Under these assumptions, the actual magnitude of the supersaturation may be eliminated from the analysis and the size distribution related to the growth and nucleation kinetics. T h e important consideration in size distribution consideration is the relationship between nucleation and growth rates, not the absolute rates. In this analysis, the crystallizer was assumed to be perfectly mixed. This condition can be closely approximated in small laboratory equipment, but in large production units mixing conditions are usually considerably removed from the ideal. Consequently, high local supersaturations may exist in the neighborhood of the feed points, giving rise to abnormally high nucleation rates. Under such conditions, the above kinetic model for nucleation rate may give a nucleation rate lower than that observed. These preliminary results show that the crystallization of calcium sulfate as hemihydrate has several advantages. The most important one is the fact that the hemihydrate has a much higher growth rate and a much lower relative nucleation rate than does gypsum. The combined effect of these desirable properties is that a larger average crystal size can be obtained in a shorter holding time. I n addition, a more concentrated acid can be produced, which reduces the concentration required to produce product acid. A more suitable habit is formed and is less prone to modification by impurities or concentration effects. I n the work discussed here, the hemihydrate was in such a form that it was easily filtered and washed free of phosphate. Hydration of the hemihydrate in the filter was not a problem, although in a commercial application this may well be troublesome. Apparently, because of the crystallization advantage of producing hemihydrate, a process producing this crystal product would have considerable advantage in the wet acid industry.
Nomenclature
proportionality constant, temperature dependent proportionality constant for growth proportionality constant, homogeneous nucleation proportionality constant, heterogeneous nucleation proportionality constant, relating growth and suspension density proportionality constant, relating nucleation population density and suspension density crystal size suspension density, mass of crystals per unit volume population density number of crystals linear growth rate supersaturation time
SUPERSCRIPTS i
=
j
= order of suspension density effect = nuclei
o
order of nucleation
literature Cited
Bransom, S. H., Dunning, W.J., Millard, B., Discussions Faraday SOC.5 , 83-95 (1949). McCabe, W. L., Ind. Eng. Chem. 21, 112-19 (1929). Murray, D. C., “Size Distribution Dynamics in Continuous Crystallization,” unpublished Ph.D. thesis, Iowa State University of Science anh Technology, Ames, Iowa, 1964. Randolph, A. D., A.I.Ch.E. J . 11, 424-30 (1965). Randolph, A. D., Larson, M. A,, A.1.Ch.E. J . 8 , 639-45 (1962). Timm, D. C., “Crystal Size Distribution Dynamics,” unpublished Ph.D. thesis, Iowa State University of Science and Technology, Ames, Iowa, 1965. Wolff, P. R., “Suspension Density Transients in a Mixed Suspension Crystallizer,” unpublished M.S. thesis, Iowa State University of Science and Technology, Ames, Iowa, 1965 RECEIVED for review February 20, 1967 ACCEPTED September 1, 1967 152nd Meeting, ACS, New York, N. Y., September 1966.
HYDROGEN CYANIDE PRODUCED FROM COAL AND AMMONIA G. E. J O H N S O N , W. A. D E C K E R , A. J . F O R N E Y , A N D J . H. F I E L D
Pittsburgh Coal Research Center, U . S.Bureau of Mines, Pittsburgh, Pa. 15213
cyanide has been one of the country’s strongest U . S. production increased from 174,000,000 in 1960 to 350,000,000 pounds in 1964, a 100% increase over the 4-year period. The growth of production of hydrogen cyanide has been directly related to the expansion in production of synthetic textiles from acrylonitrile. Relative growth and production of hydrogen cyanide and acrylonitrile are as follows: YDROGEN
51 growth petrochemicals in recent years.
Production of Hydrogen Cyanide and Acrylonitrile Hydrogen Acrylonitrile, Cyanide,a Million Lb. Million Lb. Year 22!Ib 174 1960 250b 21 1 1961 360b 266 1962 455b 293 1963 593b 350 1964 371 (6 months)C 1965 a U . S. Department of Commerce ( 1 9 6 6 ) . b U. S. T a r t f f Commission (1961, 7962, 7963, 1964, 1965). c U . S. Business and Defense Services Administration ( 1 9 6 5 ) .
