(4) Cathers. G. I., Jolley, R. L., Moncrief, E. C., X‘ucl. Sa. Eng. 13, 391 (1962). (5) Chilenskas, A. A , , Gunderson, G. E., Argonne National
Laboratory, “Engineering Development of Fluid-Bed Fluoride Volatility Processes,” Part 7, “Corrosion of Nickel in Process Environments,” ANL 6979 (1965). (6) Edlund, M. C.. Chem. Eng. Progr. Symp. Ser. No. 51, 60, 78-86 (1964). (7) Leva, M., “Fluidization,” McGraw-Hill, New York, 1959. (8) Miles, F. T., Heus, R. J., Wiswall, R. H., Brookha\.cn National Laboratory, “Extraction of Protactinium from Solid Thorium Fluoride,” BNL 482 (T-109) (1954). (9) Miller, P. D., Stephan, E. F., Stiegelmeyer, \V. N., Fink, F. W., Battelle Memorial Institute, “Corrosion during Oxidation-Fluorination Processing,” BMI-X-209 (Oct. 19, 1962). (10) Nicholson, E. L., Ferris, L. M., Roberts, J. T., Oak Ridge National Laboratory, “Burn-Leach Processes for Graphite-Base Reactor Fuels Containing Carbon-Coated Carbide or Oxide Particles,” ORNL-TM-1096 (1965). (11) Pall, D. B., Ind. Eng. Chem. 45, 1197 (1953). (12) Pashley, J. H., Kaufmar, H. L., Oak Ridge Gaseous Diffusion Plant, ORGDP K-L-1859 (classified).
(13) Reilly, J. J., Regan, \V. H., Wirsing, E., Hatch, L. P.: IND. ENG.CHEM. PROCESS DESIGN DEVELOP. 2, 127 (1963). (14) Schmidt. H. I V . , “Reaction of Fluorine with Carbon as a Means of Fluorine Disposal,” Natl. Aeronaut. Space Administration: Rept. NACA-RM-E-57E02 (1957). (15) Simons, J. H., Fluorine Chem. 1, 406-7 (1950). (16) Spence, R. \V.; Intern. Sci. Technol. 43, 58-65 (July 1965). (17) Steunenberg. K. K.: Vogel, R. C., “Reactor Handbook,” 2nd ed., Vol. 11; Chap. 6, Interscience, New York, 1961. (18) Vogel, R. C., Carr, W. H., Cathers, G. I., Fisher, J., Hatch, L. P., Horton, R. W., Jonke, A . A , , Milford, R. P.? Reilly, J. J., Strickland, G., “Fluoride Volatility Processes for the Recovery of Fissionable Material from Irradiated Reactors Fuels,” 1964 Geneva Conference (A/Conf. 28/P/250). (19) Zenz. F. A,; Othmer, D. F., “Fluidization and Fluid-Particle Systems,” Reinhold, New York. 1960. RECEIVED for review .4ugust 3. 1965 ACCEPTED October 20. 1065 Work performed under the auspices of the U. S. Atomic Energy Commission.
REACTIONS O F COAL IN A P L A S M A JET R.
D . G R A V E S , W A L T E R
K A W A , A N D
R . W .
H I T E S H U E
Bureau of Mines, U. S. Department of the Interior, 4800 Forbes Aae., Pittsburgh. Pa.
Reactions of a high-volatile A bituminous coal in argon plasmas were studied. The effects of coal feed rate, coal particle size, and plasma temperature on yields and product composition were examined. Products were a solid residue and gas containing hydrogen, methane, acetylene, diacetylene, and carbon monoxide. Yields of gaseous products increased with increased temperature and decreased coal particle size. Acetylene, the principal hydrocarbon gas produced, was obtained in yields as high as 15 weight of moistureand ash-free coal. In all experiments, plasmas contained sufficient energy to heat the coal to temperatures considerably higher than needed for complete devolatilization, but the solid residues contained 10% or more volatile matter, indicating that thermal equilibrium between the coal and plasma was not reached. More efficient utilization of the heat available in plasmas will b e required through changes in plasma generator design and/or operating techniques to obtain higher coal temperatures and higher yields of gaseous products.
