Catalytic activity of coal ash on steam methane reforming and water-gas shift reactions

Fuel Processing Technology, 16 (1987) 279-288 ElsevierSciencePublishers B.V., Amsterdam-- Printed in The Netherlands

279

Catalytic A c t i v i t y of Coal Ash on S t e a m M e t h a n e R e f o r m i n g and Water-Gas Shift R e a c t i o n s W.J. CHEN, F.R. SHEU and R.L. SAVAGE Department of Chemical Engineering, Ohio University, Athens, OH 45701 (U.S.A.) (ReceivedMay 23th, 1985;acceptedOctober 10th, 1986)

ABSTRACT The catalytic activity of Clarion 4 A coal ash on water-gas shift,steam methane reforming and carbon dioxide methane reforming reactions was studied in an atmospheric fluidized bed over temperatures ranging from 370 oC to 900 °C. The rates of steam methane reforming and water-gas shift reactions over ash were found to be several times the corresponding non-catalytic reactions but were one order of magnitude lower than the reactions over commercial catalysts. Also, it was found that the ash affected the water-gas shift reaction more significantlythan the steam methane reforming reaction probably because of the larger amount of iron oxide in the coal ash. However, there was no appreciable reaction for carbon dioxide reforming up to 900 °C.

INTRODUCTION Since the discovery that the minerals in coal promoted solvent refined coal process [ 1 ], there has been growing interest in the catalytic activity of coal ash on other coal conversion processes. Huettinger and Krauss measured the kinetics of hydrogasification [2] and water vapor gasification [3] and concluded that iron in coal catalyzed high temperature char gasification reactions. On the other hand, there is also a need to isolate and determine the catalytic activity of coal ash on gas phase reactions during coal gasification. Some of the gas phase reactions such as water-gas shift and methane reforming are also catalyzed by iron. In fact, iron-based commercial catalysts have long been used in the production of synthesis gas by natural gas reforming. One interesting observation from Huettinger and Krauss data was that the catalytic production of methane decreased simultaneously with an increase in carbon monoxide when the steam feed was raised. The present authors think this was a positive indication that catalytic methane reforming was at work. Furthermore, the gas phase reactions are probably more important than the char gasification reactions in high temperature partial gasification of coal [ 4 ]. This paper attempts to study the catalytic activity of coal ash on the following three reactions: 0378-3820/87/$03.50

© 1987ElsevierSciencePublishers B.V.

280 H 2 0 + C O ~ H 2 +C02

(1)

CH4 + H 2 0 ~ - C O + 3 H 2

(2)

CH4 + CO2~ 2CO + 2H2

(3)

The kinetics of water-gas shift and methane reformingreactions over commercial catalysts havebeen wellstudiedin the past. Numerousmodels [5-15] havebeen suggestedfor each reaction. Laupichler'smodel [5 ] was most widely used for water-gas shift reactions and has been found to correlate both the homogeneousnoncatalyticand the industrialcatalyticdata well.Severalmodifications havebeenproposed. Atwoodet al. [6 ] simplifiedLaupichler'smodel to account for the high steam and carbon dioxide concentrationand Kodama et al. [ 7] modifiedLaupichler's model for short reaction time. Other models, such as the Bohlbro [8] and Goodridge and Quazi [9] modelswere simply empirical models. Therefore, Laupichler's model is still consideredadequate for water-gas shift reactions, i.e., r = k (PH2 o Pco - Pco~ PH~/K

(4)

where k is the rate constant, P the partial pressure, and K the equilibrium constant. For steam methane reforming, Akers and Camp [ 10 ] proposed the first and the simplest model. Many complicated models [11-13] followed since, but all of them were concerned with the commercial nickel catalyst. Therefore, a simple model like Akers and Camp's may be as good as the others when coal ash is the catalyst because it has little or no nickel. A similar expression can be used for the carbon dioxide methane reforming reaction, i.e., r = k PCH4

(5 )

EXPERIMENTAL Preparation of coal ash

Ohio Clarion 4A coal was incinerated to produce ash for this study. The coal was first pulverized. The portion between 40 and 100 mesh was collected in several crucibles and then placed in a muffle furnace. The furnace was then heated up and maintained at a temperature of about 760 ° C with a continuous supply of oxygen. In order to ensure that all the carbon and hydrogen were completely combusted, the crucibles were removed from the furnace at hourly intervals and the residues were stirred. After twelve hours, a small amount of ash was removed for analysis by the Perkin-Elmer 2400 elemental analyzer. If the ash was found to be carbon-free, it was then cooled and collected in a plastic

