INDUSTRIAL AKD ENGINEERING CHEhlISTRY
March, 1934
315
TI. Semi-Industrial Production of Aromatic Hydrocarbons from Natural Gas in Persia W. H. CADMAN,Anglo-Persian Oil Company, Ltd., London, England The work described in Part 11 deals with the plant has yet to be found. It is pointed out that pyrolysis of natural gas at the oil fields in Persia the chemist in ihisjield is ahead of the metallurgist. Attention has to be paid to the catalytic action of o n a semi-commercial scale for the production of liquid aromatic hydrocarbons. It toas done simul- the material o n the gases, which influences both taneously with the more academic research on the benzene formation and carbon deposition in the thermal decomposition of pure hydrocarbons de- reaction tubes. The importance of turbulent flow is scribed in Part I . The whole of this work was emphasized. The reaction products identified were: benzene, planned primarily in connection with the utilization of natural gas in Persia. The yields of benzene toluene, xylene, styrene, indene, naphthalene, anunder varying conditions of temperature and pres- thracene, phenanthrene, chrysene, butadiene, and sure were found and optimum conditions deter- isoprene. The gases remaining after benzeneproduction contained about 20 per cent of ethylene, which mined. The method of desulfurizing the gas is described, is a potential source of further liquid hydrocarbons. Since the experimenls described in this paper were and the carious furnaces and retort materials which completed our knowledge of the art of pyrolyzing were tested in the experiments are reciewed. Special hydrocarbon gases has advanced considerably, and heat-resisting alloy steel tubes in a radiant heat further development work on a semi-commercial type of furnace were found to stand u p best, but the ideal long-life metal for a n industrial pyrolysis scale is now in progress. ROM the industrial viewpoint the production of liquid aromatic hydrocarbons from gaseous paraffins and olefins is of considerable importance owing to the vast quantities of these hydrocarbons available for development. In 1929 the quantities of paraffin hydrocarbons available, expressed in niillions of cubic feet, have been stated to be: methane, 2,015,000; ethane, 336,150; propane, 111,415; and butane, 64,980 (in niillions of cubic meters: methane, 57,059; ethane, 9516; propane 3127; and butane 1840). The potential production of aromatic gasoline from these gases, methane excepted, is 2,950,000 tons per year or 8100 tons per day. In Persia the natural gases are designated as follows: High-pressure gas is that part taken from the highpressure separators operating under the natural pressure of the (1)
oil. (2) Low-pressure gas is Feparated from the crude oil when
the latter is reduced to approximately atmospheric pressure. This separation is made in flow tanks at 3 to 6 inches of water pressure. (a) Unstripped gas is low-pressure gas which has not had its gasoline removed by absorption or compression. (b) Stripped gas is low-pressure gas from which gasoline has been removed. (3) Accumulator gas is obtained during the removal of gasoline and consists mainly of propane and butane. These gases vary in composition with operating conditions in the oil field and seasonal changes, but approximate analyses made on a Bone and Wheeler apparatus, and subsequently confirmed by a Podbielniak apparatus, when the work t o be described was in progress are: HIGHPREISCRE GAS
HsS Methane Ethane Propane Butane Pentme+
UNSTRIPPED STRIPPED LowLOTPRESSCRE
PRE5SCRE
. .*CCCMCL4TOR
GAS
Gas
%
74
%
R
4.0
12 29
12
40
11.5 2.5
10
21 18 9
4
Trace
76
18
1 1
24 21
G ' "IS were: EXITGh0 (NpFree)
Butadiene Unaccounted for
0.43
0.10
INLET Gaa (HIS- and Na-Free) 45.0 2.4 1.27 1.0s 0.28
Here, after benzene production, is a gas containing 20 per cent olefins, essentially ethylene, representing a 50 per cent volume conversion on the ingoing gas. This exit gas is a potential source of further liquid hydrocarbons or alternatively other products such as alcohols, ketones, etc. CARBON The deposition of some carbon appears to be inevitable in pyrolysis operations for the production of benzene. Certain materials inhibit the production of carbon and, when used under suitable conditions, carbon deposition in the reaction plant may become almost negligible. Cleaning, when
INDUSTRIAL AND ENGINEERING CHEMISTRY
320
necessary, may be accomplished by air-blowing afher the hydrocarbon atmosphere has been replaced by an inert one. REFININQOF BENZENE Benzene derived from the pyrolysis of natural gas appears to differ from coal-tar benzene only in the relative proportions of its constituents, and the same general methods of refining are applicable in both cases. Where the product of pyrolysis is intended as a blending stock, the following are some of the methods of treatment available: inhibitors, sulfuric acid treatment, hydrogenation, fractionation and the separate treatment of the fraction, and vapor-phase refining. Bound up with this question of refining is also the maximum permissible sulfur content of the benzene. The normal product using metal tubes contained 0.5 per cent sulfiir, and this, for blending purposes, was considered satisfactory. Pyrolysis benzene is a very suitable material for the testing of inhibitors, and considerable success can be expected by this method. Sulfuric acid refining has proved satisfactory, in that the product, after treatment, is stable and of good quality, but this method is wasteful of \That may be useful unsaturated material. Details need not be given-as these follow the normal refining lines. Hydrogenation may be applied to the benzene ns a mhole, and this has been tried with some success, but attention was concentrated on the hydrogenation of such highly unsaturated fractions of the benzene as the styrene fraction. The purpose of this was to produce extremely useful antiknock materials like ethyl benzene and so avoid loss of unsaturated hydrocarbons in refining. The more saturated parts of the benzene might then be refined using sulfuric a d . Vapor-phase refining methods, using such substances as zinc chloride, did not prove to be highly successful in the case of pyrolysis benzene.