70
HE Bureau of Mines has made a n exploratory study of the T r e a c t i o n of coal in a plasma jet in which plasma temperatures up to about 15,000° C. can be attained. The objective was to obtain new knowledge of coal chemistry which may lead to new methods of producing chemicals from coal. Plasma consists of more or less ionized gases or vapors, and is present in any electrical discharge. Temperatures generated in electrical discharges are high enough for molecular bonds to be broken and for the resultant atoms to dissociate into ions and electrons. The complex mixture of ions, electrons, neutral atoms, and excited molecules produced is called a plasma and is referred to as the fourth state of matter. I t conducts electricity, is a good heat conductor, and is luminous. The highly excited species that exist in plasmas can react to produce compounds whose formation is thermodynamically unfavorable a t conventional conditions. A plasma jet is formed when plasma generated from a fluid flowing through a n electrical arc confined within a chamber is made to flow out of the chamber through a nozzle or orifice. Various plasma-jet devices have been used in recent years to study high-temperature chemical synthesis. Leutner and Stokes ( 6 ) >using a consumable graphite anode, produced acetylene by reaction of the graphite with hydrogen in an
argon-hydrogen plasma. About 34% of the carbon consumed was converted to acetylene. Adding graphite in powdered form did not increase the yield of acetylene. These same investigators also made methane react in argon plasmas and converted 80y0 of the carbon in the methane to acetylene. Acetylene synthesis has also been reported by Freeman and Skrivan (5), who made methane react in argon and argonhydrogen plasmas, Baddour and Iwasyk ( 3 ) ,and Baddour and Blanchet ( Z ) , who made hydrogen react with carbon vapor obtained from a consumable graphite electrode in a high intensity arc reactor. A few experiments have been made by Ammann et al. (7) using coal in a graphite anode in another high intensity arc system; however, there have been few published data on the reactions of coal in a plasma jet. Bond et al. a t B.C.U.R.A. have done some research of this type, however, and have produced acetylene with a carbon content equivalent to 20y0 of the carbon in the coal fed using an argon plasma jet (4). The plasma-jet device used in the present study was a commercial plasma gun designed for spray-coating applications. Experiments were made with high-volatile A bituminous (hvab) coal using argon as the working gas and also as a carrier gas in which coal was entrained and fed into the plasma gun. VOL. 5
NO. 1
JANUARY 1 9 6 6
59
water entering and leaving the plasma gun are measured. Product gases are sampled after steady-state conditions are reached. ,4t the conclusion of an experiment, solid products are washed from the cooling system, receiver, filter, trap, and interconnecting lines with water. T h e washings are combined, and then the solids are filtered and dried in air a t 70' C. for 20 hours. Calculations. Total power input is determined from electrical current and voltage measurements a t the gun, whereas net power input to the plasma is calculated as the difference between total power input and the power loss to the cooling water. T h e enthalpy of the primary gas and the average plasma temperature are determined from the net power input into the plasma, the working gas flow rates, and a plot of specific enthalpy as a function of temperature. This calculation assumes no energy loss other than that to the cooling water. Yields of solids are determined from weighings of coal fed and solids recovered. Yields of gaseous products are calculated from analyses of product gas and the known coal and argon feed rates. This calculation assumes constant coal feed rate and steady-state conditions a t the time the gases were sampled.
T h e effects of coal rate, coal particle size, and plasma temperature on yields and product composition xiere examined. Experimental Procedure