281 TABLE 1 Elemental analysis of ash Component

Weight %

Si02 A1203 CaO MgO Na20 K20 Fe203 S03 Ti02 P205

38 24 3.3 1.1 1.0 1.5 25 4.0 1.0 0.1

bottle. The elemental analysis of ash is given in Table 1 [16]. The average particle size of ash is 285/lm with a nitrogen surface area of 6 m2/g. Fluidized bed reactor

A schematic diagram of the experimental apparatus is shown in Fig. 1. It consists of a steam generator, a mixing chamber, and a cyclone centered around a fluidized bed reactor. The reactor was fabricated from a type 304 stainless j

Vent ;: ~ Sampling

Point

Temp. Controller 8, Indicalor o

~-~ ~

SteamReactant

Fig. 1. Diagram of experimental set-up.

Sampling Point

282 steel pipe of 7.6-cm internal diameter and 122-cm length. The temperature of the reactor was maintained by 46-cm-long surrounding electrical heating elements through a temperature controller. The 6000 W heater could maintain temperatures up to 900 ° C. A shallow fluid bed of 5.1 cm to 12.7 cm height was used in the kinetic study so the reactor could be approximated by a stirred tank reactor. This simplified the subsequent analysis considerably. The parameters under investigation included the weight of coal ash, feed gas composition and temperature. During each experiment, the reactor was first loaded with a measured amount of ash. Then, the ash was treated in-situ with fluidizing air at 540 °C for three hours. The reactor was then heated up to the reaction temperature with fluidizing nitrogen. Upon approaching the temperature set point, water at the desired flow rate was admitted through the steam generator, where the evaporation occurred at a temperature of 260 °C and then the product steam was passed to the reactor. A measured amount of the other reactant gases were fed simultaneously from compressed gas cylinders. After steady state operation was established, several product and feed gas samples were taken and analyzed by gas chromatography. At the end of the experiment, the weight of coal ash was checked to make sure that the loss due to elutriation was not significant. In order to determine the possible catalytic effect of the reactor wall, blank tests were made by replacing ash with sand and conducting experiments under similar conditions. RESULTS AND DISCUSSION Both the water-gas shift (eqn. (1) ) and steam methane reforming reactions (eqns. ( 2 ) and ( 3 ) ) over ash were appreciable, though the reaction of methane and carbon dioxide was not significant up to 900 ° C. The results are summarized in Tables 2 and 3, where the reaction rate in the last column is calculated by the following relation:

-r=Fx/W

(6)

where F is the feed reactant flow rate, x is the conversion and W is the mass of the ash. The nitrogen in the feed was not only used as an inert to control the feed composition, but also as a tracer to calculate the product flow rate. In Table 2, the top 11 entries summarize the tests of the dependency on concentration, while the remaining entries summarize the tests of the dependency on temperature for the water-gas shift reaction. By plotting the first eleven rates against the concentration driving force proposed by Laupichler in Fig. 2, a reasonably good fit was obtained as indicated by the linear relation. Furthermore, the rate constant can be determined from the slope of the straight line. The temperature dependency of the rate constant was studied from 669 K to 824 K and was found to follow the Arrhenius equation as shown in Fig. 3. In

Wt. o f a s h (kg)

0.250 0.250 0.250 0.250 0.250 0.250 0.315 0.315 0.315 0.315 0.315

0.33 0.33 0.33 0.33 0.33 0.33 0.33 0.33 0.33

Temp. {K)

814 814 814 814 814 814 814 814 814 814 814

669 699 749 758 765 772 777 814 826

0.1678 0.1678 0.1678 0.1678 0.1678 0.1678 0.1678 0.1678 0.1678

0.1557 0.1557 0.1678 0.1678 0.1904 0.1678 0.1557 0.2801 0.1678 0.1557 0.1904

CO

0.0402 0.0825

0.0571 0.0846 0.0571

0.0571

0.0825 0.0825 0.0402 0.0402 0.0402 0.0825 0.0402

H2

0.0846 0.0571 0.0571 0.0211 0.0846

C02

Feed gas flow rate ( g m o l e / m i n )

0.179 0.179 0.179 0.179 0.179 0.179 0.179 0.179 0.179

0.0750 0.0463 0.0803

0.0463 0.0738 0.0750 0.1109 0.0715 0.0965 0.0750

N2

0..1666 0.1666 0.1666 0.1666 0.1666 0.1666 0.1666 0.1666 0.1666

0.1443 0.1443 0.1733 0.1733 0.1268 0.1666 0.1854 0.2333 0.1733 0.1443 0.1856

HzO

S u m m a r y of experimental results for water-gas shift reaction CO + H20--.CO 2 + H 2

