US1847238A - Process of treating hydrocarbons - Google Patents

Description

March 1,

Gaps Inlet,

1932. F. E. FREY ET AL 1,347,238 Y PROCESS OF TREATING HYDROCARBONS Filed May 21, 1929 2 Sheets-sheet 1 5 Oufifiefi vlgvi. w 9 M 7 --r 1 v f ,j I t l Exofihermzc Cracking Chamber & scrubbqr Extradim Flam?! Z l J Bcnzol Guild affier exo kerm ic cracking Thermocouple g 157M317 Gas to 6e 4/ undergo exothermic 2i cracking 5 .Colal diluent gas Z'q I confirol extent of cradrvng ZQW 2 RE. Frey -A- wycr d. 5 G.G.Obcrfcll z INVENTORS 4 i f I ATTORNEY Patented Mar. 1, 19 32 "UNITED STATES PATENT oFFica 7 FBEDERICIK E. FREY, JESSE A. GUYER, 4ND GEORGE G. OBERFEIL, OF BARTLESVILLE,

OKLAHOMA, A.SSIGNORS 'IO PHILLIPS PETROLEUM COMPANY, OE BARTLESVILLE,

GKLAHOMA IEROGESS or TREATING. nxnRocAnBoNs" Application filed May 21, 1929. Serial No. 364,810.

The present invention relates to the cracking of hydrocarbons in the gaseous and/or Vapor phase to convert the same principally into aromatic compounds typified by benaene,

toluene, and xylene.

It has been proposed to convert propane,

butane and the like, partially into crude benzol by cracking both in the presence of and in the absence of catalysts. The prior art 10 workers, in an empirical way, have cracked ethane, propane and butane and mixtures thereof in a heated tube, and obtained light oils, gaseous products and a heavy tar.

' However, in investigations of this character, the chemical reactions occurring were not thoroughly understood, and insufli'cient importance was attached to the time factor involved in the conversion of the hydrocarbons. We have made extensive investigations to study in a somewhat exhaustive manner the sequence of changes occurring during the I cracking of hydrocarbon gases, and particularly propane and butane into aromatic compounds and oils and tar, and to further de- 2 termine in a fairly quantitative degree the effect of temperature on the nature of these consecutive changes and-on the velocity with which they occur; Our. experiments show that at temperatures of 1250 to 17 50? F., and under atmospheric pressure, the cracking of hydrocarbons, typified by propane, butane andjmixtures of the same, proceeds endothermically with an increase in volume and theformation' of simple gaseous olefines to a maximum of 35 to 50% by volume, by familiar reactions in which the paraflin molecule splits into-two molecules, one an'olefine, and the other a simple paraffin 'or hydrogen. We have ascer- 40 tained definitely that in the case of propane or butane, the heat absorbed by the endother mic reaction reaches a maximum of about 700 B. t. u. per pound'at approximately 1562.

F., at which state of maximum heat content,

the olefine content is approximately at the maximum. At approximately the tempera 'ture specified, this condition isattained in somewhat less than 0.002 minutes. Athigher cracking temperatures within the range (1250 to 17 F.) both olefine content and heat content attain a somewhat higher maximum value than at the lower ones, The time'- consumed by this stage of reaction decreases rapidly with increase in the temperature at which it occurs.

The higher temperatures of cracking as beforestated give a higher content of the olefines than the lower temperature for that time of cracking, characteristic for each temperature, giving the maximum volume percent of olefines. The chief reason for this is that the decomposition of ethane to ethylene and hydrogen C H C H +'H 37900 cal. is reversible and rapid and endothermic at temperatures within our cracking range. The higher the temperature,'the more completely is'the ethane dissociated, at equilibrium. Ethane, ethylene and hydrogen are all formed in the early cracking reactions,

"and, though influenced of course by other reactions in which ethylene is destroyed, a hydrogenation equilibrium is nevertheless approached. As a result, much of the ethane we find in the gases produced bytcracking. at a low cracking temperature, appears as ethylene and hydrogen at a high cracking temperature and augment the unsaturates content of the gas, and, since the dissociation is endothermic, the heat content of vthe gas. The unsaturatesattains a maximum 10-15 percent higher at 1562 F. cracking temperature, than at 1112 F.