About 50% of the total output of hydrogen cyanide goes into the production of acrylonitrile; most of the remainder is used in production of adiponitrile and methyl methacrylate (Fugate, 1962). However, in recent years acrylonitrile and adiponitrile are being produced by processes which generate hydrogen cyanide as a by-product (Chemical W e e k , 1966). ‘The bulk of acrylonitrile is used in production of acrylic fiber (Orlon, Acrilan, Dynel, Zefran, etc.) ; a smaller amount in production of nitrile rubber, the adiponitrile in manufacture of nylon. T h e manufacture of sodium cyanide utilizes about 7% of hydrogen cyanide production. The remainder goes to a large number of relatively small uses, including ferrocyanides, acrylates, ethyl lactate, lactic acid, chelating agents, optical laundry bleaches, and pharmaceuticals. The Andrussow process, the major commercial process used to make hydrogen cyanide, involves the reaction of methane, ammonia, and air over a platinum catalyst a t 1000° to 1200’ C. (Sherwood, 1959). The platinum catalyst is usually alloyed with rhodium (10 to 20%). Conversion by the Andrussow process in a single pass is limited to about 69y0 of the ammonia (about 75% with gas VOL. 7
NO. 1
JANUARY
1968
137
Hydrogen cyanide has been produced by reaction of powdered coal with ammonia at 1250" C. Yields as high as 0.7 cu. foot per cu. foot of NH, consumed were obtained. The resulting ammonia conversion of about 75 volume approximates conversions obtained commercially in processes utilizing natural gas with platinum catalysts. Coals of higher volatile matter contents gave the best yields of hydrogen cyanide. Coal gases of higher methane contents when reacting with ammonia gave the higher yields of hydrogen cyanide. Cost studies indicate that hydrogen cyanide can b e profitably coproduced with carbon black from coal and ammonia in a 40,000,000-pound-per-year capacity plant at the current market price of hydrogen cyanide ( 1 1.5 cents per pound), if credit is taken for carbon black and char by-products, These figures are based on heating the reactor electrically. If a cheaper heating method were devised, the economics of the process would b e more favorable.
70
recycle) and 53% of the methane. A typical analysis of the reaction gases leaving a catalytic reactor is (volume per cent) : nitrogen 56.3, water vapor 23.0, hydrogen 7.5, hydrogen cyanide 6.0, carbon monoxide 4.4, ammonia 2.0, methane 0.5, carbon dioxide 0.2, and oxygen 0.1. The catalytic Degussa process is similar to the Bureau of Mines method, in that heat is provided externally to ceramic reactors. The process is not in general use, but the off-gas from the ammonia-methane reaction contains more than 20% hydrogen cyanide. Utilization of methane and ammonia is reported as 91 and 8570, respectively. In its search for new uses for coal the Bureau of Mines has been investigating the production of hydrogen cyanide from coal. Although hydrogen cyanide is present in coke oven gases, and at one time was recovered as a by-product, this source of the gas has not been commonly used in the United States since the development of the newer methane-ammonia processes. Although the production of hydrogen cyanide from coal is technically feasible, production in yields that could be competitive is a problem of major concern. Equipment and Procedure
Figure 1 is a flowsheet of the experimental unit. Predried coal ground to -325-mesh is dropped in free fall through a i,'oa