Vol. 26, No. 3
INFLUENCE OF PRESSURE ON PYROLYSIS OF HYDROCARBONS LARGE-SCALE PYROLYSIS OF NATURAL GAS. On the semitechnical scale, tests were carried out using metal tubes in the furnace section a t a pressure of 30 pounds per square inch gage. This, unfortunately, was a limiting pressure owing to the construction of the plant. The heating system consisted of 12 X 2 inch i. d. tubes in series, of which the first eight were ordinary 2-inch steel tubes and corresponding return bends, while the four hottest tubes u-ere of H. R. 4. steel with return bends of the same material. A lagged expansion box was used. Operation a t 30 pounds per square inch was compared with operation a t 6 pounds per square inch using stripped gas of specific gravity 1.00 to 1.01. The throughput was increased from 1800 to 2800 cubic feet per hour a t the higher pressure. The benzene production was little changed and was of the order of 1.1 gallon per 1000 cubic feet of nitrogen- and hydrogen sulfide-free gas. Under 30 pounds per square inch pressure the plant was easier to run than a t the lower pressure, heat transfer wasimproved, throughput increased, and yield remained the same.
ACKNOWLEDGMEKT dcknomledgnient is due the Anglo-Persian Oil Company, Ltd., for making this semiworks-scale investigation possible and for granting permission to publish thp results. To R. V. Wheeler, J. F. Thorpe, J. Jameson, G. H. Coxon, A, E Dunston, and F, B. Thole appreciation and thanks are due for criticism and valuable suggestions as the work proceeded. The miter wishes to acknowledge his indebtedness to E. ?;. Hague for assistance in preparing Part I1 for publication, and also for the identification of the various hydrocarbons present in the crude benzene. RECEIVED September 7 , 1933
Thermal Decomposition of the "Coal Hydrocarbon" H. H. LOWRY,Coal Research Laboratory, Carnegie Institute of Technology, Pittsburgh, Pa.
C
OAL is not a hydrocarbon. Even if we leave out of
consideration the technically very important constituents ash, moisture, sulfur, and nitrogen, the data in Table I show that, in a high-rank bituminous coal, there may be one to nine oxygen atoms to every hundred atoms comprising the coal substance. Only in the anthracites, which are not of commercial interest from the viewpoint of use in any process of recovering by-products from thermal decomposition, do we approach a substance having the composition of a hydrocarbon. An attempt wiL1 be made later to show that a study of the thermal decomposition of anthracite may lead to a clearer understanding of the nature of the solid residue obtained in all normal destructive distillations of solid organic complexes. COMPOSITION OF COALS TABLEI. CHEMICAL (In atomic per cent")
COAL
- ---I POI
C
H
3s +3 47 1 4 66 1 8 7R . .+12
45 l t 3
0 16 1 3
12 * 2 Lignite b *4 Bituminous I n i h r a c i t r 1*1 .__.I_ ..I-..4 Calculated on the basis % C %H %0 100. For each 100 atome of C H, and 0 bituminous coal8 normally contam 1 to 2 atoms of N and 0.3 3 atoms df S.
+
tb
+
41 f4 39 i b 20112 ~
-
Free hydrocarbons exist in coal in only relatively insignificant amounts, if at all. That certain solvents may
extract hvdrocarbons from coal at temperatures from 80" to 260" C.-temperatures below active gaseous decomposition of coal-is frequently cited as evidence to the coiitrary. This argument appears to be insufficient. A substance dissolves in a solvent owing to the fact that its partial pressure in the solvent is less than its normal vapor pressure-in this respect solvent extraction is analogous to distillation. The fact that solvents extract materials from coal a t lower temperatures than are effective in distillation does not necessarily signify their prior existence as such in the coal. For distillation a vapor-pressure gradient of an entirely different order of magnitude is necessary than for solvent extraction. I n this connection consider, for example, sugar which may be distilled only a t extremely low pressures (90) but may dissolve readily in the appropriate solvent. Also, it is clear that thermal decomposition in the presence of a solvent is easier than in a vacuum (or a t atmospheric pressure). Elementar) electrostatics states that the attraction between two oppositely charged bodies is less the higher the dielectric constant of the medium in which the bodies exist. Referring specifically to coal, the presence of a solvent therefore loosens the entire solid structure in such a way that less kinetic energy in the form of heat is necessary to cause rupture of the bonds holding the solid together than in the absence of the solvent. This explanation accounts for the greater effectiveness of
