000
Ind. Eng. Chem. process
ms. mv. 1904, 23,aoa-aw
u = superficial gas velocity Subscripts F = gas P = particle Superscripts I = input I1 = output (response) Greek Letters a, =
axial gas thermal dispersion coefficient
= bed void fraction I.L = gas viscosity p = density e
Literature Cited Andrleu, J.; Smith, J. M. Chem. Eng. J . 1980, 20, 211. AsbJornsen, 0. A.; Amundsen, K. Chem. €ng. Sd. 1970, 25, 943. Asbpnsen, 0.A.; Wang, B. Chem. Eng. Scl. 1971, 26, 585. Bradshaw. A. V.; Johnson, A.; McLachlan, N. H.; Chiu, Y. T. Trans. Inst. Chem. Eng. 1970, 48, T77. Dayton, R. W.; Fawcett, S.L.; (Llmble, R. E.; Seelander, C. E. Batelle Memorial Institute, Columbus, OH. 1952; Report EMI-747.
De Acetls, J.; Thodos, G. Ind. Eng. Chem. 1960, 52, 1003. Eichorn, J.; White, R. R. Chem. Eng. h o g . Symp. Ser. 1952. 48(4), 11. Gamson, B. W.; Thodos, G.;Hougen, 0. A. Trans. Am. Inst. Chem. Eng. 1943, 39, 1. Glaser, M. 6.; Thodos, G. AIChE J. 1958, 4 , 63. Gunn, 0. J.; Pryce, C. Trans. Inst. Chem. Eng. 1969, 47, T341. Kaguei, S.; Shiorawa, 6.; Wakao, N. Chem. Eng. Sci. 1977, 3 2 , 507. Kunii, D.; Smith, J. M. AIChf J. 1960, 6 , 71. Kunii, D.; Smith, J. M. AIChP J . 1961, 7 , 29. Littman, H.; Barile, R. G.: Pulsifer, A. H. Ind. Eng. Chem. Fundam. 1968, 7 , 554. Meek, R. M. G. International Heat Transfer Congress, ASME, New York, 1961, p 770. Pulsifer, A. H. Ph.D. Dissertation, Syracuse University, 1965. Schlunder, E. U. “Elnfuhrung In die Warme- und Stoffubertragung”, 2 Aufl.; Vieweg-Verlag: Braunschwelg, 1975; p 75. Shearer, C. J. Nat. Eng. Lab. East Kilbridge, Glasgow, N.E.L. Rept. 1962; 55. Turner, G.A.; Otten, L. Ind. €ng. Chem. Process Des. Dev. 1973, 12, 417. Wakao, N. Chemlcal Engineering Department, Yokohama National University Japan, Personal Contact, 1981. Wakao, N. Chem. Eng . Sci. 1976, 31, 1115. Wakao, N.; Kaguel, S.; Funazkrl, T. Chem. Eng. Sci. 1979, 3 4 , 325. Wilke, C. R.: Hougen, 0. A. Trans. Am. Inst. Chem. Eng. 1945, 4 1 , 445.
Received f o r review November 4, 1982 Revised manuscript received December 5 , 1983 Accepted January 26, 1984
Flue Gas DenHrMcation. Selectlwe Cataiytic Oxidation of NO to NOz Hans T. Karkront and Harvey S. Rosenberg’ Battelk Memoriel Institute, Columbus Laboratmles, Columbus, Ohio 4320 1
A study was conducted on the selective catalytic oxidation of NO in simulated flue gas. Oxidation of NO to the more reactive &Os and N204(NO,) would enable wet scrubbing of NO, simultaneously with SO,. Fourteen catalysts were tested in a fixed-bed reactor at a space velocity of at least 1500 h-’ in the temperature range of 150 to 800 OF. The experimental results comprise a screening program for an engineering analysis rather than a full kinetic study. Eight catalysts exhibited optknwn performance at 200 O F , a temperatwe which is very suitable for an add-on process at a 00aCR.d power plant. Six catalysts yielded greater than 50% oxidation at this temperature. However, deactivation of the catalysts was observed after about 14 h of exposure to the flue gas. The experimental results indicate that it may be possible to solve the deactivation problem by modifying the catalysts.