Apparatus. T h e plasma-jet unit consists of a plasma gun. product-recovery system. coal feeder: power supply, cooling water facilities. and a control console. Direct current used to operate the plasma gun is supplied by t\vo three-phase alternating current transformers and selenium rectifiers. This system is rated a t 28 kw. and is capable of delivering 700 amperes a t 40 volts or 350 amperes a t 80 volts. I t also contains a high-frequency oscillator used to start a n arc. T h e electrical potential required to sustain a n argon plasma was about 20 volts. T h e plasma gun. sho1t.n schematically in Figure 1. has a 3, ],;-inch thoriated tungsten cathode and a '/,-inch-i.d. cylindrical \iater-cooled copper anode. T h e \vorking gas enters the gun radially near the cathode and flojvs do\\-n\\-ardinto thc anode. T h e gas is converted to a plasma as it flo~vsthrough the electrical discharge betxveen electrodes. Coal entrained in the carrier gas enters the plasma through a n inlet in the side of the anode. T h e reaction mixture leaves the bottom of the anode as a luminous jet and flo~vs doivn through the Lvater-jacketed cooler. which consists of a 3-inch copper tube about 12 inches long. T h e residence time of coal in the plasma is less than a millisecond. Reaction products leave the cooler a t several hundred degrees centigrade and enter the solids receiver via a dip tube. T h e receiver is 6 inches in diameter and 12 inches long and contains about 3 inches of \\.atel-. Most of the solids collect in the bottom of the receiver, while the gases and finer solid particles a t near room temperature bubble u p through the water and a tubular baffle. Gases flo\v through a cloth filter, which collects additional fines, and are sampled, metered? and vented. Operating Procedure. T h e system is assembled, purged with argon, and pressure-tested. Flows of cooling water and working gas are established, and then the electrical discharge between electrodes is started. T h e flow of carrier gas is established, and the coal feeder is started. Coal is fed into the plasma gun for 10 minutes unless a coal feed stoppage or other system failure causes a premature termination of a n experiment. Initially: pressure of the system is about 3 to 4 p.s.i.g. but gradually increases several pounds per square inch as solids accumulate on the filter. Temperatures of cooling
Results and Discussion
Reactions of 70 X 100 M e s h Coal. Experiments with 70 X 100-mesh coal (L. S. sieve) were made using feed rates of about 3.1 and 1.1 pounds per hour. Argon rates were constant at 1.17 standard cu. feet per minute (scfm) as the primary gas and 0.26 scfm as the carrier gas. The power input a t each coal rate was varied to give average plasma temperatures ranging from 3400' to 7700' C. a t about 3.1 pounds per hour and from 3900' to 8600' C. a t about 1.1 pounds per hour, Reaction products were a solid residue and a gas containing hydrogen, methane, acetylene. diacetylene. and carbon monoxide. Operating conditions, yield data, and analyses of the coal feed and residues produced are shown in Table I. Yields are expressed as weight per cent of moisture- and ash-free (maf) coal. At a coal rate of 3.1 pounds per hour, increasing the power input (or average plasma temperature) had only a
Table I. Effects of Power Input and Coal Feed Rate on Yields and Residue Compositions (1.17 scfm argon as primary gas, 0.26 scfm argon as carrier gas. 70 X 100 mesh hvab coal in argon plasmas) 4 18 9 10 14 16 Experiment number 1.20 3.11 3 06 3.14 0 99 1 11 Average coal rate, lb./hz. 3400 6300 7700 3900 6600 8600 Average plasma temp., C., 7.3 3.8 9.4 4.1 7 5 12 6 Total power input, kw. 3 2 1.7 3,9 2.0 3 3 4 7 Net power input, kw. Products. wt. % maf coa 92.2 89.2 89.0 87.7 84 0 78 3 Solid residue, maf 0.8 1 7 0.4 1.1 0.4 1 0 Hydrogen 0.2 0.2 0.3 0.2 0 1 0 2 Methane 4 3 2.6 3.5 4.5 2.2 6 0 Acetylene 0.3 0.3 0 4 ... 0.1 0 6 Diacetylene 5 6 4.1 5.3 5 6 3.8 11 0 Carbon monoxide 99.8 99.3 00.5 94.4 95 4 97 8 Total 9 6 . 8 9 6 . 8 9 7 . 1 9 1 . 3 92 8 92 3 Carbon recovery, 7 0 96.4 98.0 05.4 94.2 100 3 104 1 Hydrogen recovery, 70 Analyses, wt. % Solid residue Coal Material 1.3 0 6 0.8 0.8 0.4 0.1 0 6 Moisture 5.1 4.7 5.4 4 2 3.8 4.4 4 6 ,4sh
Ultimate composition, maf basis H C N S 0 (by difference) Volatile matter, maf basis
60
l&EC
5.6 84.3 1.7 0.9 7.5 37.6
5.2 83.3 1.7 0.9 8.9 34.8
PROCESS DESIGN A N D DEVELOPMENT
5.1 87.5 1.6 0.9 4.9 31.2
5.0 87.0 1.5 1. o 5.5 32.6
5 3 83.4 1.6 0.8 8.9 34.4
5 87 1 0 5 33
2 4 6 8 0 9
4 86 1 0 7 31
6 0 5 9 0 1
Copper a n o d e
Figure 1.
