Kinetic Model for Wet Oxidation of Organic Compounds in Subcritical

Chapter 24

Supercritical Fluid Engineering Science Downloaded from pubs.acs.org by UNIV OF CALIFORNIA SAN DIEGO on 04/12/16. For personal use only.

Kinetic Model for Wet Oxidation of Organic Compounds in Subcritical and Supercritical Water 1

1,3

2

Lixiong Li , Peishi Chen , and Earnest F. Gloyna 1

Separations Research Program, Center for Energy Studies, The University of Texas at Austin, Austin, TX 78758 Environmental and Water Resources Engineering, The University of Texas at Austin, Austin, TX 78712

2

The methodology of developing a global kinetic model for wet oxidation (WO) of organic compounds is described. Simplified reaction schemes considering rate-controlling intermediates are used in the model development. The selectivity of product vs. intermediate formations is defined as a key model parameter to simplify model calculations and characterize the "strength" of the feed stream. This model has been previously validated for hydrocarbons and oxygenated hydrocarbons using reported WO kinetic data in the temperature and pressure ranges, respectively, from 150°C to 550°C and from 20 bar to 440 bar. The model formulation and kinetic parameters for nitrogen- and chlorine-containing organic compounds are further discussed to demonstrate the model adaptability for WO of a variety of organic compounds, wastewaters and sludges in both subcritical and supercritical water. Wet oxidation (WO), such as, the Zimpro wet air oxidation ( W A O ) process (1-2), has been commercially used in the treatment o f wastewaters and sludges for at least three decades. Early W O processes have been typically operated in a temperature range of 150°C to 350°C and pressure range of 20 to 200 bar. The operating pressures have been maintained well above the saturation pressures corresponding to the operating temperatures, so that the reactions have occurred in the liquid phase. Residence times have ranged from 15 min to 120 min, and the chemical oxygen demand ( C O D ) removal typically have been about 75% to 90%. Volatile acids and alcohols have constituted a substantial portion o f the remaining C O D (3-11). Unfortunately, incomplete (partial) W O o f some wastewaters has resulted in colored and toxic effluents (12-13). Supercritical water oxidation ( S C W O ) can overcome the shortcomings o f W A O . The complete miscibility o f oxygen and organic compounds in supercritical water creates a single-phase fluid and favorable reaction environment. In supercritical water, high destruction efficiency (> 99.99%) o f many complex organic compounds and E P A priority pollutants can be achieved with a short residence time (< 5 min) and 3

