Plasma Chemistry and Plasma Processing, Vol. 3, No. 4, 1983
P r o d u c t i o n of C a r b o n M o n o x i d e f r o m C a r b o n a n d C a r b o n D i o x i d e in a P l a s m a A r c R e a c t o r F. W. Giacobbe 1 and D. W. Schmerling 2 Received May 13, 1983; revised August 26, 1983
The production of carbon monoxide from the reaction of powdered carbon and carbon dioxide in a plasma arc reactor has been described. A description of the equipment, techniques, and results obtained are included. Periodic electrical to chemical conversion efficiencies of approximately 37% were achieved. Average yields of carbon monoxide in the product gases were as high as 90-95 % in selected experimental trials. Carbon monoxide could be produced at rates exceeding 9000 1/h (STP) with a power expenditure of about 27 kW.
KEY WORDS: Carbon monoxide; carbon dioxide; plasma; arc; jet; production; preparation.
1. INTRODUCTION The production of many major industrial chemicals is based upon the use of carbon monoxide. (1~ In some industrial manufacturing applications requiring the use of carbon monoxide, the gas is supplied via pipeline under long term take-or-pay contracts. Often, there are fluctuating demands for the products produced from the carbon monoxide and/or periodic and unexpected operating problems. As a result of these facts, the unit cost of the product produced from the carbon monoxide can become much higher than the anticipated unit cost. Therefore, in some instances, on-site production of carbon monoxide with an on/off system employing easily deliverable raw materials may be economically viable. One possible on-site method of producing carbon monoxide, from easily deliverable raw materials, involves a direct chemical reaction between Liquid Air Corporation, Technology Center, 5230 South East Avenue, Countryside, Illinois 60525. z Alloy Rods Division, Chemetron Corporation, Allegheny International Company, Hanover, Pennsylvania 17331. 383 0272-4324/83/1200-0383503.00/0© 1983PlenumPublishingCorporation
384
Giacobbe and Schmerling
solid carbon and carbon dioxide gas. The balanced chemical equation for this reaction is C(s) + CO:(g) = 2CO(g)
(1)
The effects of both temperature and pressure upon the equilibrium yields of carbon monoxide, produced according to Eq. (1), are illustrated in Fig. 1. The equilibrium yields plotted in Fig. 1 were obtained by employing standard thermodynamic methods. To simplify the calculations, ideal gas behavior was assumed and the effects of side reactions were ignored. I00
90
0
/ i
"
5O
J 0 >
20
50
/// / / / I / I ,
30
20
/
10
/
J o 60 ~
i
/
i .....
//
I ATM
,oo,,,
70
60 I
9O
i,'"/.//
0 400
50
,
4O
4o
i
/
200
I0
/
///
70
60
/'/
// ///
[
80
0 u
It/
I[XplrRIMENTAL~ ¢
600
I
,
800
1000
TEMPERATURE
r
1200
, 1400
I00 1600
(°C)
~ g . 1. Equilibrium yields of CO and CO2 vs. temperature for the reaction C(s)+ CO2(g ) = 2CO(g).
