Sulfur Condensation In Claus Catalyst

a

Adsorbents & Catalyst

Sulfur Condensation in Claus Catalyst By G. R. Schoofs, Princeton University, Princeton, NJ Published in HYDROCARBON PROCESSING, February 1985

CONTENTS: Catalyst Deactivation Capillary Condensation Sulfur Properties Estimates & Discussion Literature Cited

2 3 4 5 6

T E C H N I C A L B U L L E T I N USA/6020-R01/0404

Page 1 of 6

Theory suggests what makes for Claus catalyst deactivation from sulfur condensation

CATALYST DEACTIVATION by pore plugging occurs in many heterogeneous reactions: •Coke formation in petroleum processing, especially hydrocracking1 and hydrodesulfurization2-4 catalysts •Steam reforming and methanation catalysts5 •Ammonia synthesis catalyst5,6 •Automobile exhaust catalysts7,8 All involve deposition of solids in catalyst pores. However, Claus catalyst deactivation involves catalyst pore plugging by liquid sulfur. The Claus process converts hydrogen sulfide produced by petroleum desulfurization units and gas treatment processes into elemental sulfur and water. More than 15 million tons of sulfur are recovered annually by this process. The overall reaction may be written as10 2H2S + (1 + a)O2 Õ 2H2O + ((2 - a)/n)Sn + aSO2 (1) where, a = 0 to 2 and n = 2 to 8 and possibly more Commercial Claus plants appear to operate at thermodynamic equilibrium.11,12 Depending on the H2S content of the feed and the number of reactors, total H2S conversion to elemental sulfur can exceed 95 %.

A

Figure 1—Schematic diagram of a typical Claus plant. Converter 1 Acid Gas Air

Reaction Furnace

Reheater 1

Condenser 1

Converter 2 Reheater 2

Condenser 2

Reheater 3

Condenser 3

Sulfur Pit

Page 2 of 6

Converter 3

Tail Gas

Condenser 4 Liquid Sulfur

USA/6020-R01/0404

100

Conversion, %

90

Figure 2—Theoretical equilibrium conversion for oxidation of hydrogen sulfide with stoichiometric air to sulfur vapor and water at a total pressure of 1 atm.

80 70 60 50 400

600

800

1000

1200

1400

1600

Temperature, °K

Fig. 1 shows a typical Claus plant. Acid gas and air enter a furnace operating at about 1,000°C to achieve a 70% conversion to elemental sulfur and also to produce an H2S:SO2 molar ratio of 2:1. Principal side reactions in the furnace include formation of COS and CS2 from residual hydrocarbons. H2S and SO2 react further over alumina catalyst in one or more subsequent catalytic converters (2) 2H2S + SO2 Õ 2H2O + 3/nSn Typically COS and CS2 are also hydrolyzed in the first catalytic converter COS + H2O Õ H2S + CO2 (3) CS2 + 2H2O Õ 2H2S + CO2

(4)

Reaction 2 follows reactions 3 and 4. Sulfur is condensed and recovered from the gas streams leaving the furnace and the catalytic converters. Fig. 2 shows the equilibrium sulfur yield as a function of temperature for the overall reaction.13 Thermodynamics provide a strong incentive to operate the catalytic converters at low temperatures. The first converter typically operates at about 350°C in order to hydrolyze COS and CS2. The second and subsequent converters usually operate just above the dew point of sulfur vapor. However, sulfur may still condense in the catalyst pores, since small pores can hold a condensed liquid at vapor pressures below the normal condensation point. Sulfur condensation in catalyst pores causes a decline in Claus process performance, since it reduces the catalyst surface area and since sulfur exhibits little catalytic activity.14 Although sulfur condensation is well recognized as USA/6020-R01/0797

one of several mechanisms of Claus catalyst deactivation, the nature of the problem and its implications for optimum Claus catalyst design have not been fully explored. The physics of capillary condensation provides a basis for estimating catalyst pore sizes that plug with sulfur as functions of the catalytic converter temperature, and of the difference between the catalytic converter temperature and the sulfur dew point temperature.