Introduction Generally speaking, two types of technologies have been considered for NO, control on power plant boilerscombustion modification and flue gas denitrification. The latter technology is required only for very stringent emission regulations. However, flue gas desulfurization (FGD)is required on new power plant boilers. Since the leading FGD processes involve wet scrubbing, there is considerable interest in developing a wet scrubbing process for the simultaneous removal of NO, and SO2. Wet methods for NO, removal are limited by the relatively inert nature of NO. The NO, in flue gas is approximately 90% NO. This difficulty can be overcome by oxidation of NO to the more reactive NOz in the gaseous phase using ozone or C102prior to absorption (oxidation absorption). Ozone is the be& oxidizing agent but it is very expensive. The cost of C102is 30 to 40% lower than that of ozone, but the use of C102 introduces a considerable amount of chloride into the scrubbing liquor, thus causing Energy Technology R&D,Southern Sweden Power Supply,
S-21701 Malmo, Sweden.
waste disposal problems. Ozone and CIOz oxidize NO to NO2 within a 1-s residence time. NO can be oxidized by the oxygen in flue gas, but with low NO concentrations the noncatalytic reaction is very slow and the rate decreases with increasing temperature. For flue gas containing 5% oxygen and 750 ppm NO, a residence time of about 150 min (space velocity of about 0.4 h-l) is required to convert 90% of the NO to NO2 at a temperature of 300 O F . Several oxidation absorption-reduction processes are under development in Japan for the simultaneous removal of NO, and SOz from flue gas. In this type of process, NO is first oxidized to the more reactive NO2 which is then absorbed. Because of the nature of the process chemistry in the liquid phase, usually 3 to 4 mol of SO2 are needed for each mole of NO, for efficient removal of NO,. This mole ratio is not a problem for flue gas from a boiler fired with high-sulfur coal. Most of the absorbed NO, is reduced to nitrogen or NH3 by sulfite ions (absorbed SOz) so that the formation of undesirable nitrites and nitrates is greatly reduced or eliminated. Any NH3 produced must be recovered for use, decomposed, or disposed of after appropriate waste treatment.
0196-4305/84/ 1123-0808$01.50/0 0 1984 American Chemical Society
Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984 809
One of the simplest and potentially least expensive methods for converting NO to NO2is by catalytic oxidation using the oxygen present in the flue gas. It would be desirable to find a catalyst that: (a) oxidizes virtually all of the NO to NO2at a space velocity of lo00 h-' or higher; (b) shows optimum performance in the temperature range of 300 O F down to 150 OF; (c) does not oxidize SOzto SO,; and (d) does not become poisoned by sulfur compounds or other species present in the flue gas. Although the oxidation absorption-reduction processes under development are based on complete oxidation of NO to NO2,it is still possible to remove NO, along with SOz if some of the NO, exists as NO. In this case, NO, is absorbed in the form of N2O3 and N204,rather than Nz04 alone. For efficient removal, the molar ratio of NOz to NO must be at least one. The size of a catalytic reactor is inversely proportional to the space velocity. Therefore, the lower limit of space velocity is determined by practical engineering considerations with regard to reactor size and pressure drop across the reactor. In flue gas processes employing catalytic reactors, the lowest space velocity used is about lo00 h-' and this value can be considered as a guideline for engineering design. Flue gas leaves the air preheater of a coal-fired boiler at a temperature of about 300 O F and usually enters an electrostatic precipitator for particulate removal. Therefore, if the catalytic reactor can be operated at 300 O F or below, catalytic oxidation can be an add-on process and particulates should not have an adverse effect on the catalyst. If an operating temperature between 300 and 700 O F is required, the air preheater would have to be divided into two units with the reactor inserted between the units. For an operating temperature above 700 OF, the reactor would have to be located before the air preheater. Either of these latter two cases would be difficult to retrofit. Also, the particulates would have to be removed ahead of the catalyst or the catalyst would have to tolerate the full particulate loading of the flue gas. An operating temperature below 300 O F can be achieved by spray cooling the flue gas with water. At about 125 O F , the flue gas becomes saturated with water (adiabatic saturation temperature) and cannot be cooled further by evaporation. Therefore, the catalytic reactor should be operated at a minimum temperature of 150 O F to prevent condensation of water on the catalyst. After catalytic oxidation of NO to NO2,the flue gas can be sent to a scrubber for simultaneous removal of NO2 and SO2. The scrubber will cool the flue gas to its adiabatic saturation temperature. If SOz is oxidized to SO3, condensation of sulfuric acid mist may occur in the catalytic reactor and in the ductwork downstream from the reactor and may cause corrosion problems. Large amounts of SO, would also be expected to influence the wet scrubbing process with regard to process chemistry and NO, and sulfur removal efficiency. Absorption of SO, results in formation of sulfates rather than sulfites, and acid mist is very difficult to scrub from a gas stream. The preceding analysis was used to form the basis of a study on the catalytic oxidation of NO. The study began with a comprehensive literature search which w a followed by experimental work. The experimental results comprise a catalyst screening program for an engineering analysis rather than a full kinetic study.