Production of CO from C and CO 2 in a Plasma Reactor
385
It can be seen in Fig. 1 that there is an excellent agreement between the theoretical curve and the average of six experimental measurements made at 800°C and a total system pressure of 1 atm. (2) Note also that essentially complete conversion of carbon and carbon dioxide to carbon monoxide can occur at temperatures greater than 1000°C, if the total system pressure is 1 arm. Even at system pressures as great as 100 atm, a temperature of 1600°C is high enough to produce almost pure carbon monoxide. The high carbon monoxide yields possible at elevated temperatures indicates that the production of carbon monoxide via Eq. (1), in a plasma arc, may be a practical method of producing carbon monoxide. Although high-temperature plasma arc methods have been utilized (3) since 1880 to effect various chemical reactions, no systematic attempts to produce carbon monoxide with a plasma arc reactor were found during a detailed literature survey. For these reasons, an experimental study involving the production of carbon monoxide, in a plasma arc reactor, was initiated. The objectives of this study were to determine the feasibility of and the economics related to the production of carbon monoxide in a plasma arc reactor. The techniques and results of the initial experimental work carried out are reported in this paper. 2. EXPERIMENTAL 2.1. Plasma Arc Reactor
During the course of experimental work, a number of plasma arc reactors were tested. A diagram of the most efficient and reliable plasma arc reactor tested is shown in Fig. 2. This reactor was used to produce carbon monoxide by a direct high-temperature reaction between powdered carbon (Germantown Bear Brand or B-5 Lampblack made by the Monsanto Chemical Company) and a plasma produced in pure carbon dioxide. The carbon dioxide plasma was initiated by a high-voltage, high-frequency electrical discharge (3000 V / 4 MHz, for 0.5 sec) between the cathode and anode of the reactor. A 160-kW constant-current plasma torch power supply (Model C-2509-MC-2A, made by the P & H Welding Equipment Company for the Chemetron Corporation) was used to sustain the plasma. The most radical difference between the plasma arc reactor illustrated in Fig. 2 and conventional solid/gas plasma arc reactors (g) involves the powdered carbon injection site. Powdered carbon, carried in a stream of carbon dioxide~ was injected into the plasma arc reactor above the arc discharge region. This method of injection was attempted unsuccessfully in at least two studies of plasma reactions with powdered coal. (5'6) Experimentally, it was found that carbon injection above the arc discharge region,
386
Giacobbe and Schmerling
~
CARBON ELECTRODE
O'I ~ ' ~ ' - POWER CONN.(-)
WATER OUT
--WATER
/--
CO2+CARBON,N
IN
INSULATOR
~,.~]~- POWER CONN.(+)
I;iil
PLASMA BODY WATER I N -
"WATER OUT
WATER OUT COOLING CHAMBER
~ L
~_WATER iN
I
CO OUTJ
0 I 2 1,1,1 ,=,l APPROX.
SCALE
(iN)
Fig. 2. Plasma arc reactor. instead of beneath it into the plasma flame, produced significantly higher carbon monoxide yields in the product gases. Consumable carbon electrodes (0.794 cm o.d. x 30.5 cm long) were used in the plasma arc reactor because nonconsumabte tungsten electrodes or inserts deteriorate rapidly in a carbon dioxide plasma. The plasma arc reactor employed a nontransferred arc mode of operation. A transferred arc reactor was tested but discarded because of short circuiting due to the deposition of a carbon layer, within the reactor, during operation. Normally,
Production of CO from C and CO 2 in a Plasma Reactor
387
the plasma arc reactor consumed energy at the rate of approximately 27 kW. At this rate of electrical energy consumption, in an atmosphere of carb o n / c a r b o n dioxide, the 0.974-cm carbon electrodes were consumed at a rate of approximately 12.7 cm/h. This rate of electrode sublimation was not great enough to provide enough carbon for 100% yields of carbon monoxide at the high input rates of carbon dioxide normally used. For this reason, additional carbon was injected into the carbon dioxide gas entering the reactor above the arc discharge region. The carbon electrodes were manually advanced, during most of the experimental runs, to compensate for sublimation effects. Total carbon dioxide flow rates into the plasma arc reactor ranged between 25 and 100 liters/min at STP (0°C and 1 atm). Part of the carbon dioxide (about 50% at a total flow of 100 liters/rain) carried powdered carbon into the plasma reactor. The balance of the carbon dioxide entered the reactor through the gas entry ring surrounding the carbon cathode. The carbon dioxide flow through the gas entry ring inhibits short circuiting through a carbon layer which may form between the cathode and anode of the reactor. 2.2.