heated reaction zone in the presence of ammonia at rates up to 1.10 pounds per hour. A revolving-disk feeder especially designed by the Bureau of h h e s to feed coal at low rates was constructed; it delivered to within +5% of the desired feed rate. A special coal feed system is used to prevent agglomeration and possible plugging of the reactor by heating the coal rapidly through its plastic range (about 400" C.). The reactor itself is a 4-foot length of vitreous refractory mullite, I1/8-inch i.d. and l3/~-incho.d., jacketed with two electrical resistance heaters. The top heater (maximum temperature 850" C.), 12 inches long and wound with Nichrome wire, serves as a preheater for the coal and gas. A Kanthal heater (AI-Cr-Co-Fe alloy, maximum temperature 1250" C.) encloses the center 20 inches of the tube, or the reaction zone. The bottom section of the reactor is exposed to the atmosphere for rapid cooling of the product gases. The bottom of the reactor tube fits into a 4-liter side-arm flask or char receiver in which the heavier solids are collected. The fine solids and carbon black produced are collected in an electrostatic precipitator. After the product gases leave the precipitator they pass through a cooler; then they are either metered or sent through absorbers to remove the hydrogen cyanide for analysis. All of the piping and vessels are stainless steel or glass, because of the corrosive nature of the gases. Since the gases are toxic, the unit is completely enclosed, and the enclosure is well ventilated to prevent accumulation of escaped gases. The maximum allowable concentration of hydrogen cyanide in air
I hopper - f e ed e f
(0-350g / h r )
A
,Thermocouple
Insulated e l e c t r i c a l resistance heaters
ctor(ceromic tubel Purge
Gas meter
Figure 1. 138
Flowsheet of hydrogen cyanide unit
l&EC PROCESS DESIGN A N D DEVELOPMENT
is 10 to 20 p.p.m.; its explosive limits are 5.6 to 40% by volume in air (Sax, 1951). T h e whole structure (6 X 6 X 15 feet high) is covered with steel sheeting. I t has an exhaust blower (400 cu. feet per minute) on the roof and access doors at both ground and 8-foot levels. Figure 2 shows the exterior of the unit and Figure 3 shows its interior. Product gas can be recycled to the top of the reactor adjacent to the cooled feed tube along with part of the feed gas. This flushes away any tar vapors which might adhere to the walls and cause plugging. T h e remainder of the feed gas (0 to 10 scfh) enters the reactor with the coal. Ammonia, helium, methane, nitrogen, or air, or mixtures of these gases fed from cylinders have been used as feed gas. Before startup, the system is purged with inert gas. After the reactor has been heated to 1250' C. the desired flows of coal and gas are started. Ammonia and nitrogen or helium are the gases usually used. T h e gas flow is generally split, part entering the top of the reactor adjacent to the cooled feed tube, and the remainder entering with the coal. The dry powdered coal is fed through a steam-jacketed tube (S/ls-inch 0.d.) which extends into the preheat zone of the reactor. The coal leaves the end of the feed tube which is at the temperature of the steam to enter the preheat zone of 850' C. The temperature of the coal rises very suddenly to 850' C. because of the high heat-transfer rate to the small particles. T h e carrier gas (usually helium, a n inert gas) fed with the coal keeps the partides in motion and helps prevent agglomeration as the coal rapidly passes through its plastic range. Proximate and ultimate analyses (American Society for Testing Materials, 1966; Fieldner and Selvig, 1951) are made of the char and heavier solids collected in the char receiver and of the lighter solids collected by the electrostatic precipitator. Mass spectrometric and chromatographic analyses are made on spot samples of the product gas. For cyanide determinations, metered amounts of product gas are bubbled through two scrubbers in series containing solutions of sodium hydroxide. Titration with silver nitrate solution determines the total cyanide present
Experimental Results and Discussion T h e initial tests were made with a metallic reactor tube, b u t because of low yields of hydrogen cyanide and failure of the metal at the temperatures employed, the metal tube was replaced by a ceramic reactor tube. In all the tests of this report with high volatile A bituminous (hvab) coal, Pittsburgh seam coal from Bruceton, Pa., was used. Its ultimate analysis is as follows (in per cent): carbon 75.6, ash 8.4, oxygen 8.0, hydrogen 5.1, nitrogen 1.6, and sulfur
1.3.
Figure 2.