CAPILLARY CONDENSATION

Lord Kelvin recognized that the vapor pressure of a liquid contained in a small diameter capillary is less than the normal value predicted for a free surface.15 For the case of evaporation from a capillary partially filled with a liquid d=

-4gV cos q RT ln(P/Po)

(5)

where d = diameter of the capillary or pore containing condensed liquid, cm g = surface tension of the liquid, dyne/cm V = molar volume of the liquid, cc/mole q = contact angle, degrees R = Universal Gas Constant = 8.314 x 107 erg/mole/°K T = absolute temperature, °K P = partial pressure of the condensing gas in the capillary or pore, mmHg Po= normal vapor pressure of the liquid, mmHg. This equation provides the basis for estimating pore sizes and their distribution by nitrogen desorption.16 To account for hysteresis in adsorption isotherms for Page 3 of 6

porous solids, Cohan proposed that on condensation (adsorption) the pores fill radially, but that for evaporation (desorption) the pores empty vertically.17 Although the nature of hysteresis is not yet fully resolved, some data quantitatively support this mechanism.18 Assuming that this mechanism applies to sulfur condensation in porous alumina (the principal Claus catalyst), then d =

-2gV RT ln (P/Po)

(6)

Eq. 6 applies to pores whose diameters are at least four times the dimension of the liquid molecule, yet also sufficiently small that a curved meniscus forms in the pore. For sulfur condensation, Eq. 6 applies to pores which are at least 15 Å in diameter. Although the upper limit of applicability is more difficult to identify, it probably exceeds the size of pores having significant catalytic activity. Liquid sulfur has a curved meniscus in glass capillaries with diameters larger than 0.1 cm.19 A lower limit of applicability of Eq. 6 does not imply that reactants will still have access to micropores. Sulfur molecules cannot condense in micropores in a macroscopic sense, but they may “strongly adsorb” in them. Steijns and Mars present evidence suggesting that sulfur does “strongly adsorb” in micropores in a variety of porous solids.20 Additionally, sulfur condensation in pores having diameters greater than roughly 15 Å may prevent reactants from getting into micropores if the only access to micropores is via slightly larger pores. Hence pores with diameters smaller than that given by Eq. 6 will probably not be catalytically active since they will probably not be accessible to reactants.

PROPERTIES OF SULFUR

Strictly speaking, Eq. 6 should be applied to each sulfur allotrope. However, because necessary data are not available, sulfur will be considered as a lumped component. The molar volume, V, may be written where

V = M/r

(7)

corresponding to S1.22 Furthermore, this is conservative; selecting a higher molecular weight allotrope predicts a more severe pore plugging problem. West has measured the density of liquid sulfur as a function of temperature,23 the data may be expressed analytically as r = 2.0363 - 6.0137 x 10 -4 T for 433°K < T < 633°K

(8)

Kellas19 and Fanelli24 have measured the surface tension of liquid sulfur as a function of temperature. Their data agree within 3.5%, and may be written as g = 81.16 - 0.0566T for 433°K < T < 720°K

(9)

Fanelli also noted that the surface tension of sulfur saturated with H2S does not differ from that of pure sulfur. The Antoine Equation with the proper constants gives the vapor pressure of pure sulfur as a function of temperature23 log10 (Po) = 6.84359 - 2500.12/(T - 86.85)

(10)

Although impurities may alter, the vapor pressure of sulfur, Eq. 10 will be used in the absence of more accurate data. If a Claus converter operates at a temperature T which is D°K above the sulfur dew point temperature, then the partial pressure of gas-phase sulfur may be approximated by P = Ps = Po evaluated at (T - D) (11) where Ps = partial pressure of gas-phase sulfur, mmHg D = a safety-margin parameter, indicating the difference between the catalytic converter temperature and the sulfur dew point temperature, °C or °K It is implicitly assumed that the distribution of gasphase sulfur molecules is an equilibrium distribution. Alumina very effectively catalyzes the reactions which lead to such an equilibrium distribution.26 Substituting Eqs. 7 through 11 into Eq. 6 gives estimates of the extent of pore plugging in Claus catalyst.