Carbon Feeder
Carbon dioxide gas flow rates into the plasma arc reactor were monitored with Fischer-Porter rotameters. A precision rotameter (type 10A1800) was used to monitor the total carbon dioxide gas flow rate. The estimated errors in the flow rates measured with this rotameter were about +1.5 liters/rain. Separate rotameters were used to measure gas flows to the carbon feeder (Fig. 3) and directly to the plasma arc reactor. Product gas flow rates were also monitored, after cooling and filtering, with a positivedisplacement Roots Meter (Model 1.5M125). Carbon powder flow rates were estimated by measuring the time required to transfer a known mass of carbon from the carbon reservoir to the carbon feeder/plasma arc reactor system. Combining this data with the total carbon dioxide flow rate allowed an estimate of the c a r b o n / c a r b o n dioxide ratio. The estimated errors in the c a r b o n / c a r b o n dioxide ratio determined in this way were about +0.05 g/liter. 2.3. Plasma
Quenching
The hot plasma and excess carbon leaving the reactor were rapidly quenched in a water-cooled expansion chamber and then by a watercooled/tube-in-shell heat exchanger. The heat exchanger was made in two separate sections. Each section consisted of four 2.54 cm o.d. standard wall
388
Giacobbe
~ COn+f~A~R~N
~ ---~
and
Schmerling
ROTARYVALVE
~
M I PELLER
! ...........I!t!l!IUjt!Jl MOTORDRIVE 0
i
2
I,
, , i , E,
i
APPROX. SCALE
tiN) Fig. 3. C a r b o n f e e d e r .
copper tubes soldered in a symmetrical square array inside of a 0.914 m length of 7.94 cm o.d. copper tubing. The hot product gases and excess carbon from the plasma arc reactor were conducted through the 2.54-cm copper tubes and cooling water passed through the 7.94-cm copper shell. In a typical run, the total flow of water through the plasma arc reactor and heat exchanger was about 760 liters/h. The overall increase in water t e m p e r a t u r e was approximately 20°C. Rapid quenching of the product gases was essential in order to produce high carbon monoxide yields. This is a direct result of the fact that carbon monoxide is thermodynamically unstable but k]netically stable with respect to carbon and carbon dioxide at low temperatures. A gradual decrease in
Production of CO from C and C 0 2 in a Plasma Reador
389
the temperature of carbon monoxide, especially in the presence of carbon, (7) would tend to permit the formation of the equilibrium ratios of carbon monoxide/carbon dioxide predicted in Fig. 1. The final ratio would depend mostly upon the temperature range in which the approach to equilibrium is inhibited for kinetic reasons. (8-tl) 2.4. Carbon Powder Filter
The carbon powder filter used to separate the excess solid carbon from the cooled product gas stream was a modified commercial Torit dust collector (76.2 cm long× 61.0 cm wide× 76.2 cm high). The main modification involved the removal of the internal gas circulation fan. A 5.08-cm pipeline carried the cooled product gases and excess carbon powder into the base of the dust collector. On the opposite side of the filter element, on top of the unit, a second 5.08-cm line carried the filtered product gas to the Roots Meter. A diagram of the entire plasma arc system which indicates the relative positions of all major system components is shown in Fig. 4.
,CARBON HOPPER
PLASMA REACTOARC R ~ --~
E R ~ WAT OU3; ( ~
n r- ~
ROTARYVALVE I ¢:~=~/CARBON FEEDER _
_
C02
/- CO
DISCHARGE
INLET COOLING CHA WATERIN ROOTS METER-~E=~ GAs COOLER
CARBON
WATER OUT
FILTER
) GAs COOLE-/
VALVEJ Fig. 4, Plasma arc system.