Enclosure surrounding hydrogen cyanide unit
The effect of varying the coal-ammonia feed rates is illustrated in Table I. Hydrogen cyanide yields were computed from the wet-chemical method of analysis which is considered the more reliable method, since it was determined from proportionated gas samples taken continuously throughout the test (sample volume of 0.2 to 2 cu. feet). Only spot gas samples (sample volume 0.1 cu. foot) were used for the chromatograph and mass spectrograph analyses. In test C-241 the coal-feed rate was 0.37 pound per hour and the ammonia-feed rate was 1 cu. foot per hour. (All process gas volumes reported are corrected to standard conditions of 0' C . and 760-mm. mercury pressure.) A hydrogen cyanide yield of 0.6 cu. foot per cu. foot of ammonia reacted was ohtained, corresponding to about 12% of hydrogen cyanide in the product gas. In test C-243 the hvab coal-feed rate was reduced to 0.19 pound per hour, while the ammonia-feed rate was increased to 2.3 cu. feet per hour. A yield of 0.4 cu. foot of hydrogen cyanide per cubic foot of ammonia consumed was obtained, corresponding to about 13% of hydrogen cyanide in the product gas. VOL 7
NO. 1
JANUARY 1 9 6 8
139
Identifcation of Coal Hvab, Pittsburgh seam
Source of Coal Bruceton, Pa.
Activated carbon, Grade SXWC Anthracite
Union Carbide and Carbon Corp. Anthracite Research Center, Schuylkill Haven, Pa. Erie, Colo.
38.5
SteDheson. W. Va.
17.5
The chemical nature of the volatile matter and the oxygen content of the coal may also affect the hydrogen cyanide yield. The ratio of H2 to C O in the off-gases varied from 2.3 to 1 for subb coal and 2.4 to 1 for lignite to 7 to 10 to 1 for hvab. The higher carbon monoxide values obtained with subb coal and lignite are due to the higher oxygen contents of these coals, being 17.1 and 20.3%, respectively, compared with 8.0% for the hvab coal. Tests with Air in Feed Gas. Tests were made to determine the effect of oxygen in the treating gas on the yield of hydrogen cyanide. I t was thought that the heat for raising the temperature of the reactants to reaction temperature could be supplied by direct contact with a hot flue gas containing oxygen instead of by electric heating (Table 111). I n test C-125, a maximum yield of hydrogen cyanide was produced for tests with air in the treating gas-0.12 cu. foot of hydrogen cyanide per cubic foot of ammonia consumed, or 9.2% hydrogen cyanide in the product gas. When more than 0.5 cu. foot per hour air was fed into the reactor, moisture condensed on the walls of the solids-collection flask; the yield of hydrogen cyanide decreased. The moisture could have been absorbing the hydrogen cyanide, since hydrogen cyanide is highly soluble in water. This approach was abandoned. Tests with Methane and Ammonia. Tests were made without coal feed but with ammonia and methane to determine the resulting hydrogen cyanide yields for comparison. I n one series of tests 2.0 cu. feet per hour of methane reacted with ammonia in flows varying from 0.3 to 2.5 cu. feet per hour at 1250' C. As illustrated in Figure 4, yields of 0.18 to 0.61 cu. foot of hydrogen cyanide per cubic foot of ammonia consumed were obtained, equivalent to 1.2 to 13.8% of hydrogen cyanide in the product gas. The yield of hydrogen cyanide reached a maximum a t a feed ratio of methane to ammonia of about 1 to 1. A series of tests was made in which the methane and ammonia flow rates were maintained a t 2 feet and 1 cu. foot per hour, respectively, while the reactor temperature was increased from 1000' to 1275' C. Hydrogen cyanide yields are plotted with temperature in Figure 5, indicating increased yields with increased temperature. Maximum yields of about 0.6 cu. foot of hydrogen cyanide per cubic foot of ammonia consumed were obtained a t the higher temperatures, while at 1000' C. only 0.09 cu. foot of hydrogen cyanide per cubic foot of ammonia consumed was formed. T o explore the use of coal-derived gases in the formation of hydrogen cyanide from ammonia, synthetic mixtures of a low-temperature carbonization gas, a coke oven gas, and a producer gas were prepared and reacted with ammonia a t 1250' C. Gas flows were adjusted to give a minimum methaneammonia ratio of 1 to 1. I n Figure 6 the hydrogen cyanide yields obtained are plotted against the methane contents of the coal gases. Hydrogen cyanide yields increased with increasing methane contents of the feed gas.
Beulah, N. Dak.