Our experiments. indicate that. continued exposure to the cracking temperature causes the olefines produced by the initial cracking to undergo an exothermic conversion or de-' composition into aromatic oils and tar after a period of no less than tentimes as long as that required to e'flect the endothermic decomposition or conversion.

In endothermic cracking, it happens that the reaction in the useful temperature range (1250-17 is so rapid that a practical coefficient of heat transfer through confining surfaces as in a tube coil, develops a temperature at which the reaction-will just absorb the heat furnished. Furthermore, in the early stages of endothermic cracking, the velocity of cracking is greater than in the later stages, and a somewhat lower temperature is therefore developed than in the later stages of endothermic cracking. The time of cracking at single temperature, moreover, is more or less incidental to the economically rapid introduction of the endothermic heat. In our experience, a tube coil operating at full capacity permitted the introduction of all the endothermic heat with an exit temperature in the gas of 14:004425 F. An exit temperature above l i50 F. showed the instability characteristic of an incipient exothermic reaction. The value of 0.002 minutes at 1562 F., is a maximum value, and an exposure time for exothermic cracking ten times thatlong, 0.02 minutes, is a minimum value for developing a maximum amount of volatile oils. Unlike the endothermic cracking, this value has been determined fairly accurately by us and is long enough to be significant and indicate that a reaction chamher is desirable, and indicate the size to be used. A value for this ratio as high as 50 is permissible, corresponding to an exothermic cracking time of 0.10 minutes which gives a substantially optimum yield of volatile oils also, but how much greater the maximum value would be, since the time of endothermic cracking given of 0.002 minutes is a maximum value we have not determined.

' A maximum yield of benzene and toluene is thus developed, accompanied by a somewhat less amount of tar, and very little carbon. During this period, relatively little change occurs in the volume of the gas. However, of course, the olefine content of the gas decreases progressively. The heat evolved during this stage in cracking propane or butane to a maximum yield of simple aromatics, is about 350 B. t. u. per pound, at a cracking temperature of 1552 F., at which temperature a period of 0.04 to 0.08 minutes is required.

We have discovered that the temperature is related to the duration of this exothermic cracking stage by the empirical formula T: 124.5 180 log t in which T=temperature in degrees Fahrenheit and t=time in minutes.

This formula applies for temperatures varying approximately between 1250 to 1550 F. It is somewhat less accurate when utilizin higher temperatures up to'17 50 i ln regard tosaid formula, it may be stated that experiments were conducted with the object of determining the role of cracking time as well as temperature in producing an optimum yleld of benzol. A gas of the composition was submitted to cracking in a non-catalytic silica tube under conditions suitable for the A production of benzol. The experiments were conducted at atmospheric pressure and the carbon, tar, benzol and gases were separately measured. The effect of time and temperature on benzol yield is shown in the following table, in which time is expressed in minutes and benzol yield in gallons per thousand cubic feet of gas.

TabZe.Efi'ect of time of cracking on benzol yield at several temperatures.

Time Yield 1292" F.

At a iven temperature, the yield of benzolity without greatly influencing the sequence of changes taking place'during cracking. The maximum yield of benzol was roughly constant throughout the rather wide temperature range studied within the time range appropriate to and unique for each temperature.

The data of the table show the minimum time required to obtain a substantially optimum yield of benzol at several temperatures withina rather widerange. The values are 1 approximately 0.5 minutes at1292 F., 0.2 minutes at 1382 F., 0.012 minutes at 1562 F., and somewhat less than 0.003 minutes at 1742 These values may be used to determine a time-temperature relation which is expressed in a compact form by the equation .T= 1245 180 log t,

found to be required for forming a yield of benzol, it is obvious that the same timetemperature relation is applicableto the treatment of gaseous olefines to form benzol and to all the simpler paraflins which decompose into gaseous paraffins and gaseous olefines. 'Methane, because of its great stability to heat is 'not converted into henzol under the above mentioned time-temperature conditions.