Schematic diagram of plasma generator
slight effect on the extent of coal decomposition. Yields of solid residue decreased from 92.27, a t 3400' C . to 89% a t 7700' C . Yields of hydrogen and acetylene increased almost linearly from 0.4 to 1.17' and from 2.6 to 4.5%, respectively. The yield of methane was constant a t about 0.270, while the yields of diacetylene showed no significant trend. With a coal rate of about 1.1 pounds per hour, the effect of increasing the plasma temperature on yields \vas similar. Yields of residue decreased, Lvhile yields of hydrogen increased from 0.4 to 1.77,, yields of acetylene increased from 2.2 to 6.O'%, and yields of diacetylene increased from 0.1 to 0.67, \vith increasing temperature. 'The amount of methane produced was again constant at about 0.27,. At both coal rates, yields of carbon monoxide increased with increasing temperature. T h e effect of decreasing the coal feed rate from 3.1 to 1.1 pounds per hour can be seen by comparing the results from experiment 9 with experiment 18 and experiment 10 with experiment 14. Estimated plasma temperatures in each pair of experiments \and 18 (1.1 pounds of coal per hour). However, as yields and
ultimate compositions of the residues produced in the two sets of experiments were similar a t comparable conditions, this change \cas ineffective. Assuming that better mixing was achieved u i t h the extended nozzle, it appeared that heat transfer rate rather than the degree of mixing limited the temperature rise and extent of coal decomposition. Reaction of -325-Mesh Coal. To reach higher coal temperatures, the particle size of the coal feed was decreased to -325-mesh. Experiments were made with power input as a variable, and average plasma temperatures of 4800°, 7300', and 8800' C . were obtained. Argon rates were constant a t 1.17 scfm as working gas and 0.26 scfm as carrier gas. Average coal rates were 1.03, 0.84, and 0.74 pounds per hour, although intended to be 1 .O pound per hour. Product distributions and the analyses of the coal feed and solid residues produced are shown in Table 11. As power input was increased, yields of solid residue decreased from about 74 to 45% and yields of hydrogen and acetylene increased. KO trends were observed in the yields of methane and carbon monoxide. Trace quantities of diacetylene were formed in each experiment. There was considerably more decomposition in all experiments with -325-mesh coal than with 70 X 100 mesh coal. The - 325-mesh coal evidently reached higher temperatures. Volatile matter contents of the residues from - 325-mesh coal \vere much lower, and yields of gaseous products were much higher. I n experiments with 70 X 100 mesh coal in argon plasmas, the highest yield of acetylene obtained was 6%. The solid residue produced from this experiment contained 31 weight % volatile matter on a maf basis. As sho\vn in Table 11, acetylene yields obtained from -325-mesh coal were about l o ? 12, and 157,, and the residues contained about 17, 10, and 12% volatile matter. These results clearly show that particle size had a pronounced effect on the extent of coal decomposition. The highest acetylene yield is in approxi-
Table II.
Effect of Power Input on Yields and Residue Compositions
(1.17 scfm argon as working gas, 0.26 scfm argon as carrier gas. -325-mesh hvab coal in argon plasmas) Experiment number Average coal rate, lb./hr. Average plasma temp., O C. Total power input, kw. h'et power input, kw.
38 41 40 1 03 0 84 0 74 4800 7300 8800 4 8 7 5 102 2 4 3 7 4 9
Products, wt. yc maf coal Solid residue, maf Hydrogen Methane Acetylene Diacetylene Carbon monoxide Carbon dioxide Total
Carbon recovery, 7c Hydrogen recovery, 7, Analyses, wt.
.Material Moisture
73 6 2 4 2 7 9 5 Trace 18 1 1 4 107 7
62 9 3 0 0 5 12 3 Trace 11 4
103 2 126 6
45 3 0 15
3 9 6 4
Trace
24 3
0 0
0 0
90 1
89 5
89 4 115 8
80 9 130 5
7c
Ash
Ultimate composition, maf basis H C N
S 0 (,by difference) Volatile matter, maf basis
VOL. 5
Coal 1 1 8 3
Solid residue 0 5 0 6 0 7 112 114 119
4 9 81 9 1 5 1 5 102 37 2
3 89 1 1 5
NO. 1
2 2 5 1 0 17 4
2 5
2 5
90 0
90 9
1 1 1 1 5 3 10 2
1 1 4 12
JANUARY
1966
0 1 5 3
61
mate agreement with the results obtained by Bond et al. (4) in a similar apparatus. However, the -325-mesh coal was still not completely devolatilized, although it is estimated that coal temperatures of about 2900' to 6300" C. would have been reached if thermal equilibrium had been attained. Total yields and carbon and ash recoveries were low in experiments 40 and 41 (Table 11), possibly because of losses of extremely fine solid residue that filtered through the recovery system. In all three experiments, the recoveries of hydrogen and oxygen were high, and in experiments 38 and 40, the yields of carbon monoxide obtained would not be possible even if all the oxygen in the coal had been converted to carbon monoxide. I t is likely that some cooling water leaked into the plasma. The reaction of this water with coal or coal decomposition products would account for the high hydrogen recoveries and the high yields of carbon monoxide. The predominance of acetylene in the gaseous products shown in Tables I and I1 and the absence of liquid products indicate that, although the coal was heated only to temperatures high enough for partial devolatilization to occur, the volatiles released attained much higher temperatures.