Current address: Phillip Morris R&D, P.O. Box 26583, Richmond, VA 23261

0097-6156/93/0514-0305S06.00/0 © 1993 American Chemical Society


306

SUPERCRITICAL FLUID ENGINEERING SCIENCE

the reaction can be accomplished in a totally enclosed facility (14-23). Therefore, SCWO provides a technically viable and environmentally attractive option to managing the growing organic sludge and toxic wastewater treatment problems. However, W O reaction mechanisms detailing elementary reactions of complex organic compounds and organic mixtures (such as wastewaters and sludges) appear to be tedious. Data describing these mechanisms are only available for a few simple compounds. Furthermore, the values of the reported kinetic parameters may differ considerably. Therefore, general and reliable kinetic models are needed for the development of W O processes. Such a generalized kinetic model has been developed for W O of hydrocarbons and oxygenated hydrocarbons (24). The model has been validated using reported W O kinetic data in the temperature and pressure ranges, respectively, from 150°C to 5 5 0 ° C and from 20 bar to 440 bar. This generalized kinetic model can be used to describe W O of nitrogen- and chlorine-containing organic compounds. It is the objective of this paper to discuss the methodology of this model development. Existing Global Kinetic Models Global kinetic models have been widely reported and successfully used within the stated experimental conditions to describe W O of organic model compounds, wastewaters, and sludges. A kinetic database has been initiated (24). Additional references have been summarized in Table I. Most subcritical W O of organic compounds and mixtures can be described by a first-order reaction with respect to the organic reactants and first-order reaction with respect to oxygen. An oxygen mass-transfer factor can be used in global rate equations to account for the nonzero-order oxygen dependency (9,31-33). For supercritical W O , the oxidation rates appear to be first-order and zero-order with respect to the reactant and oxygen concentrations, respectively. In most W O experiments there is no effort made to differentiate the effect of side reactions. Depending upon reaction conditions and reactants involved, the rate of side reactions, such as hydrolysis (34-36), pyrolysis (35), dehydration (36-37), and liquefaction (38) varies considerably. The kinetic models discussed in this paper, unless otherwise specified, describe the overall reaction which may include oxidation, hydrolysis, pyrolysis and dehydration steps. Pressure is another factor which can affect the oxidation rate in supercritical water. At a given temperature, pressure variation directly affects water density, and in turn changes the reactant concentration. Furthermore, the properties of water are strong functions of temperature and pressure near its critical point Due to limited information, the effect of pressure is not included in this paper. A generalized W O kinetic model is available (24). The model is derived from a simplified reaction scheme involving the formation and destruction of rate-controlling intermediates. The assumption is made that as W O reactions proceed, the initial compounds are converted to (a) oxidation end products, (b) unstable intermediates, and (c) relatively stable intermediates. Conversion of all intermediates to the oxidation end products requires a number of parallel and consecutive reactions. Quantification of the W O reaction mixtures requires the treatment similar to that used in the kinetic lumping technique (39-41). Since the global W O rate depends on the end product formation rate, as well as the formation and destruction rate of stable intermediates, it is necessary to identify key rate-controlling reaction intermediates. Comparison of kinetic parameters for different organic compounds can assist in identifying rate-controlling components. For example, the activation energies for subcritical W O of volatile acids are higher as compared to those found for organic compounds possessing a lower oxygen content. Acetic acid exhibits the highest activation energy, 167.7 kJ/mol (3). When the same data are fitted using a pseudo-first-order model, the activation energy increases to 182

02

2

2

0

0

0

o

Phenol

Phenol (COD)

Phenol (Total phenolics)

Pyridine n

a

(29)

0.2 0.5

0

1

91.5

-

batch

m

1.5

0.5

57.5

-

a

240-350

0.3-1

1.5-4.5 0.01-0.05

353-383 673-773

4-15

4-15

1

1 413-453

423-468

0.5

1

93

48

0

(29)

0.1

1

batch

1.59Χ10

(28)

0.128

26-48

300-500 418-453

0.5

batch

batch

(27)

0.128

70-140

833-903

1

flow

4

(27)

-

245

833-903

*, Kinetic parameters are defined by -d[C]/dt = k [ C ] [ 0 ] and k = k°exp(-E /RT), where [C] and [O] are concentrations of oragnic reactants and oxidant, respectively; E is in kJ/mol; Τ is in K; R = 8.314 J/mol-K; and k° = 1/sec (first-order), liter/mol-sec (second-order), etc.. -, Not reported. [ C ] = feed concentration The concentration of compounds labeled with C O D is quantified by chemical oxygen demand method; and concentration of other compounds is quantified by chromatographic techniques. The excess oxidants are used in all tests. Kinetic parameters are reported for the overall reaction in water.

2

2

(26)

-

245

673-773

70-140

(23)

0.3-1

240-350

423-483

300-500

(25)

2-20

(23)

(30)

(26)

(3)

-30 33-35

20-200

References

543-593

0

[C] (g/L)

Temperature Pressure (Κ) (arm)

56.6

1.98xl0

batch 7

1

Phenol

112

02

Phenol

0.5

1

batch

89.5

02

m-Xylene(COD)

-

02

0.66 0

0.99 1

2.04xl0

flow

141.7

0

1

178.9

0.3-0.6

0

1.5

1

7

2.51xlO

flow

Methane

2

0

Methane

28.5

45.3

182

n

9.0x10*

batch

12

02

3

4.40xl0

2,4-DichloiOphenol

batch

3xl0

2

Kinetic Parameters* k° Ea m η

batch

0

Oxidant Reactor Type

Alcoh. Dist. Waste (COD) 02

Acetic Acid

Compounds

Table I. Global Kinetic Models for Wet Oxidation of Organic Compounds



308

kJ/mol, Table I. The refractory nature of acetic acid and its validity as a key W O intermediate are elaborated elsewhere (24). Simplified W O Reaction Schemes