This formula gives the minimum time of cracking to develop a substantially optimum yield of light oil at any temperature within the range. Since the oil yield in the early stages of exothermic cracking increases rapidly with prolonging of exposure, the for mula may, perhaps, give an oil-yield as much as 25 percent less than optimum, but it does define the more sensitive lower exposure time limit. An exposure of from 2 to 6 times this length will actually give an optimum light oil yield, and exposure times in the range 2 to 6 times the formula value would actuallybe used, the selection depending ,on gas-depletion, tar yield permissible, etc., within the range 12501550 F. From 1550 to 1750 F.

the high rate of self-heating and complications due to heat transfer, renderedour work less accurate, but we have actually obtained optimum yields at exposure times twice that calculated by the formula at temperatures up to 1922 F.; well above the range.

A several fold increase in the period of cracking above the shortest period above specified, which will givea virtually optimum yield of volatile oil causes relatively little change in yield. The oil produced at 1562. F. by the cracking period of .02 min-' utes given by the formula contains 20% .or,

so of unsaturates, chiefly butadiene and cy- "clopentadiene. The remainder is benzene,

The oil produced .by'

toluene and xylene. a longer cracking period, say .10 minutes contains over benzene, and less than 3% unsaturated hydrocarbons; the optimum yield of volatile oil is substantially constant over the temperature range of. 1250 F. to

We propose to take advantage of the exotherinic stage of the cracking operation and to control said'stage so as to crack hydrocarbon gases to form oils in a reaction chamber in which converslon occurs at a temperature advantageously hi her than that of the gas,

and this without t e addition of extraneous heat duringthe exothermic stage, this being.

possible by reason of a predominance of exothermic reaction. Broadly stated, our disf coveries indicate that the oil forming stage in the cracking operation takes place ,exothe'rmically within the approximate ran 0 of 950 F. to 17 50* F., the range preferab y being between .1100 F. and '1750 F; The

composition of the gas to be exothermically cracked or converted, may vary greatly. We have ascertained that parafiins higher than methane will absorb heat in cracking. Gaseous olefines, especially ethylene, have a posi tive heat of formation. In cracking to form oils of higher carbon content, methane, which has a high negative heat of formation, is formed in a relatively large amount, whet-her hydrogen is, or isnot present. These facts account for most of the exothermic effect. The maximum exothermic effect will be obtained with a gas containing a maximum concentration of gaseous olefines and a minimum concentration ofparafiins higher than methane. The calculated temperature rise during the exothermic stage of the cracking of butane, for e'Xample,-.

at approximately 1562- F., to give a maximum yield of simple aromatics, is approximately 350 -F. A temperature rise, from this cause, ofover 200 F. has been obtained in large scale operations, which will be specifically set forth hereafter. If a rise in temperature of 50 F. be considered the minimum which will give a practical advantage, the weight percent of gaseous olefines need not greatly exceed that of the paraflins higher than methane.

In this connection it maybe well at this time to direct attention to the accompanying drawings, in which Figs. 1,2,3 and 4 are diagrammatic views, partly in vertical section, ofsuitable apparatus for use in practicing our improved processes.

Referring to Fig. 1, we have passed, for example, 8924 cubic feet per hour of av as consisting chiefly of butane through a. coil 5 .of three and one-half inch tube, 200 feet long, into which heat was introduced by convection of combustion gases in the cooler parts, and radiation in the hotter. As the I .rate offiring was increased, the temperature of the gas" leaving the tube coilincreased rapidly, until it reached 1300 F. After reaching this point, the temperature rose very slowly to 1400 F. while the rate of firing was greatly increased. The linear temperature gradientin the tube coil was from 1300 at the beginning to 1400 F. at-the end of the last section of the coil, in which-section the heat-increment was lmown to be many times large enough to heat the gas 100 F. Both I by the formula, while causing depletion of the these effects show heat absorption to be due sorption is such that 20 pounds will absorb 80 200=16,000 B. t. u. in the time of contact of the gas, which was about 0.005 minutes in the section of the tube coil where cracking took place. The temperature would show almost no gradient from point to point of the cracking section if the reaction rate did not decrease as the original constituents were destroyed, but since it does, a small temperature gradient is to. be expected, the equilibrium temperature attained being somewhat higher near the end of the endothermic stage of cracking than at the beginning.