41.1
Economic Evaluation
32.6
The Bureau of Mines Process Evaluation Group, Morgantown, W. Va., made a preliminary cost study of an integrated plant to produce hydrogen cyanide by reaction of ammonia with coal, based on experimental results including a yield of 0.6 cu. foot of hydrogen cyanide per cubic foot of ammonia. Electrical heating was assumed as in the bench-scale tests; a plant capacity of 40,000,000 pounds per year was chosen. The total estimated capital investment was $12,930,000, including costs for power generation. Based on a coal cost of $4.00 per ton and an ammonia cost of $100 per ton, the operating costs before profit and taxes
The yield of hydrogen cyanide per pound of coal was approximately doubled when the coal-feed rate was halved and the ammonia-feed rate doubled (tests C-241 to 243), while the hydrogen cyanide yield per cubic foot of ammonia consumed decreased by one third. I n test C-239 (Table I) the coal-feed rate was increased to 0.73 pound per hour, while the ammonia flow was maintained a t 1 cu. foot per hour. A hydrogen cyanide yield of about 0.8 cu. foot per foot of ammonia reacted was obtained, equivalent to about 13% of hydrogen cyanide in the product gas. In the next listed test, (2-198, the coal-feed rate was 0.37 to 0.40 pound per hour, while the ammonia flow was only 0.2 cu. foot per hour. The hydrogen cyanide content in the product gas was only 5% (Table I) but about 100% of the ammonia was converted with 1 cu. foot of hydrogen cyanide formed per cubic foot of ammonia used. This yield of hydrogen cyanide approximates the stoichiometric yield according to the following reactions:
+ NHa+ CHk + NHs C
+ Hz H C N + 3H2
HCN
+
In typical commercial units using catalysts for the methaneammonia reaction, the ammonia conversion attained is about 75%. Our yield values include the slight amount of hydrogen cyanide that may be formed when coal is heated to high temperatures without adding ammonia. In general, the tests of Table I indicate that an excess coal feed is desirable in order to attain maximum utilization of the ammonia, since the ammonia is by far the more expensive raw material. Tests with Coals of Different Ranks. I n addition to the hvab coal, lignite, subbituminous, low-volatile bituminous, and anthracite coals, coal char, and activated carbon were tested as raw materials for producing hydrogen cyanide. I t is thought that the volatile matter in coal reacts with the ammonia to form hydrogen cyanide; therefore coals with higher volatile content should produce more hydrogen cyanide. The volatile-matter contents on a moisture-free basis of the various materials tested are as follows:
Volatile Matter, Test (2-123, 124, 125, 198, 239, 241, 243 C-207
c-210 C-246 (2-249 C-252
Subb, Laramie seam Lvb. Pocahontas ~ b 3 .seam Lignite, unnamed seam
Pretreated hvab,a Bruceton, Pa. Pittsburgh seam Treated with air at ,200' C.
C-254 a
% 34.0
2.0 7.6
Table I1 shows the results of these tests. Lignite with 41% volatile matter produced the most hydrogen cyanide, 0.4 cu. foot per cu. foot of ammonia consumed; activated carbon, containing the least volatile matter (2.0%), produced the least hydrogen cyanide, 0.007 cu. foot per cu. foot of ammonia consumed. 140
l h E C PROCESS D E S I G N A N D DEVELOPMENT
Table 1.
Test (2-241 C-243 C-239 C-198
HCNa 11.8 13.1 13.2 5.4 ~~
~
HzS
Data from Tests with Hvab Coal and Ammonia with Helium at 1250' C.
iVHa
0~.~ .0
0~.~ .0
0.0 0.1 0.3
17.5 0.5 0.0
HP 77.7 6816 75.9 78.3
0 2
0.1 019 0.1 0.3 Feed
Product Gas, 7 0 Nz co 6.7 9.6 618 4.0 4.4 10.6 3.5 11.1
COt 0.3
5 ..~ 1
0.5 0.3
1.0 6.9 5.7
O.i
CHI ~
CZHI CaHs 0.4 0.1 0:0 0.0 0.6 0.1 0.5 0.0 Yield. Cu. Ft.