The previously proposed reaction scheme for W O of organic compounds can be illustrated as follows:

ki A + O2

—

B+

C

O2

where C = [Oxidation end products]; Β = [Rate-controlling intermediate]; and A = [Initial and intermediate organic compounds other than B]. Application of this generalized reaction scheme requires the définition of Group A and identification of Group B. Three classifications of organic compounds which are characteristic of many organic waste streams, are defined as follows: Category I - hydrocarbons and oxygenated hydrocarbons; Category Π - nitrogen-containing organic compounds; and Category ΙΠ - chlorinated organic compounds. Both categories Π and ΠΙ may or may not contain oxygen. For category I organic compounds, acetic acid is considered to be the key ratecontrolling intermediate (24). In this case, the reaction scheme becomes,

ki CmHuOr+ ρθ2 \ t a

mC02 + (η/2)Η2θ

k^X qCH3COOH + q02

where C H O is a collective term including the initial compounds and unstable reaction intermediates. Using the same notations given in the generalized reaction scheme, three groups of organic substances can be further specified. Group A includes the initial organic compounds and all unstable intermediates except acetic acid, Group Β contains the refractory intermediates represented by acetic acid, and Group C is designated as the oxidation end products (carbon dioxide and water). The concentrations of Groups A and Β may be expressed in terms of total organic carbon (TOC), chemical oxygen demand (COD), total oxygen demand (TOD), or biochemical oxygen demand (BOD). The concentration units for these measurements are usually expressed in mass per unit volume. The mass concentration measurements are particularly useful and convenient for quantifying organic contents in wastewaters and sludges. For nitrogen-containing organic compounds (category II), nitrogen gas is confirmed to be the predominant SCWO end product, regardless of the oxidation state of nitrogen in the starting material (42-44). This observation agrees with thermodynamic and kinetic calculations. The formation of ammonia and nitrous oxide has been reported in S C W O of various nitrogen-containing organic compounds and wastewaters (17,42-43,45). Ammonia is usually a hydrolysis product of nitrogencontaining organic compounds (17,36), and nitrous oxide is a partial oxidation product m

n

r


24.

Kinetic Model for Wet Oxidation of Organic Compounds

LI ET AL.

of ammonia. At higher temperatures, 560°C to 670*C, the formation of nitrous oxide is more favorable than the formation of ammonia (43-44). A t low temperatures near 400°C, nitrogen remains in solution as ammonia or ammonium ion (10,44,46). The activation energy for SCWO of ammonia has been reported as 156.8 kJ/mol (47), but the activation energy for nitrous oxide is not known. For low temperature applications, about 4 0 0 ° C , nitrogen conversion rate may depend on ammonia formation and destruction rates. Therefore, a simplified W O reaction scheme for nitrogen-containing organic compounds can be written as, sNH3 + t02

CmNoHriOr+

ρθ2

^

yN2 + mC02+ χ Η 2 θ

qCH3COOH + q02 where C N H O is a collective term including the initial organic compounds and unstable intermediates with or without nitrogen-containing functional groups. If Group A represents the compounds containing primarily the trinegative state nitrogen, then organic nitrogen concentrations can be monitored using the Kjeldahl nitrogen and ammonia nitrogen methods. A similar reaction scheme can be shown for chlorinated (category ΙΠ) organic compounds. m