when enough heat was introduced to raise the temperature of the gas leaving the tube to above.1450 F., the temperature fluctuated, as would be expected from the onset of the exothermic cracking, whichwould magnify fluctuations in gas rate or heat input. In our example, gas issuing at 1435 Efrem the tube coil* with a specific gravity of 0.94 (air=l.00) and over 40 percent unsaturates was passed through an insulated cracking chamber 6, of. 273 cubic feet capacity, in which the temperature rose to 1520 F. and the gas left the chamber at 7 with a specific gravity of 0.55 and a high content of benzol. This corresponds to a 0.20 minute duration of exothermic cracking at a mean effective temperature of 1500 F. which is six times as long as the minimum time of 0.037 minutes for developing an optimum light oil yield. A much longer exposure than 0.20 minutes would result in destruction of light oil, but" we have found an exposure atthis temperature six fold greater than the minimum given cracked gas in heating value, produces a maximum yield of light oil more free from gum forming unsaturates than with the minimum exposure time.

The converted gaseous mixture leaving the reaction chamber 6, may be passed by suitable conduits 8 and 9, first through a scrubber 10, and afterwards through an extraction plant 11; the light. oils being separated from the gas in the latter, and being partially suitable for motor fuel purposes.

As indicated, our process of exothermically cracking hydrocarbons may be applied to other hydrocarbons than the gaseous parafiins. For example, the gases formed in the pressure distillation of petroleum, contain paraflins higher than methane, as well as gaseous olefines, and this product suitable as the initial starting material. A certain amount of endothermic cracking, somewhat less than that required for ethane, propane, and butane, is necessary before such gases are in a suitable condition to undergo exothermic crackmg.

As an example of pressure still gases suitable for treatment by our process, the following is typical:

CO2 and 1- Olefines 9.

Hydrogen 5.

Nitrogen "*"T'f Calorific value 1800 B. t. 11. per cubic foot. The calorific value indicates 50 percent by weight or so of paraflins higher than methane.

The olefines are largely propylene and butylenc. This gas would require some endothermic cracking at about 1400 to initiate exothermic cracking. A propane-butane concen- ,trate of cracking still gases would be even more suitable for cracking to benzol. Such a material containing 5 percent propylene, 25 percent propane, 35 percent butylenes, 35 percent butane has been obtained by us from pressure still gases.

The initial starting material may furthermore be the gases formed in vapor phase cracking, and these are particularly suitable since they contain as high, in some cases, as 75% olefines. Such gases will crack exothermically after only such a pre-heating as is necessary to initiate a moderate rate of exothermic reaction. As an example of a gas produced by vapor phase cracking, it may contain 50 percent unsaturates, with a calorific value of 1800 B. t. u. With the cracking conditions used, such a gas would contain less ethane than unsaturates, and would need little if any endothermic cracking to initiate exothermic cracking.

Petroleum may be cracked at temperatures of 1200 to 1600 F. almost wholly toproduce gases containing a large proportion of gaseous olefines, and such gases will undergo, according to the present invention, exothermic cracking. I

From the above, it is seen that gases typified by ethane, propane and butane, and mu;- tures of these gases, are fairly rapidly endothermically cracked at temperatures somewhat above 1250 F. with a considerable absorption of heat and with the formation of It is to be noted that an increase in temperature 'infboth stages of the -cracking operatlon,

that'is, theendothermio stage and the exothermic stage, decreases the time consumed in each stage to crack and obtain the desired results.