390
Giacobbe and Schmerling
2.5. Analysis of Product Gases P r o d u c t gases were analyzed with a standard Orsat apparatus. Gases to be analyzed were t r a p p e d by water displacement in a 15.I-liter stainless steel drum. Before each analysis, the Orsat apparatus was t h o r o u g h l y flushed with samples of the gas to be analyzed to r e m o v e traces of the previous gas sample. T h e estimated errors in the Orsat results were a b o u t ± 3 % of the readings obtained. Most of the product gases were disposed of by burning in air after they had passed through the R o o t s Meter. A p r o p a n e torch flame was used to ignite the combustible gas mixture p r o d u c e d by the plasma arc reactor. The p r o p a n e flame was n e e d e d to sustain combustion in gas mixtures low in c a r b o n monoxide. A t c a r b o n m o n o x i d e levels a b o v e about 5 0 % by volume, the gas mixture would continue to burn in air, once ignited. The burning gas mixture p r o d u c e d a light blue flame when all traces of p o w d e r e d c a r b o n were effectively filtered out of the gas. D u r i n g all experimental trials, laboratory air was continuously m o n i t o r e d with an M S A C a r b o n M o n o x i d e Indicator (model 70). A n audible alarm within the indicator was set to ring if the level of carbon m o n o x i d e in the air exceeded the threshold limit value of 50 ppm. (12) 3. R E S U L T S A N D D I S C U S S I O N A mental carbon during
s u m m a r y of experimental data, collected during some runs carried out during this study, is given in Table m o n o x i d e concentration as high as 9 9 % by volume one of the experimental runs. T h e c a r b o n / c a r b o n
of the experiI. N o t e that a was achieved dioxide ratio
Table I. Operating Conditions and Results Obtained During Production of Carbon Monoxide in Plasma Arc Reactor Total COz input flow rate at STP (liters/min) 88 88 88 88 62 73 60
Average CO Electrical yield in to chemical C/CO 2 Total product conversion ratio Current ~ Voltage Power run time gases efficiency (g/liter) (A) (V) (kW) (rain) (% by vol.) (%) Electrical characteristics
0.45 0.45 0.45 0.45 0.55 0.40 0.40
130 130 135 130 140 135 135
200 200 200 200 200 200 200
26 26 27 26 28 27 27
10 11 10 16 21 12 32
92 90 93 92 99 83 85
37 36 36 37 28 25 21
Production of CO from C and C02 in a Plasma Reactor
391
delivered into the plasma arc reactor during this run was approximately 0.55 g/liter. This was very close to the carbon/carbon dioxide ratio of 0.54 g/liter which is required to produce carbon monoxide in 100% yields. The carbon dioxide flow rate during this run was only about 62 liters/min. At higher carbon dioxide flow rates the carbon/carbon dioxide ratio dropped to approximately 0.45 g/liter. This accounts for the drop in carbon monoxide yield from approximately 99 to 92%. Electrical to chemical conversion efficiencies were also calculated from the experimental data. These efficiencies are included in Table I. The conversion efficiencies were calculated from the ratio of the energy required to produce carbon monoxide and the total electrical energy actually consumed. The energy of formation of 20,610 cal/mole for carbon monoxide, produced according to Eq. (1), was used in these calculations. It may be seen that electrical to chemical conversion efficiencies of approximately 37% were achieved in two of the experimental runs indicated in Table I. It is possible that even higher electrical to chemical conversion efficiencies could have been achieved in these runs if the carbon/carbon dioxide ratio had been higher than 0.45 g/liter. The electrical to chemical conversion efficiencies achieved in the course of this work are quite high compared to those achieved by other researchers who have attempted to synthesize selected chemical compounds via plasma arc chemistry. For example, acetylene, (13~ cyanogen, (t4-16~ and hydrogen cyanide (16~1s~ have all been synthesized in plasma arc studies directly from the elements, but the highest electrical to chemical conversion efficiency achieved in these studies was less than 2%. Generally, the plasma arc reduction of selected metal oxides in hydrogen is even less efficient. (18-2°~ The most efficient plasma arc reaction encountered, during the course of the literature survey made in connection with this study, involved the production of acetylene from methane.(21'22) The acetylene was produced from pure methane in a multimegawatt AC arc. The reported data indicate an electrical to chemical conversion efficiency of approximately 68 %. This figure may be close to the upper limit of electrical to chemical conversion efficiency possible in high-temperature plasma arc systems since some energy must always be lost as high-temperature products are cooled to ambient temperatures. From an economical standpoint, electrical to chemical conversion efficiencies are more important than the achievement of extremely high carbon monoxide yields in the product gases, provided that the excess carbon and carbon dioxide in the product stream can be economically separated and recycled. Due to experimental difficulties, total carbon dioxide flow rates higher than 62 liters/min, containing a carbon/carbon dioxide ratio of at least 0.54 g/liter, were not tested with the equipment described above. However,
392
o t h e r e q u i p m e n t c a p a b l e of h a n d l i n g h i g h e r c a r b o n constructed and tested. Attempts were made to c h e m i c a l c o n v e r s i o n efficiencies, i n c r e a s e c a r b o n rates, and automate feeding the carbon electrode T h e r e s u l t s r e l a t e d to t h e s e e f f o r t s will b e r e p o r t e d
Giacobbe and Schmerling
d i o x i d e flow r a t e s was i m p r o v e e l e c t r i c a l to monoxide production into the plasma torch. in a s e p a r a t e p a p e r .