CsHio 0.0 i :i
0.3 0.0
Total Coal Length HCNlcu. HCN~CU. off-gas f t . NHa HCNIlb. ft. IVH~ He-free, of run, He, "a, feed, coal feed consumed min. cu. ft./hr. 1b.hr. cu. ft./hr. cu. f t /hr 0.60 (2-241 1 .oo 1.59 0.60 1 .oo 0.37 5.06 30 0.40 3.31 0.27 (2-243 0.49 2.35 0.19 4.90 30 1.01 0.74 0.76 5.59 15 (2-239 0.98 1 .oo 0.73 15 C-198 2.12 0.20 0.37-0.40 0.51 0.98 0.98 3.61 Wet-analytical method of analysis for H C N only; other components are average of 2 chromatograph and 2 mass spectrometer analyses, HCN-free basis.
. .
a
Table II. Data from Tests with Various Coals and Ammonia with Helium at 1250' C. Product Gas, yo Test HCNa HzS NHa Hz 0 2 N2 co COz CHI CZHI CaHs 0.0 4.9 28.0 0.5 2.4 0.0 0.0 C-246 subb 5.4 0.0 64.2 0.0 C-249 lvb 0.1 14.2 3.6 3.0 0.2 Tr 0.0 78.9 0.0 4.2 0.0 6.1 26.3 C-252 lignite 6.4 0.0 64.6 0.1 0.5 2.4 Tr. 0.0 0.0 C-254 pretreated, hvab 3.8 0.8 10.6 14.3 0.1 0.0 72.3 0.0 1.9 0.0 0.0 0.3 21.4 2.9 0.4 1.5 C-210 anthracite 0.2 0.0 73.4 0.0 0.0 0.0 0.0 23.0 0.1 C-207 activated carbon 0.25 0.0 70.0 6.6 0.3 0.0 0.0 0.0 Feed Total off-gas Length of He-f ree, He, "3, cu. f ./hr. cu. ft./hr. Solids feed, lb./hr. cu. ft./hr. run, min. 0.42 (2-246 subb 0.91 1 .oo 6.88 30 '2-249 Ivb 0.47 0.97 1 .oo 3.92 30 C-252 lignite 1.06 1 .oo 0.37-0.40 6.38 15 C-254 pretreated, hvab 1.04 1 .oo 0.35 5.80 30 C-210 anthracite 1.02 1.15 0.37-0.40 4.15 15 C-207 activated carbon 1.06 1.15 0.62 3.23 15 Yield, Cu. Ft. HCNICU.f t . NHa HCN/cu. f t . NHa H C N / l b . coal feed consumed C-246 subb 0.88 0.37 0.37 C-249 lvb 0.34 0.16 0.16 0.41 C-252 lignite 1 .05 0.41 C-254 pretreated, hvab 0.22 0.63 0.22 C-210 anthracite 0,007 0.021 0.007 C-207 activated carbon 0.013'~ 0.007 0,007 a Wet-analytical method of analysis for H C N onb; other components are average of 2 chromatograjh and 2 mass spectrometer analyses, HCN-free basil. b Per pound of carbon.
.
Table 111.
Product Gas Analyses and Yields of Hydrogen Cyanide from Tests with Hvab Coal, Ammonia, and Air at 1250' C. Product Gas, 7, Length Feed of Run, NHa , Air Coal, CHI h'Ha H2 Nz co Min. cu. ft./hr. cu. ft./hr. Ib./hr. Test HCNa C-123 10.3 1.1 10.2 62.8 16.2 9.7 20 5.90 0.52 0.17-0.20 C-124 10.9 0.6 6.9 62.1 17.2 13.2 20 2.96 0.53 0.17-0.20 (2-125 9.2 9.6 48.4 31.5 0.4 10.1 20 3.08 1.08 0.17-0.20 Total Yield, Cu. Ft. Off-Gas, HCIN/CU.f t , H C N ~ C fUt .. NHz Cu. Ft./Hr. HCN/l b . coal NH3 feed consumed (2-123 4.18 2.30 0.073 0.078 C-124 3.05 1.77 0.112 0.120 C-125 3.70 1.82 0.110 0.123 a Wet-analytical method of analysis for H C N only; other components are average of 2 chromatograph and 2 mass spectrometer analyses, HCN-free basis.
would be 5.82 cents per pound of hydrogen cyanide product, allowing by-product credit. Addition of 12% gross return on investment would give production costs of 9.7 cents per pound of product when $4.00 per ton coal is used. The current market price of hydrogen cyanide is 11.5 cents per pound (Oil, Paint and Drug Reporter, 1966).