Broadly, in accordance with our 1nvent1on,

we propose to conduct thecracking operation in.two stages, an endothermic stage, and. an exothermic stage, and this, under peculiarly suitable circumstances for each stage; and

in general, theiheatrequired forthe endothermic stage, may be introduced through any confining surface. For example, the gas may be passed through the coil tube 5, through which heat is applied externally. The temperature obtained in difierentparts of the tube are, generally speaking, the result of a balance between the rate of endothermic reactionand the rate of heat introduction.

Our experiments indicate that a temperature range of 1150 to 1500 embraces the more practical values. Owing to the high velocity of the reactions occuring, cracking may beconducted economically in large scale operations at a temperature between 1300 F. and

1450 F. with the gas at a pressure of only a little above atmospheric.

The temperature balance has been explained above. It may bementioned that it can be .put moreclearly in mathematicalform, but this would not be especially useful, since the-several controlling factors cannot all be evaluated, as, for example, the

change of velocity of endothermic heat ab sorption with extent of endothermic cracking, but the practical temperature range of 1300 F. to 1450 F. or 15 0O F. for conducting the endothermic cracking near atmospheric pressure is practicahle for an iron or alloy tube coil in'which the maximum practical coeflicient of heat transfer is'used.

The exothermic stage may be conducted in the chamber 6, through the walls of which no heat is introduced, and in which the carbon which forms during'the operation may deposit. 4

Regardingthe insulated chamber, we have used acylindrical'chamber of 273 cubic feet capacity of a length of about four times the diameter. The insulation wassupplied by building the wall 18 inches thick 0 and Silocel insulation brick. The lining is of firebrick. It is encased in a steel shell and stands on one end. The gas, after endothermiccracking was introduced atthe bottom at a temperature near that at which itleft the tube coil and was. discharged fromthe top after exothermic cracking had taken -place during its passage. The direction of flowj from bottom to top is preferred to t e,reverse direction, because the small decrease in specific gravity during exothermic cracking would, in the latter .arrangement,,tend to cause circulating and consequently, discharge part of the 'larly desirable i gas undercracked and part overcracked. 1

sulation, heat developed by the reaction may be used to attain this higher temperature without mixing additional heating gases Our experimentsindicate further that the which would lower the quality of the gases.

produced by the process, andthis is particuthe initial starting materials are such and the process is carried out soas to produce a fuel gas. In extreme cases, however it may be necessary to impart a helpful temperature increment prior to the exothermic cracking stage. However, if this is necessary, and if heated gases are used for this purpose, only a small amount thereof is necessary to impart a helpful temperature increment, since little of-the added heat is absorbed by endothermic reaction.

Regarding the addition of heat prior to exothermic cracking in the chamber 6a see Fig. 2), the following is an example: If the raw gas were simply heated to 1300 F. about 1300 B. t. u. per pound would be required.

If products of combustion at nearly flame temperature were added by way ofpipe 12, to complete endothermic cracking, enough would need to be added to impart ZOO-B. t. u. per pound of heat of reaction, whereupon exothermic reaction would set in and a gas be produced containing benzol and perhaps I 35 per cent inerts. This not only would reduce the quality of the gas, but, since the presence of tar and carbon suspension would rule out good heat recovery in some cases, the sensible heat carried by the inerts would need to be lost. If, however, the raw gaswere heated t6 0 F. and then all the endother- 15 mic heat introduced in-a tube coil within the range 13001400 F. no such heating; gas

. III

would need to be added and yet, the tube-coil would be exposed to a temperature'only 100 F. higher. If a checkered paSSageWQ 6 ternatelyheated and-used for endothermic cracking, or'if the' gaswere cracked under superatmospheric pressure at a lower temperature by reason of a longer time of residence in the tube: coil, or'if the endothrmically cracked gas'were conducted. through an interconnecting tube and cracking chamber, in which a loss'of IOU-400 F.- of sensible heat was permitted, the gas supplied to the cracking chamber 6 or 6;; would be endothermical ly cracked, but toocool sometimes to undermic cracking. The gas produced would contain only a little inerts.