Background A comprehensive literature search was conducted in order to obtain background information on the selective catalytic oxidation of NO to NO2and to identify candidate
Table I. Comparison of Catalysts for Oxidation of NO activity COOS MnO, > Fe20, > C;O
MnOz = COO > none NiO > Cr203 Cu = Cr3+> Fe = co2+ > Ni*+ = MnZ+
-
temp, carrier OF gas comp reference Alz03 480-700 500 ppm NO, Seiji (1975) 200 ppm soz, 4% 02, bal H20 and
NZ
570-750 750 ppm NO, 4% oz, bal Nz zeolite 500-700 300 ppm NO, 6.4% 02, bal Nz
Miyadera et al. (1975) Arai et al. (1977)
catalvsts. The first documented studv on catalvtic gas phase oxidation of NO to NO2 dates back to the eahy 1920's (McKee, 1921). Since then, over 50 experimental investigations have been reported in the literature. Most of the studies are based on the behavior of a *cleann flue gas containing no SO2. The first catalysts investigated consisted of single compounds such as silica gel, activated carbon, and different minerals. Research and development have gradually been focused on more complex catalysts involving metals in the form of elements, oxides, or exchanged (adsorbed on molecular sieves) ions. Over 20 different metals have been investigated in the form of mixtures or on a carrier. It is difficult to compare the various investigations because of the different conditions used in each case. However, a few individual studies have compared the activities of different catalysts (Seiji, 1975; Miyadera et al., 1975; Arai et al., 1977). Table I summarizes the findings of these studies which include mostly the same metallic elements, but on different carriers. The catalytic activity of the various metals is reversed when switching from an alumina to a zeolite carrier. One investigation consisted of test runs in a fixed bed reactor with Cr203 catalyst on eight different carriers (Takayasu et al., 1976). The results clearly showed A1203 carrier to be superior, especially in the presence of SOz and H20. These investigators also studied the effect of the method of catalyst preparation on the performance. Two samples of Crz03catalyst on A1203carrier were prepared, one by impregnation and the other by coprecipitation. The impregnated catalyst exhibited twice as much oxidation of NO as the coprecipitated catalyst, both in the presence and absence of SOz and H20. Most studies indicate that SO2 and H 2 0 suppress the oxidation rate. However, one investigation found that the oxidation increased from 40 to 75% over a Cr3+-zeolite catalyst when 8.3% H 2 0 by volume was added to the flue gas (Arai et al., 1977). Other investigators state that SO2 and H 2 0 poisoning are reversible for Cr203and Mn02 on A1203, respectively (Takayasu et al., 1975, 1976). The statements are based on the observations that the oxidation increased when SO2and/or HzO were removed from the gas stream. The test runs were performed for a period of a few hours. Still others found the poisoning effect to be enhanced at low temperatures (Seiji, 1975; Hattori et al., 1977). A concentration of 500 ppm SO2 was enough to cause a strong deactivation for metal oxides on Al,O, (Takayasu et al., 1976; Hattori et al., 19771, while 200 ppm SOz caused strong deactivation for ion-exchanged zeolites (Arai et al., 1977). One of the studies found that 500 ppm SO2can be tolerated if V205,Sbz03, or Bi203is added to the catalyst (Hattori et al., 1977). Two investigations showed that single metal oxides on Alz03work better than mixtures of oxides on A1203. One
810
Id.Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984
Table 11. Catalysts Selected for Investigation cat. active no. species carrier 1 2 3 4 5 6
Cr3+ Fez+ cu2+ Fe203/MnO/Zn0 Mn02/Pb0 CuO/ MnOz/Sb203
7 8 9
Cr203
Ni0 coo
10 11 12 13 14
Fe303/Mo03 CuO/Mo03 Bi203/Mo03 V20s/Mo03 Pt
size shape Prepared in the Laboratory 13X zeolite 8-12 mesh spheres 13X zeolite 8-12 mesh spheres 13X zeolite 8-12 mesh spheres none 3/16 in. pellets 'la in. irreg particles none none in. irreg particles
A1203 A1203 4 0 3
A1203
A1203 A1203 A1203
-41203
Commercially Available '/ 32. in, pellets in. pellets in. pellets Provided by Catalyst Suppliers granules in. granules in. granules 'I8 in. granules in. spheres 'I4 in.