Credit has been allowed in the cost figures for a 7.6y0 yield of carbon black and the excess char produced in the process. Some of the char and the scrubbed product gas (containing about 7501, of hydrogen) are consumed in the steam plant for power generation. Electrical heating, which was used in the test unit and in the cost figures, is one of the most expensive VOL. 7
NO. 1
JANUARY 1 9 6 8
141
06
/
5-
I
d
w
d
z I U
o 1 05
10
15
20
25
20
Figure 4. Effect of ammonia flow rate on yield of hydrogen cyanide from methaneammonia reaction a t 1250' C.
40
60
80
100
CH, IN F E E D G A S , percent
NH3 FLOW,scfh
Figure 6. Variance of hydrogen cyanide yield with methane content of feed gas in reaction with ammonia at 1250' C.
Methane flow 2 SCFH
types of heating, accounting for more than 40% of the capital costs in the estimate. If cheaper conventional heating could be used, production costs would be lowered considerably. Conclusions
Figure 5. Effect of temperature on yield of hydrogen cyanide from reaction of methane (2 SCFH) and ammonia (1 SCFH) 142
I&EC PROCESS DESIGN A N D DEVELOPMENT
Hydrogen cyanide has been produced from coal and a m monia at 1250' C. in bench-scale studies. A metal reactor could not be used because the metal failed at the temperatures required, and the yield of hydrogen cyanide was low. The yield was improved greatly when a refractory ceramic reactor was used. Hydrogen cyanide yields approximating stoichiometric of 1 cu. foot of hydrogen cyanide per cubic foot of ammonia reacted were obtained at low flows of ammonia. At higher ammonia flows, ammonia conversion of about 75% was obtained, which is the usual conversion attained in commercial units using natural gas and a platinum catalyst. The low-volatile coals gave low yields of hydrogen cyanide; the high-volatile coals gave the best yields. The results indicated that the hydrogen cyanide is produced by reaction of ammonia with the hydrocarbons in the coal. Yields of hydrogen cyanide from reaction of ammonia with gas mixtures containing methane are directly related to the methane content of the gas. Cost studies indicate that hydrogen cyanide can be produced from coal and ammonia at a price approximating the posted sales price.
literature Cited American Public Health Association, New York, “Standard Methods for the Examination of Water and Wastewater,” 11th ed., pp. 355-6, 1960. American Society for Testing Materials, “A.S.T.M. Standards with Related Materials.” “Gaseous Fuels. Coal and Coke.” Part 19. 1966. Chem. Week 99, No. 23, 65 (Dec. 3, 1966). Fieldner, A. C., Selvig, W. A., “Methods of Analyzing Coal and Coke,” Bur. Mines Bull. 492 (1951). Fugate, W. O., “Hydrogen Cyanide,” in “Encyclopedia of Chemical Technology,” R. E. Kirk and D. 0. Othmer, eds., 2nd ed., Vol. 6, pp. 574-85, Interscience, New York, 1962. Oil, Paint and Drug Reporter 189, No. 23, 20 (June 6, 1966).
Sax, N. I., “Handbook of Dangerous Materials,” pp. 197-8, Reinhold, New York, 1951. Sherwood, P. W., Refining Engr. 31, No. 2, C-22-26 (February 1959). U. S. Business and Defense Services .4dministration, “Chemicals,” Quarterly Industry Rept. 12, No. 4, 40 (December 1965). U. S. Dept. Commerce, “Current Industrial Reports, Inorganic Chemicals and Gases,” Ser. M-28A, 6 (Jan. 28, 1966). U. S. Tariff Commission, Pub. 34, 53 (1961); 72, 53 (1962); 114, 56 (1963); 143, 55 (1964); 167, 55 (1965). RECEIVED for review May 5, 1967 ACCEPTED September 14, 1967 Reference to trade names is made for identification only and does not imply endorsement by the Bureau of Mines.