TABLE 2

32.4 32.2 31.0 31.0 30.5 30.4 30.3 28.7 28.1

28.1 28.1 29.7 29.4 34.5 30.2 26.4 45.9 29.0 27.8 32.1

CO

32.1 32.0 30.8 30.7 30.3 30.1 30.0 28.5 27.9

25.8 25.9 30.8 30.4 22.1 30.0 32.3 36.7 30.1 25.5 31.2

H20

0.32 0.45 1.64 1.74 2.15 2.30 2.41 3.95 4.52

18.3 18.3 10.8 11.1 10.4 18.5 11.7 8.7 11.5 18.7 16.1

H2

in product gas ( a t m × 100)

Partial pressure

0.32 0.45 1.64 1.74 2.15 2.30 2.41 3.95 4.52

18.7 13.4 14.1 7.4 19.0 2.5 15.0 8.7 14.8 19.1 5.0

C02

0.298 0.416 1.534 1.593 2.010 2.150 2.250 3.687 4.216

2.790 2.766 3.682 4.085 3.134 3.026 3.761 8.524 3.579 2.555 4.885

Reaction rate - r(co) ( gmole/kg h )

b,0 O0

284 TABLE 3 Summary of experimental results for steam-reforming reaction CH4 + H20-~ CO + 3 H2 Temp. (K )

Wt. of ash Feed flow (gmole/min) (kg) CH4 H20 N2

Conversion CH4 reacted CH4 feed

Partial pressure CH4 out (atm)

Reaction rate - r (CH,} (gmole/kg h )

1196 1196 1196 1196 1042 1079 1115 1140 1160 1172

0.482 0.482 0.482 0.482 0.550 0.550 0.550 0,550 0.550 0.550

0.1952 0.1953 0.1991 0.2049 0,0430 0.0559 0.0911 0.1197 0.1193 0.2040

0.0756 0.1103 0.0525 0.1271 0.1360 0.1338 0.1282 0.1217 0.1115 0.1095

1.630 2.390 1.157 3.025 0.461 0.599 0.977 1,284 2.071 2.188

0.0671 0.0983 0.0467 0.1186 0.0983 0.0983 0.0983 0.0983 0.0983 0.0983

0,240 0,240 0,240 0,240 0,240 0,240 0,240 0.240 0.240 0.240

0,385 0.354 0.405 0,346 0.354 0.354 0,354 0.354 0.354 0.354

addition, the activation energy and the frequency factor were calculated from the graph to be 21700 cal/gmole and 3.26 × 107 gmole/kg ash h atm 2, respectively. Similar work was also carried out for the blank test. The catalytic activity of the blank test was significant, but it was only 10-20% of the overall I0"001 9.101 8.20 7,30 v

"Z ctJ 0

"6

q,$

E

o

!

6.405.50" 4.603.702.80"

/.

1,90' I.OC 6.00

8)5 ~d.50 ~2.75 iS.00 ~z25 ( PCOPM20"PH2PCo2/ k ) alm 2

Fig. 2, D e p e n d e n c e of - rco o n P c o PH2o - Pco2

PHJKfor w a t e r - g a s

shift reaction.

285 5.00. 4.50, 40O 3.50 3.00

2.50 C:

2.001.501.00" 0.50 0.00 1.10

L;'O

1.30

~.40

I.,50

1.6o

I

Tx io3fl/k) Fig. 3. Arrhenius plot of rate constant for water-gas shift reaction. activity for the temperature range studied. A comparison of the rate constants at 700 K is presented in Table 4, together with a value recalculated from Moe's [ 17 ] study on the commercial catalyst. The activity of commercial catalyst is one order of magnitude higher t h a n that of coal ash. The primary reason is due to the iron content which is considerably higher in the commercial catalyst t h a n in coal ash. The other reason is probably due to the structure properties, such as porosity and surface area. Table 3 summarizes the results for the steam methane reforming reaction at different concentrations and temperatures. In this case, the Akers model was found to be adequate to describe the concentration dependency again, the Arrhenius equation could be employed to describe the temperature dependency TABLE 4 Comparison of rate constants for water-gas shift reaction Temperature (K)

700

k (gmole/h kg of catalystatm 2) Moe

Ash

Blank

296

5.32

0.70

286

5.20-

2.98.

•

,r,,

o 2.762.54. .c 2.3; u~

2.1( w ~

1.88

z

o_

~.66 I.L~ "~ 1.44 W

1.22 1.00 ).04

0.66

0.68

0]0

0.'12

0.14

PCH 4 ( o t m ) Fig. 4. Dependence of - rcH, on PCH, for s t e a m - m e t h a n e reforming reaction.