M = molecular weight of liquid sulfur, 32 g/mole r = density of liquid sulfur, g/cc

Although liquid sulfur contains S1 through S8 and also larger chains and rings,21 the thermodynamically appropriate molecular weight of liquid sulfur is 32 g/mole, Page 4 of 6

USA/6020-R01/0797

Maximum pore diameter that plugs with liquid sulfur, Å

80 70

= 2.5°C

60 50 40 30 20 10

= 5.0°C

Figure. 3—Sulfur condensation in Claus catalyst pores. Note: D is the difference between the catalytic converter temperature and the sulfur dew point temperature.

= 7.5°C = 10.0°C

0 160 180 200 220 240 260 280 300 320 340 360 Temperature, °C

ESTIMATES AND DISCUSSION

Fig. 3 shows the maximum Claus catalyst pore diameters that will plug with sulfur as functions of catalytic converter temperature and the safety-margin D, the difference between the catalytic converter temperature and the sulfur dew point temperature. Pores with diameters smaller than those given by Fig. 3 will not exhibit significant catalytic activity. Fig. 3 also indicates that for a Claus plant which operates on a policy of a fixed safety-margin, sulfur condensation tends to be less troublesome in successive reactors. This results primarily from the vapor-liquid equilibrium behavior of sulfur. The pore size distributions of many commercially available Claus catalysts are centered around a diameter of approximately 40Å.27-29 Based on the results shown in Fig. 3, significant pore plugging and loss of catalytic activity should occur for operation within 5° to 10°C of the sulfur dew point temperature. This estimate agrees with what has been observed in commercial practice.11,30 In the presence of sulfur condensation, a Claus catalyst with more small pores will be less active than a Claus catalyst with fewer small pores. As the catalytic converter temperature approaches the sulfur dew point temperature, a catalyst with more small pores loses more surface area because more of its pores will be plugged with sulfur, and also because small pores have more surface area relative to large pores. Hence a Claus catalyst with fewer small pores should retain more surface area and therefore better activity at lower temperatures, which are more

USA/6020-R01/0797

thermodynamically favorable for H2S and SO2 conversion. Total surface area of a fresh Claus catalyst is not a true measure of catalyst activity; additionally, pore size distribution may be an equally important parameter. Experiments might provide firmer estimates of these effects and lead to the development of improved Claus catalysts. NOMENCLATURE a, n - Fixed variables d - Diameter of the capillary or pore containing condensed liquid, cm M - Molecular weight, g/mole P - Partial pressure of the condensing gas in the capillary or pore, mmHg Po - Normal vapor pressure of the liquid, mmHg Ps - Partial pressure of gas-phase sulfur, mmHg R - Universal gas constant, 8.314 x 107 erg/mole/°K T - Catalytic converter temperature, °K V - Molar volume of the liquid, cc/mole q - Contact angle, degrees g - Surface tension of the liquid, dyne/cm r - Density of the liquid, g/cc. D - The difference between the catalytic converter temperature and the sulfur dew point temperature, °K