In the case pf gases such as those from' 10 vapor phase cracking, whichwould absorb a negligible amount of'endothermic heat, the

gas could be preheated to 1100 F. in the tube coil 50, at which temperature the life of the tube would beflonger and the. cost lower, than at higher temperatures. The gas'fcould then be heated to1400 F. by introducing hot products of combustion through pipe 12' f (rapid mixing should reduce heat absorption from the water gas reactions forming carbon monoxide), and the gas could then undergo self-heating and exothermic cracking to henzol. is

We will now proceed to described the process broadly, and, thereafter, several variations thereof.

Gases containing a substantial amount, preferably over 15% of hydrocarbons higher than methane, that is, hydrocarbons such as ethane, propane, and butane, are subjected, preferably in the tube coil .5 or 5a, to sufiicient preheating accompanied 'by endothermic cracking toinitiate the exothermic stage of cracking to produce the olefines, which is then the conditions previously described, until a maximum yield of aromatic compounds and oils has developed. For example; butane may be passed through the tube coil in which it is cracked, preferably within a temperature range of 1300 to 1400 F., and discharged at 1450 F. with a specific gravity of 1.08 (air=1.00), and anhnsaturated content of over by yolume. This gaseous product may then be passed through insulated -reaction chamber at the exit of which temperature ranges of preferably 14:50" to l650 F. are developed. At temperatures within this range, the optimum yield of aromatic com= pounds, or oil's, is obtained.

'However, if it is desired, the process, by a slight variation, will produce carbon and drogen as the substantial end products. This will occur if the exothermic reaction is allowed to proceed far enough to a temperature above 1600 F., and that temperature is maintained long enough to accomplish the desired decompositionf The hydrogen containing gas produced in accordance with the above will contain methane. -However, this may be reduced by using in the exothermic reaction chamber a suitable catalyst, such as carbon. Equivalent means may be used to increase thehydrogen content, and thereby make the hydrogen suitable for hydrogenation.

allowed to proceed'in the chamberv 6, under' If hydrogen for hydrogenation is desired, the chamber 6 or 6a may be filled with coke to act as a catalyst for increasing the hydro,- gen content, but such means cannot be used to increase it greatly. The gas is, however, suitable for conversion into a purer hydrogen by other means, such as decomposition with steam.

The exothermic cracking chambers may be subjected to fluctuations in temperature, and 7 it is desirable to provide means for reducing the temperature of fluctuations and inequalities during the cracking operation. This may be accomplished by disposing-fire-brick or other suitable heat absorbing material in the exothermic cracking chamber.

The use of a checker ,brick fillingfor the reaction chamber, as indicated at 14.- in Fig. 3, would not remove fluctuations in temperature of long period, due to changes in self-heatas flow rate or temperature of incoming gases varied, but the brick will smooth out fluctuations of short period by absorbing heat when the temperature of the gases rises above the brick temperature, and give it off when the gas falls below. Fluctuations of longer period or temperature drift however might need to be controlled by introduction of cold diluent gases through the pipe 12 in Fig. 2.

The checker brick might be of further advantage in smoothing out inequalities of cracking in this, that the small pressure gradient it would cause might assist in giving all portions of gas an equal time of passage through the chamber, as by reducing edd currents out of the main path of the gas. X vertical cylindrical cracking chamber filled with crackerbrick could be used. It might, for example, have about the same cubic content of empty space as if not brickwork were used, but brick would take up about 46% of space," making .the chamber somewhat larger.