obtained in yields as high as 1 5 iceight Yc of maf coal. HOWever, the coal was not heated to temperatures high enough for complete devolatilization. To obtain higher coal temperature and yields of gaseous products, future ivork by the bureau will be concerned Lcith changes in plasma generator design and;'or operating techniques that may result in more efficient utilization of the heat available in plasmas. Higher coal temperatures should be attainable by feeding smaller coal particles and increasing the residence time of the coal in the plasmas.
literature Cited (1) hmmann, P. R., et al.. Chem. Eng. Progr. 60, 52-7 (1964). (2) Baddour; R. F., Blanchet, J. L., ISD. E s c . CHEM.PROCESS DESIGK DEVELOP. 3,258-66 (1964). (3) Baddour, R. F.. Iwasyk, J. M., Ibid., 1, 169-76 (1962). (4) Bond, R. L . >et al.. .Vature 200, 1313-14 (Dec. 28, 1964). (5) Freeman. M. P., Skrivan, 3. F., Hydrocarbon Process. Petrol. Rejner 41, 124-8 (.August 1962). (6) Leutner, H. LV., Stokes, C. S.: Znd. Eng. Chem. 53, 341-2 (1961).
RECEIVED for review January 21. 1965 ACCEPTEDAugust 13, 1965
Conclusions
Acetylene was the principal hydrocarbon gas produced when hvab coal was injected into argon plasmas and was
Division of Fuel Chemistry, 148th Meeting, ACS, Chicago. Ill., September 1964.
CATALYTIC OXIDATION OF METHANE R E l J I M E Z A K I A N D C H A R L E S C . W A T S O N Department of Chemical Engineering, Yniuersity of Wisconsin, Madison, Wis.
The catalytic oxidation of methane was investigated b y experiments conducted with a flow-type integral reactor. The reaction temperature ranged from 320" to 380" C. a t a pressure of 1 atm. The catalyst employed was palladium on alumina. The experimental results indicated that the reaction rate could be controlled b y a surface reaction in which the gaseous methane reacts with the adsorbed oxygen, producing adsorbed carbon dioxide and adsorbed water.
c
the total catalytic oxidation of light hydrocarbons has become increasingly important with respect to the problem of combustion removal of hydrocarbons in process tail gases such as the exhausts of internal combustion engines. The hydrocarbons from the exhausts have been said to contribute to the generation of smog. Combustion elimination of hydrocarbons is, therefore, one of the challenging problems to be solved for the prevention of smog formation. While recent development of automotive afterburner devices may be a sufficient remedy a t least to defer this problem, industrial exhaust gases containing unburned hydrocarbons and other organic contaminants will continue as a field for application of catalytic combustion processes. Since methane is a representative compound in the light hydrocarbon family and is the most difficult member of this family to oxidize, it was selected for this investigation. Many investigators ( 7 , 7-70) have reported the experimental results of the catalytic oxidation of methane. Their results are, however, useful only for the selection of a suitable oxidation catalyst. For use in equipment design and development studies, the results of kinetics research ought to be expressed in terms of practical models, as realistic as possible, to facilitate extrapURRENTLY
62
I & E C PROCESS D E S I G N A N D D E V E L O P M E N T
olation. Especially useful relations were developed by Langmuir and Hinshelwood, and are commonly called LangmuirHinshelwood models. These models, or reaction rate equations, result from a simple kinetic treatment in which a t least one reactant species must react chemically with one or more uniform sites upon the surface of a solid catalyst. Techniques of applying these concepts to the analysis of rate data of industrial importance were developed by Hougen and LYatson ( 4 ) . and the resulting rate expressions are sometimes also referred to as Hougen-Watson models. In the present work the kinetics of the total oxidation of methane was investigated by consideration of various possible Langmuir-Hinshelwood type models and a satisfactory rate equation was developed. Equipment
The experimental equipment is described in detail by Mezaki ( 5 ) . Figure 1 shows a schematic flow diagram of the entire equipment. Air, nitrogen, carbon dioxide, and methane were measured through previously calibrated flowmeters into the reactor
