysis runs were made using a gas oil with a density of 0.81 g/cm ³ . The gas oil had a boiling point range from 190 to 350°C and contained, by weight, 28.00% aromatics, 32.00% cyclic paraffins, 24.13% isoparaffins, 15.60% n-paraffins, and 0.27% sulfur in sulfur-containing hydrocarbons. The furnace had four coils and a rated total capacity of 20,000 kg hydrocarbon feedstock/hr. Pyrolysis was conducted at an exit temperature of 820°C. Runs were conducted with a gas oil flow rate of 5000 kg gas oil/hr/coil and steam flow rates of 3000 kg stea /hr/coil (with additive) and 4500 kg steam/hr/coil (without additive) .

The run without the additive mixture had to be decoked after 40 days. For the run with the additive mixture, the following additive mixture was used (as expressed on a weight basis): 97.3% calcium propionate; 2.4% equal parts potassium propionate and sodium propionate; and 0.3% mono-ammonium acid phosphate. Ammonium sulfide was not used in the additive mixture since the gas oil contained appreciable sulfur atoms. The amount of additives employed in ppm in the feedstream were varied as desired between 0.5 to 40. The flow rate of additives was adjusted to control the pressure drop at a constant value throughout the entire run. Whenever the pressure drop in the coil increased substantially, the rate of additive mixture flow was increased to obtain a

higher ppm of additives in the feedstream. After 90 days of operation, the unit was shut down for survey. Even with the reduced steam flow, no evidence of coke formation in the coils was found; in addition, no coil corrosion was noted.

Further results are presented in Table 3. Table 3 illustrates the composition of the pyrogas at the point of discharge from the furnace. Data to the left represents the furnace with the additive mixture. Data to the right represents furnace without the additive mixture.

TABLE 3

"Furnace without the additive mixture is shut down for coke burning. "Percentage of product yield from feedstock.

Example 4

Table 4 represents the comparative data for pyrolysis runs for naphtha, both with and without the additive mixture. The runs are analogous to, and the additive mixture proportions were the same as, those discussed in Example 2. Flow rates were 5000 kg

naphtha/hr/coil and 3000 kg steam/hr/coil (without additive mixture) and 5000 kg naphtha/hr/coil and 1900 kg steam/hr/coil (with additive mixture) . Temperature upon exit from the furnace was 835°C. Moreover, the additive mixture was the same as that in Example 2. The level of additives used during the course of the additive mixture run varied from about 5-20 ppm of feedstock, depending upon the differential pressure across the pyrocoil. Table 4 illustrates the composition of the pyrogas at the point of discharge from the furnace. Data to the left represents the furnace with the additive mixture. Data to the right represents furnace without the additive mixture.

Without the additive mixture, the furnace had to be decoked after 40 days, whereas the furnace operated for 150 days with the additive mixture disclosed in Example 2. Even after 150 days, no coke had formed in the coils for this latter run.

The outer wall temperatures presented in Table 4 were measured using a pyrometer. No substantial change in the temperature of the pyrocoil walls of the furnace was noted using the additive mixture throughout the 150- day run. In the run where no additive mixture was used, a steady elevation in temperature was observed which reached a maximum after 40 days of run time. As the temperature of the pyrocoil walls increased, the

differential pressure across the pyrocoils increased as well. Both effects indicate the growth of coke deposits on the inner tubular wall of the pyrocoils.

Carbon monoxide content at beginning and end of the run using the additive mixture was practically constant at 0.1% of mass, which is only 25-30% greater than that of the furnace not using the additive mixture. A similar effect was also seen above in Tables 1, 2 and 3. From these tables, it is also apparent that, with the addition of the additive mixture, the concentration of C0 ₂ in the pyrogas upon discharge from the furnace is increased about 20-30%, which suggests the occurrence of coke deposit gasification with the presence of the additive mixture. Table 5 presents data pertaining to C0 ₂ content of the pyrogas in Example 4, before C0 ₂ is removed in the purification unit.

TABLE 5

C0 ₂ Content, Projected Values, Actual Values,

% of mass % of mass % of mass

Average 0.004

Maximum Not exceeding 0.2 0.02

Minimum 0.001

It is also important to note that maintenance of the proposed rate of additive mixture flow based on the differential pressure of the pyrocoil permits pyrolysis to be conducted with relatively low releases of CO and C0 ₂ . Only when large amounts of coke are deposited due to uneven (nonuniform) additive mixture flow is excessive CO and C0 ₂ formation possible during decoking. Therefore, the additive mixture flow technology as described herein is useful for the management of CO and C0 ₂ formation in conjunction with the desired compounds. The presence of ammonium sulfide, (NH ₄ ) ₂ S, and mono-ammonium acid phosphate, (NH ₄ )H ₂ P0 ₄ , salts, also discourages the formation of unwanted CO and C0 ₂ impurities: ammonium sulfide decomposes in the furnace with the formation of hydrogen sulfide, H ₂ S, which suppresses the reaction that initiates CO formation; mono-ammonium acid phosphate loosens coke buildup on the tubular walls, permitting partial physical removal of coke from the pyrocoil via flow of the pyrolysis products. Moreover, as seen from Example 4 (and the preceding examples) , the use of the additive mixture increases furnace run time by a factor of about 2-3. The output of high pressure steam from the heat exchangers of the TLX was also seen to increase by about 30% due to the lowered (2-3 times lower) rate of coke and resin, formation in the heat exchanger tubes.

The additive mixture also effectively reduces coke deposition in the TLX's, especially in the inlet portion of the unit. In Example 4, the inlet (high temperature) portion and up to 60-70% of the TLX's were completely free of coke during the entire 150-day run. Toward the exit (low temperature) portion of the TLX, small coke deposits were found. These coke deposits were analyzed upon completion of the 150-day study. The results are shown in Table 6, wherein the upper data represents furnace with additive mixture and the lower data represents furnace without additive mixture.

Ca Content Fe Content Ni Content Carbon in terms of in terms of in terms of Content CaO Fe-0 ₃ , NiO % of mass % of mass % of mass

7.0 trace trace trace 83.0

3.3 1.0 0.02 0.012 87.28

As is apparent from the data in Table 6, the Ca content in terms of CaO is increased in the furnace using additive mixture from 3.3 to 7%, indicating the presence of Ca in the TLX and its activity in the coke gasification reaction. Moreover, the absence of Fe, Cr and Ni in the coke deposits of the furnace using additive mixture indicates an absence of corrosion in the pyrocoils and tubes of the TLX.

The preceding examples can be repeated with similar success by substituting the generically or specifically described reactants and/or operating conditions of this invention for those used in the preceding examples. From the foregoing description, one skilled in the art can easily ascertain the essential characteristics of this invention, and without departing from the spirit and scope thereof, can make various changes and modifications of the invention to adapt it to various usages and conditions.