Fluctuations in the rate of flow, the temperatures and composition of the gases enter-- 1 ing the chamber to undergo exothermic cracking, are magnified in;the temperature attained in the exit end of the chamber 6, 6a or (SbQbecause of the high rate of self-heating at the high temperature developed. Since 5 small variations in extent of cracking greatly effect the characteristics of the final product,

the extent of cracking must be confined within narrow limits. When the rate of flow of the gas undergoing cracking'and heat input. preceding introduction to the cracking chamber is varied, as a means of maintaining a constant extent of cracking owing to the inertia of the system, due chiefly to storage of heat, considerable fluctuation in cracking re-'* sults, and excessive cracking occurs periodi cally.

A closer control may be obtained. by the introduction of a cooling diluent in a small 7 amount through the pipe 12, as heretofore mentioned, either in the gas before it enters I the exothermic reactioli chamber, or at some drop due to heat loss through the chamber walls. The temperature rise, due to selfheating, depends on the composition of .the gds entering the chamber, the temperature at which it enters, and the temperature-time relation as it passes through the chamber.

For a given constant composition, temperature, and velocity of gas entering the reaction chamber, there would be a constant extentof. cranking maximum temperature attained if no heat were thrown back by the self-heated gas to'gas which had as yet .undergone less self-heating, but since there is, the heat thrown back canincreas'e the rate of self-heating gradually even under these constant conditions of the entering gas. The

- resulting temperature climb would eventually result in overcracking. The rate of firing the tube, coil 5% in which the endothermic cracking took place could he suddenly decreased to. arrest the temperature rise. But

if this were done, the temperature climb in the top oithe cracking chamber could still go on while the furnace reached the new lower temperature equilibrium, and when the temperature rise did stop, if the rate of heating had beenreduced a little too much,.in order to bring about a quick response, the chamber temperature would drop, less heat would be thrown back, less self-heating thentake place, and the chamber temperature would.

continue to drop. Heat loss through the chamber walls. could reduce this fluctuation somewhat, but this would necessitate more heat introduction in the tube coil, which it is desirable to keep at a minimum. The efl'ect ofmagnifying in outlet temperature small fluctuations in inlet temperatureisthen' to be expected.

The introduction of cold diluent gas through pipe 12to arrest temperature climb, does so,'both by reducingthe inlet gas temperature, and hence rate of initial self-heating, and by decreasing the time of passage,

permit the exothermic cracking temperature to drop off, so a slight average excess of heat would need to be introduced. The amount of diluent gas would need to be great enough. to prevent excessive temperature rise, when the fluctuation in operating conditions causes maximum values. With ideally constant conditions in the tube furnace, the correctdegree of exothermic heating would take place, and the amount of diluent gas would be vanishingly small. Presumably, with the firing ot the tube furnace and the flow rate of raw gas under good control, the amount of diluent would neverneed to exceed 5% of the gas being cracked, and the amountwould vary between 5% and zero as fluctuations were oflset when they appeared. 4

The temperature attained, the unsaturates content and specific gravity of the completely cracked gas, are all properties which vary greatly enough with extent of cracking to perniittheir continuous determination to be used as an index of extent of cracking and means of controllingthe introduction of di-' luent gas. Temperature is a convenient lndex to use, and it could be'used either to regulate .the rate of continuous introduction of diluent gas or percentage of time open of an openand-shut valve device. For example, the introduction of the cold diluent gas may be advantageously controlled by use ofa thermocouple 15 (Fig. 2) or the like, which utilizes conditions at the top of the cracking chamber 6a to control the'operation of valve means 16 which regulates the admission ofthe cold gas.

Our experiments also indicate that the formation. of carbon proceeds rapidly in the exothermic cracking stage, if the cracking continues after the maximum yield of benzol has developed. The'high rate. of cracking at the conclusion of the exothermicstage aggravates the trouble dueto'carbon forma-,

tion. The extent of cracking may be better controlled, and carbon decomposition in the exit pipe 17 (Fig; 3) reduced by introducing at thepoint of exit from the chamber, enough cold gas to' reduce the temperature to such a degree that further cracking is virtually arrested. .The introduction of cold-gas to arrest cracking would have to be done between theexit 17 and the thermocouple 1'8 measuring the maximum temperature developed and used to control extent of cracking. The amount to be introduced through pipes 19, for instance, to reduce the temperaturefrom a reasonable value of 1550 ,F. to 1400 F., at which latter temperature cracking would be nearly arrested, would be about 10% byweight of the gas'cracked. The introduction will be contiguous and at a and hence time of self-heating. Both these Hired rate.