Page 5 of 6

LITERATURE CITED 1 Wolf, E.E. and Alfani, F., Catalysis Reviews, Vol. 24, 1982, p. 329 and references cited therein. 2 Richardson, J.T., Industrial and Engineering Chemistry, Process Design & Development, Vol. 11, 1972, pp.12, 18. 3 Stanulonis, J.J., Gates ,B.C. and Olson, J.H., AIChE Journal, Vol. 22, 1976, p. 576. 4 Newson, E., Industrial and Engineering Chemistry, Process Design & Development, Vol. 14, 1975, p. 27. 5 Bartholomew, C.H., Catalysis Review, Vol. 24, 1982, p. 67 and references cited therein. 6 Chen, H.C. and Anderson, R.B., Journal of Catalysis, Vol. 28, 1973, p. 161. 7 Bomback, J.L., et al, Environmental Science and Technology, Vol. 9, 1975, p. 139. 8 Chou, T.S., and Hegedus, L.L., AIChE Journal, Vol. 24, 1978, p. 255. 9 Grancher, P., Hydrocarbon Processing, Vol. 57, No. 7, July 1978, p. 155. 10 Baglio, J.A., et al, Industrial and Engineering Chemistry, Product Research & Development, Vol. 21, 1982, p. 408. 11 Lieberman, N.P., Oil & Gas Journal, Vol. 78, No. 11, March 17, 1980, p. 155. 12 Maadah, A.G., PhD thesis, Oklahoma State University, 1978. 13 Gamson, B.W. and Elkins, R.H., Chemical Engineering Progress, Vol. 49, No. 4, April 1953, p. 203. 14 “Activated Alumina for Claus Catalysis,” Aluminum Company of America, Pittsburgh, PA, May 1977. 15 Lord Kelvin, Proceedings of the Royal Society of Edinburgh, Vol. 7, 1870, p. 63. 16 Smith, J.M., “Chemical Engineering Kinetics,” McGraw-Hill, New York, NY, 1981, Ch. 8.

The author Gregory R. Schoofs is a graduate student at Princeton University working toward a PhD in chemical engineering. After receiving bachelors’ degrees in 1980 in chemical engineering and economics from the University of California, Berkeley, he worked briefly for Chevron Research Co. His research involves behavior of molecules on solid surfaces with emphasis on adsorption and catalysis. He is a member of the AIChE and ACS.

Cohan, L.H., Journal of the American Chemical Society, Vol. 60, 1938, p. 433. 18 Cohan, L.H., Journal of the American Chemical Society, Vol. 66, 1944, p. 98. 19 Kellas, A.M., Journal of the Chemical Society, London, Vol. 113, 1918, p. 103. 20 Steijns, M. and Mars, P., Industrial and Engineering Chemistry, Product Research & Development, Vol. 16, 1977, p.35. 21 Meyer, B., Chemical Reviews, Vol. 76, 1976, p. 367. 22 Stull, D.R. and Prophet, H., “JANAF Thermochemical Tables,” 2nd ed., NSRDS-NBS-37, U.S. Department of Commerce, National Bureau of Standards, Office of Standard Reference Data, 1971. 23 West. J.R., Industrial and Engineering Chemistry, Vol. 42, 1950, p. 713. 24 Fanelli, R., Journal of the American Chemical Society, Vol. 72, 1950, p. 4016. 25 ”Lange’s Handbook of Chemistry,” 12th ed., John A. Dean, ed. McGraw-Hill, New York, NY, 1979, Section 10. 26 Berkowitz, J., in “Elemental Sulfur,” Beat Meyer, ed., Wiley Interscience, New York, NY, 1965, Ch. 7. 27 Burns, R.A., Lippert, R.B., and Kerr, R.K., Hydrocarbon Processing, Vol. 53, No. 11, Nov. 1974, p. 181. 28 Data sheets, Rhône-Poulenc Industries, Paris, France. 29 Norman, W. S., Oil & Gas Journal, Vol. 74, No. 46, Nov. 15, 1976, p. 55. 17

a

United States of America Almatis AC, Inc. Adsorbents & Catalysts 109 Highway 131, Vidalia, LA 71373 USA Phone: 800-533-4511 or 318-336-9601 • Fax: 318-336-9922 Website: www.almatis.com Email: [email protected] Europe Frankfurt, Germany Asia Tokyo, Japan

Page 6 of 6

USA/6020-R01/0404