EXPLORATORY STUDIES OF A PROCESS FOR CONVERTING OIL SHALE AND COAL T O STABLE HYDROCARBONS J A M E S E. F L I N N AND G E O R G E F. S A C H S E L Chemical Process Development Division, Columbus Laboratories, Battelle Memorial Institute, Columbus, Ohio
43207
Oil shale and coal are retorted in a flowing hydrogen stream at pressures of 150 to 450 p.s.i.a. The resulting hydrogen-hydrocarbon vapor stream is passed directly through a cobalt-molybdenum catalyst bed for removal of sulfur, oxygen, and nitrogen impurities and for olefinic bond saturation. The condensed vapors yield a clear, stable, low viscosity, liquid hydrocarbon. Volumetric yields of liquid product in excess of the Fischer assay value appear possible from the process. This concept of retorting differs from past approaches in that the direct production of a gasoline product from coal or oil shale is the goal.
shale and coal have long been recognized as potential sources of the hydrocarbon products which are now produced almost exclusively from crude petroleum and natural gas. However, an economic means of extracting hydrocarbons of value from these solids has yet to be accomplished commercially in the United States. With regard to oil shale, only a few of the numerous ex situ processing schemes, which have been proposed over the years, have received widespread attention. Among these are the Bureau of Mines gas-combustion retort process, the Union Oil Co. down-draft retorting process, and the Oil Shale Corp. rotating-kiln (Tosco) process (Prien, 1964). Until recently (Chemical and Engineering News, 1966; Chopey, 1965), it appeared that the Tosco process was closest to achieving commercialization. These three processes, while differing in method, seemed to have one objective in common-viz., the recovery of a hydrocarbon liquid from the oil shale? which, after a minimum of processing, could be utilized directly as feed to a conventional oil refinery as though it were a crude petroleum. T h e situation with regard to coal is similar. The more advanced projects for producing liquids from coal currently being sponsored by the Office of Coal Research all have as their objective the production of a synthetic crude which can be processed directly in existing oil refineries (Chemical Week, 1967; Fumich, 1966). These projects include Project Gasoline, H-Coal Process, Project COED, and Project Seacoke, developed by the Consolidation Coal Co., Hydrocarbon Research, Inc., FMC Corp., and Atlantic Refining Co., respectively. IL
This paper presents the results of some exploratory experiments on a process (Huntington, 1963a, 1963b, 1966) which departs somewhat from the above-mentioned objective. Applicable in principle to both coal and oil shale, this process was designed to produce in a single step a low boiling range liquid hydrocarbon product substantially free of sulfur, nitrogen, and oxygen compounds. When applied to coal, it was believed that in addition to a liquid fuel, a char product low in organic sulfur content could be produced, thus adding to the economics of the process. Sufficient exploratory data have been obtained to confirm that a liquid hydrocarbon mixture resembling gasoline can be produced from either coal or oil shale by this process. Since most of the experiments to date have been with oil shale, the remainder of this paper deals principally with this substance, while referring to only a few experiments with coal. T h e recovery of oil from oil shale by conventional means involves: mining of the oil shale, obtaining the shale oil by a retorting process, upgrading the shale oil to a petroleum-like product, and converting the petroleum product to valuable hydrocarbon liquids such as gasoline, gasoline-blending stock, etc. The mining step itself is a major economic factor. One estimate is that mining costs represent 60% of the over-all processing cost. This explains the current interest in in situ recovery methods (Chemical and Engineering News, 1967). While mining of the shale is a major problem, retorting methods have received the major research emphasis to date. T h e three processes mentioned for retorting oil shale involve the destructive pyrolysis of the shale in an open or closed vessel a t near atmospheric pressure. T h e oil vapors obtained in this VOL. 7
NO. 1
JANUARY 1968
143