0

n

r

M

q'CH3Cl+ q 02

V

ki CmClsHnOr+pCfe

V

V

S

S

>

V

mC02 + χ Η 2 θ + sHCl

qCH3COOH + q02 Methyl chloride is a representative reaction intermediate among short-chain chlorinated hydrocarbons which are likely to be produced. The abundance of water in the W O environment enhances fast hydrolysis of these intermediates to methanol and ethanol. Therefore, in addition to acetic acid, methanol and ethanol may be used as ratecontrolling intermediates for W O of chlorinated organic compounds. Finally, the key rate-controlling intermediates identified for use in the generalized model and the reported kinetic parameters for these compounds are summarized in Table II. Results and Discussion The discussed kinetic model is generally suitable to describe W O of various organic compounds under isothermal, well-stirred batch reactor or constant volumetric flow rate plug-flow reactor conditions. As an example, the model expression is given for supercritical W O of category I organic compounds. The reaction rate may be assumed to be first-order with respect to the reactants, and zero-order with respect to oxygen. The concentrations of A and Β can be expressed as, +

[Α] = [ Α ] β " ^ ι ^ 0

1

(D

310


Table II. Kinetic Parameters for Key Rate-Controlling Intermediates


Organic Key Oxidation Compound mterrnediate End Category (Alternative) Product

Œ3COOH

I

π

NH3 (N 0)

Kinetic Parameters* Condition k E (Water) (1/sec) (kJ/mol)

C02,H 0 2

N ,H 0 2

2

2

m

CH3C1 (Œ3OH)

Reference

a

HC1, H 0

12

Subcritical Supercritical Supercritical*

4.40xl0 2.55xlO 2.63xl0

Supercritical Not available

3.16x106

157

-

-

10

182 172.7 167.1

(3) (14) (34)

Λ

(48)

-

-

Not available Supercritical

2

n

2.51Χ10

24

(49)

395

* Pseudo-first-order reaction model using oxygen. Obtained from treating the reported 7 data points. Using hydrogen peroxide (obtained from treating 57 data points). Λ

#

where [ A ] and [ B ] are the concentrations of A and Β at time = 0, respectively. Combining equations 1 and 2 produces equation 3: 0

0

[A + B] _ [A + B ]

0

[A]

/

0

k

[ A ] + [B]olki+k2-k 0

, (ki-k )

2

3

ki+k -k

3

2

c

3

. ^ , | , /

[Bio [A] +[B] 0

^ (3).

0

If [B] = 0, equation 3 can be further simplified to equation 4. 0

[A + B] _ [A + B ]

0

k

2

ki+k -k 2

c

.w

3

+

jM53L -(k k )t e

ki+k -k 2

1+

2

(4)

3

The three rate constants, k i , k and k , require further description. If k is much smaller than k i , the organic compounds may be oxidized more easily to the end products. If k becomes larger, more acetic acid will be formed. Values o f k i may be determined from the initial reaction rate data based on lumped parameters, such as, C O D and T O C ; because Group A is a collective term encompassing the initial compounds as well as unstable intermediates. Conversion of Group A to acetic acid does not change effluent C O D or T O C concentration significantly. Therefore the kinetic parameters based on C O D or T O C are directly related to k i . The kinetic parameters for acetic acid given in Table Π can be used for k3. Values of k may be derived from further treatment of equation 4. Since the W O of Group A can be considered as a parallel reaction system, the 2

3

2

2

2

point selectivity, α ϊ , is defined as the formation rate of acetic acid to that of carbon dioxide from Group A . If first-order reactions are assumed, the following relation may be obtained for oci,

24.


1

LI ET AL.

Π

k l [ A

]

Kinetic Model for Wet Oxidation of Organic Compounds

ki

£ f

311

(

5

)