REFERENCES 1. C.A. Rohrmann etal., Chemical Production From Waste Carbon Monoxide--Its Potential For Energy Conservation, Battelle-Pacific Northwest Laboratories, Richland, Washington (1977) pp. 94-134. 2. E. L. Quinn and C. L. Jones, Carbon Dioxide, Reinhold, New York (1936), p. 132. 3. J. Dewar, Proc. R Soc. London 30, 85 (1880). 4. M. Venugopalan (ed.), Reactions under Plasma Conditions, Vol. II, Wiiey-Interscience, New York (t971), p. 262. 5. E.H. McDonald and G. R. Hill, Plasma Reactions with Powdered Coal, Fuels Engineering Department, University of Utah, Salt Lake City (1966), pp. 13-26. 6o R. L. Bond, et aL, Nature 200, 1313 (1963). 7. French Patent 874, 681, Societe Francoise au Carbon alpha et Ses Derives. 8. R. F. Baddour and R. S. Timmins, The Application of Plasmas to Chemical Processing The M. I.T. Press, Cambridge (1967), p. 32. 9. P. C. Wu, The Kinetics of the Reaction of Carbon with Carbon Dioxide, Sc.D. Thesis, Chem, Eng., Mass Inst. of Tech. (1949). 10. J. Gadsby, et al., Proc. R. Soc. London A 187, 129 (1946). 11. Jo Gadsby et al., Proc. R. Soc. London A 193, (1948). 12. N. I. Sax, Dangerous Properties of Industrial Materials, 3rd edn., Van Nostrand Reinhold, New York (1968), p. 533. 13. H. W. Leutner and C. S. Stokes, Ind Eng. Chem. 53, 341 (1961). 14. C. S. Stokes and W. W. Knipe, Ind Eng. Chem. 52, 287 (1960). 15. H, W. Leutner, Ind. Eng. Chem., Process Des Dev. 1, 166 (1962). 16. A. V. Grosse, et al., Plasma Jet Chemistry, First Annual Report. Office of Naval Research Contract NONR-3085 (02), The Research Institute of Temple University, Philadelphia, Pennsylvania (1961). 17. C. S. Stokes and J. A. Cahill, Plasma Jet Chemistry, Final Report. Air Force Office of Scientific Research Grant 775-65, The Research Institute of Temple University, Philadelphia, Pennsylvania (1965). 1.8. H. W. Leutner, Ind. Eng. Chem., Process Des. Dev. 2, 315 (1963). 19. C. S. Stokes, Adv. Chem. Ser. 80, 390 (1969). 20. C. S. Stokes, et al. Plasma Jet Chemistry, Final Report. Air Force Office of Scientific Research, Grant 62-196, The Research Institute of Temple University, Philadelphia, Pennsylvania (1964). 21. W. K. Maniero et al., Westinghouse Eng. 26, 66 (1966). 22. C. Hirayama and W. K. Manierto, Am. Chem. Soc., Div. Fuel Chem., Preprints, 11, 470 (1967).