3.20 2.882.56" 2.24. 1.92" 1.60" e-

1.280 96' 0.64-

0.32-

°o b o

8~40

8~80

9~20

9.co0 I0.00

± 4 T xlo (I/k)

Fig. 5, Arrhenius plot of rate constant for steam-methane reforming reaction.

287 TABLE

5

Comparison of rate constants for steam methane reforming reaction

Temperature (K)

1280

Aker's (commercial catalyst)

k (gmole/h kg of catalyst atm. ) Ash

Blank

Gordon's [ 18] (no catalyst)

4030

56.2

8.5

0.6

as in Figs. 4 and 5. The activation energy and the frequency factor were 32800 cal/gmole and 2.23 × 107 mole/kg ash h atm 2, respectively. In order to compare the rate constants, both Akers' study with the commercial catalyst and Gordon's study without a catalyst were recalculated and included in Table 5. It can be seen that the activity of the commercial catalyst is one order of magnitude higher than that of coal ash and the wall effect is one order of magnitude higher than that of a non-catalytic reaction. CONCLUSION Coal ash catalyzed both the water-gas shift reaction and the steam methane reforming reaction. The kinetic expression for the water-gas shift reaction was found to be in agreement with the Laupichler's second-order model and could be represented by rco = - 3.26 × 107 exp ( -

21,700/RT) [Pco P H 2 0

--

PH2Pco2/K]

On the other hand, the result of the steam methane reforming reaction was consistent with the Akers first-order model and could be written as rcm = - 2.23 × 107 exp ( -

32,800/RT) PcH4

The catalytic activity of coal ash was one order of magnitude lower than that of the commercial catalyst. The catalytic effect of stainless steel wall was significant, but was only a fraction of that of coal ash.

REFERENCES 1 2 3

Probstein, R.F. and Hicks, R.E., 1982. Synthetic Fuels. McGraw-Hill, p. 299. Huettinger, K.J. and Krauss, W., 1981. Fuel, 60: 93. Huettinger, K.J. and Krauss, W., 1982. Catalytic activity of coal minerals in water vapor gasificationof coal. Fuel, 61: 291.

288 4

Savage, R.L., Chen, W.J., Sato, A., Patke, N.G., Hereth, M., Tong, J.Y. and Kneller, W.E., 1982. Process for the concurrent production of hydrogen, carbon monoxide and low-volatile, low-sulfur char by flash carbonization of coal. Int. J. Hydrogen Energy, 7: 717. 5 Laupichler, F.G., 1938. Catalytic water-gas shift reaction. Ind. Eng. Chem., 30: 579. 6 Atwood, K., Arold, M.R. and Appel, E.G., 1950. Water-gas shift reaction. Ind. Eng. Chem., 42: 1600. 7 Kodama, S., Mazume, A., Fukuba, K. and Fuku, K., 1955. Reaction rate of water-gas shift reaction. Bull. Chem. Soc. Jpn., 28: 318. 8 Bohlbro, H,, 1961. The kinetics of the water-gas conversion at atmospheric pressure. Acta. Chem. Scand., 15: 501. 9 Goodridge, F. and Quazi, H.A., 1967. The water-gas shift reaction. A comparison of industrial catalysts. Trans. Inst. Chem. Eng., 45: T274-T279. 10 Akers, W.W. and Camp, D.P., 1955. Kinetics of the methane-steam reaction. AIChE, 1: 471. 11 Bordrov, N.M. and Apelbaum, L.O., 1967. Kinetics of the reaction of methane with steam on a nickel surface. Kinet. Catal., 8: 379. 12 Ross, J.H. and steel, M.C.F., 1972. Mechanism of the steam reforming of methane over a coprecipitated nickel-aluminum catalyst. Trans. Farad. Soc., p. 10. 13 Haldor, T., 1966. Some aspects of catalytic reforming. I.G.E.J., p. 402. 14 Sodesawa, T., Dobashi, A. and Nozaki, F., 1979. Catalytic reaction of methane with carbon dioxide. React. Kinet. Catal. Lett., 12: 107. 15 Bordrov, I.H. and Apelbaum, L.O., 1967. Kinetics of the reaction of methane with carbon dioxide on a nickel surface. Kinet. Catal., 8: 379. 16 Botoman, G. and Stith, D.A., 1978. Analysis of Ohio coals. Ohio Geol. Survey, Information Cr. No. 47, p. 148. 17 Moe, J., 1962. Design of water-gas shift reactors. Chem. Eng. Prog., 58: 33. 18 Gordon, A.S., 1948. Pyrolysis of methane flowing through a porcelain tube in the region 1000. J. Amer. Chem. Soc., 70: 395.