The global reaction rate depends on activation energy levels for the three step reactions. Since acetic acid is a refractory intermediate, Ea3 is greater than either Eai or Ea2. For most organic wastes, Ea2 should approximately be equal to E a i ; because Groups Β and C represent series degradation products all derived from similar reactions. For this reason, ατ should be a weak function of temperature, and merely become a ratio of the frequency factors of the two parallel reactions. The value of α ϊ for a feed stream matching the above assumptions falls in between 0 and 1 (24). If the concentration of short-chain organic compounds, including those with two carbon atoms, other than acetic acid, such as ethanol, is high in a wastewater, the value of α ϊ may be large. Therefore, oci can be used to characterize the "strength" of the waste stream in a W O process. Two approaches can be used to apply this model for nitrogen- and chlorinecontaining organic compounds. The first approach is to express the concentrations of Groups A and Β in terms of C O D . The concentrations of ammonia and methyl chloride can be expressed by equation 2 when k2 and k3 are substituted by k4 and ks, and k$ and k7, respectively. A general expression may then be obtained for [A + B i + n + BniL where the subscripts I, II, and ΙΠ indicate the organic compound category as previously defined. Similarly, typical values of α may be obtained from experimental data for categories II and ΙΠ. The second approach is to monitor the concentrations of total carbon, total nitrogen, and total chlorine. It can be assumed that the oxidation kinetics for each species of C, N , or CI is independent of those of other species. Equation 3 can be used for each species with proper substitution of reaction rate constants. The values of k i , k2, k4, and k6 can be determined from experimental data. If experimental results are not available, an approximation may be established by selecting a waste from the existing kinetic database. It is recommended that the parameters shown in Table Π be utilized as k3, ks, and Icj. B

Conclusions A previously described global kinetic model can accommodate the W O of nitrogen- and chlorine-containing organic compounds. A similar approach may be used to describe the W O of organic compounds containing other heteroatoms. Such a model may be derived from a simplified reaction scheme based on parallel reaction pathways involving the formation of key intermediates and oxidation end products. This selectivity factor, quantified by the ratio of the parallel reaction rates, can be used to characterize the "strength" of the feed stream in W O processes. A high point selectivity value indicates that feed is more difficult to treat because a large fraction of the feed is converted to the rate-controlling intermediate. This work extends the validity of the previously proposed kinetic model for W O of organic compounds associated with wastewaters and sludges. Acknowledgments Appreciation for financial support is extended to Separations Research Program, The University of Texas (Austin, TX) and RPC Energy Services, Inc.Also, this material is based in part upon work supported by the Texas Governor's Energy Research in Applications Program under Contract No. 5127.