efl'ects take place at once on introduction of cold gas. The heat introduced in the tube coil 5 a would need always to be so great that no Since the velocity of the exothermic reaction is greatly increased by a rise of tem- 'perature, the accelerating of the temperature I fluctuation in operating conditions would rise, especially in the early stage oftheexothermic cracking, will permit a smaller size for the reaction chamber for a given amount of gas to be put therethrough. This may be accomplished by introducing the endothermically treated gas soas to provide turbulence,

thereby mixing the incoming gas with the gas which has already undergone some selfheating. In the last stages of cracking, however, turbulence is undesirable, because a portion of the undesirable liquid unsaturates. and olefine gases will be discharged before their conversion occurs. In the latter stage, moreover, temperature instability becomes serious, and the state of cracking must be controlled. This control may be accomplished by providing an exothermic cracking chamber 60 which consists of two adjoining chambers 6dand 6e, communicating by one or more constricted openings 6 f. The gas enters the lower chamber through a pipe 20, directed in such a manner that active turbulence is induced in the chamber 6d to cause mixture of the incoming gas with gas that has already undergone some exothermic cracking. The gas then passes through the restricted opening (where cold gas is introduced through a pipe 21 at an adjusted rate 'to maintain a constant extent of cracking),

into the adjoining chamber 66, through which it moves in an upward direction to an exit 22, at which cold gas may be introduced, if de-, sired, (as at 19 in Fig. 3) to arrest the crackmg.

Our experiments, in which hydrocarbons are cracked first endothermally, and then exothermally, while controlling the relationship between the temperature-em'ployed in and the duration of each cracking stage to produce an optimum yield of aromatic compounds, have been carried out under atmospheric pressure or slightly above, but it is clear that the process may be carried out at higher pressures.

The terms and expressions employed herein are used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions, of excluding any equivalents of the features shown and described, or portions thereof, butit is recognized that various modifications are possiblewithin the scope of the invention claimed.

What we claim and desire to secure by Letters Patent is:

l. The process of converting hydrocarbons, comprising thermally treating raw hydrocarbon fluids to produce a gaseous mixture of aliphatic hydrocarbons containing an olefine content sufficient to insure subsequent exothermic cracking and having a temperature between 950 F. and 1500 F." suitable for exothermic conversion, then subjecting said mixture without the addition of heat to exothermic cracking at a temperature between 1259 and 1750 for an interval of time substantially as expressed by the formula T=1245 180 log t, to convert the ole- -ture between 1250 and 17 50 F.'for an interval of time substantially as expressed by the formula T=1245180 log t, to convert the olefines into heavier hydrocarbon oils, introducing a cold diluent gas into the mixture during the exothermic reaction for control-- ling such reaction, and then separating said oils so produced.

3. The process of converting hydrocarbons, comprising thermally treating raw hydrocarbon fluids to produce a gaseous mixture of aliphatic hydrocarbons containing an olefine content sufficient. to insure subsequent exothermic conversion and having a temperature between 950 F. and 1500 F. suitable for exothermic conversion, mixing a hot diluent gas with said mixture to obtain said temperature, then subjecting the combined mixturewithout the addition of heat to exothermic cracking at a temperature between 1250 and 1750 F., for an interval of time mixture without the addition of heat through an exothermic cracking zone at a temperature between 1250 and l750- F; for an interval of time substantially as expressed by the formula T=1245180log ,t, to convertthe olefines into heavier hydrocarbon oils, introducing into the exit end of said. zone a cold diluent gas for controlling said final temperature, and then separating said oils so produced.

' FREDERICK E. FREY;

JESSE A. GUYER.

GEORGE G. OBERFELL.

Images