312


Literature Cited


1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

Zimmerman, E. J. Chem. Eng. 1958, 65, 117. Teletzke, G. H. Chem. Eng. Progress, 1964, 60, 33. Foussard, J.-N.; Debellefontaine, H.; Besombes-Vailhe, J. J. Environ. Eng., 1989, 115, 367. Keen, R.; Baillod, C. R. Water Res., 1985, 19, 767. Baillod, C. R.; Faith, Β. M.; Masi, O. Environ. Prog., 1982, 1, 217. Baillod, C. R.; Lamparter, R. Α.; Leddy, D. C. Proceedings 34th Industrial Waste Conference, Ann Arbor Science Publ: Ann Arbor, MI, 1980; 206-213. Friedman, Α. Α.; Smith, J. E.; DeSantis, J.; Ptak, T.; Ganley, R. C. J. WPCF 1988, 60, 1971. Wu, Y. C.; Hao, O. J.; Olmstead, D. G. J. WPCF 1987, 59, 39. Ploos van Amstel, J. J. Α.; Rietema, K. Chem. Ing. Tech., 1973, 45, 1205. Fisher, W. J. Water Res. 1971, 5, 187. Teletzke, G. H.; Gitchel, W. B.; Diddams, D. G.; Hoffman, C. A. J. WPCF 1967, 39, 994. Sheppard, J. D. Can. J. Chem. Eng. 1985, 63, 645. Larson, R. Α.; Ju, H.-L.; Snoeyink, V. L.; Recktenwalt, Μ. Α.; Dowd, R. A. Wat. Res. 1988, 22, 337. Wightman, T. J. M.S. Thesis ; Univ. of Cal.: Berkeley, CA, 1981. Price, C. M. M.S. Thesis, MIT: Cambridge, MA, 1981. Modell, M . ; Gaudet, G. G.; Simson, M . ; Hong, G. T.; Biemann, K. "Supercritical Water Testing Reveals New Process Holds Promise," Solid Wastes Manag., 1982, 25, 26. Timberlake, S. H.; Hong, G. T.; Simson, M.; Modell, M . S.A.E. Tech. Paper Ser., No. 82872, 1982. Thomason, T. B.; Modell, M. Hazardous Waste, 1984, 1, 453. Helling, R. K. Ph.D. Dissertation, MIT: Cambridge, MA, 1986. Staszak, C. N.; Malinowski, K. C.; Killilea, W. R. Envir. Progress, 1987, 6, 39. Li, L.; Eller, J. M.; Gloyna, E. F. presented at AIChE Summer Meeting, San Diego, CA, 1990. Lee, D. S.; Gloyna, E. F.; Li, L. J. Supercrit. Fluid, 1990, 3, 249. Lee, D. S.; Kanthasamy, Α.; Gloyna, E. F. presented at AIChE Annual Meeting, Los Angeles, CA, 1991. Li, L.; Chen, P. S.; Gloyna, E. F. AIChE J, 1991, 37, 1687. Daga, N. S.; Prasad, C. V. S.; Joshi, J. B. Ind. Chem. Eng., 1986, 28, 22. Webley, P. Α.; Tester, J.W. Energy and Fuels (in press). Willms, R. S.; Balinsky, A. M.; Reible, D. D.; Wetzel, D. M.; Harrison, D. P. Ind. Eng. Chem. Res. 1987, 26, 148. Jaulin, L.; Chornet, E. Can. J. Chem. Eng., 1987, 65,64. Mundale, V. D.; Joglekar, H. S.; Kalam, Α.; Joshi, J. B. Can. J. Chem. Eng., 1991, 69, 1149. Kulkarni, U. S.; Dixit, S. G. Ind. Eng. Chem. Res., 1991, 30, 1916. Charest, F.; Chornet, E. Can. J. Chem. Eng., 1976, 54, 190. Pruden, Β. B.; Le, H. Can. J. Chem. Eng., 1976, 54, 319. Joglekar, H. S.; Samant, S.D.; Joshi, J.B. Wat. Res. 1991, 25, 135. Lee, D. S. Ph.D. Dissertation, The Univ. of Texas: Austin, TX, 1990. Huppert, G. L.; Wu, B.C.; Townsend, S.H.; Klein, M.T.; Paspeck, S.C. Ind. Eng. Chem. Res. 1989, 28, 161. Tiffany, D. M.; Houser, T.J.; McCarville, M.E.; Houghton, M.E. Am. Chem. Soc., Div. Fuel Chem., 1984, 29, 56. Xu, X.; De Almeida, C. P.; Antal, Jr. M . J. Ind. Eng. Chem. Res. 1991, 30, 1478.

24. LI ET AL. Kinetic Model for Wet Oxidation of Organic Compounds


38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

313

Mikita, Μ. Α.; Fish, Η. T. ACS Div. Fuel Chem. Prep. 1986, 31(4), 56. Hutchinson, P.; Luss, D. Chem. Eng. J. 1970, 1, 129. Astarita, G. AIChE J. 1989, 35, 529. Aris, R. AIChE J. 1989, 35, 539. Killilea, W. R.; Swallow, K. C.; Hong, G. T. J. Supcritical Fluids (in press). Hong, G. T.; Fowler, P. K.; Killilea, W. R.; Swallow, K. C. SAE Technical Paper Series No. 871444, 1987. Takahashi, Y.; Wydeven, T.; Koo, C. NASA Conf. Publ. #10040, 95-106, 1989. Shanableh, A. M. Ph.D. Dissertation, The Univ. of Texas: Austin, TX, 1990. Gamer, C. R. Iron and Steel Engineer, 1986, 55. Webley, P. Α.; Tester, J. W.; Holgate, H. R. Ind. Eng. Chem. Res. 1991, 30, 1745. Webley, P. Α.; Holgate, H. R.; Stevenson, D. M.; Tester, J. W. SAE Technical Paper Series No. 901333, 1990. Rofer, C. K.; Streit, G. E. Oxidation of Hydrocarbons and Oxyenates in Supercritical Water, Phase II Final Report, LA-11700-MS, DOE/HWP-90, Los Alamos National Laboratory, 1989.

RECEIVED April 27, 1992

Kinetic Model for Wet Oxidation of Organic Compounds in Subcritical

Recommend Documents