observed no enhancing effect when some of a Cr203catalyst was replaced by oxides of cobalt, iron, copper, vanadium, or magnesium(Takayasu et al., 1976). In fact, the degree of oxidation decreased by about one-third. The other studied some binary mixtures and found the activity to decrease in the following order: CuO/Mn02 > COO/ CuO > CoO/Mn02 > CoO/Fe203 (Seiji, 1975). The highest rate of conversion with the best mixture, CuO/ Mn02,was 80% while with the best single oxide, COO,it was 90%, with the temperature in the range of 480 to 660 O F in both cases. Two Japanese patents discuss the use of mixed metal oxides without a carrier as oxidation catalysts for NO. One patent describes a process in which NO in flue gas is oxidized to NOz at 400 to 900 O F and a space velocity of 15000 h-l with a catalyst consisting of 52 w t % Fe203, 24 w t % MnO, and 24 w t % ZnO (Kotera et al., 1976). The flue gas is first scrubbed to remove SO2and particulates, reheated, subjected to catalytic oxidation of NO, and finally scrubbed with an alkaline solution to remove greater than 85% of the NO,. The other patent states that a catalyst containing 95 wt % Mn02 and 5 wt % PbO achieved 90% oxidation of NO in "clean" waste gas (no SO2 or particulates) at a temperature of 480 O F and a space velocity of 7500 h-' (Takeyama et al., 1975). Experimental Section Catalysts Investigated. Table I1 lists the catalysts used in this study and summarizes the basis for their selection. The catalysts are divided into three groups. The first group was prepared in the laboratory after selection on the basis of information from the literature search. The second group is commercially available, but it was also selected on the basis of the literature search. The third group was provided by catalyst suppliers in response to a written request for samples to be screened for gas-phase oxidation of NO to NOz. The catalysts prepared in the laboratory include metal ions adsorbed on a zeolite carrier, and mixed metal oxides without a carrier. The zeolite-based catalysts were prepared with an 8 to 1 2 mesh 13X molecular sieve and chlorides of chromium, iron, and copper. The type of sieve used has the following molar formula: Na20.A1,O3-2.8 f 0.2Si02.xH20. The sieve was slowly added to distilled water while stirring in order to remove excess sodium by a thorough wash. Three solutions were prepared using chromium (III), iron (III), and copper (11)chloride. Each solution was divided into five portions and each portion contained the stoichiometric amount of positive ions re-
basis for selection Arai et al. (1977) Arai et al. (1977) Arai et al. (1977) Kotera et al. (1976) Takeyama et al. (1975) Seiji (1975); Hattori et al. (1977) Takayasu et al. (1976) Miyadera et al. (1975) Seiji (1975) response to request response to request response to request response to request response to request
quired to replace all of the sodium ions in the sieve. One portion of a salt solution was added to molecular sieve and the slurry was stirred for 6 to 12 h to achieve equilibrium. The sieve was then separated from the solution and washed with distilled water. The procedure was repeated with the remaining four portions of salt solution. The degree of ion exchange of a sieve increases as the ratio of sodium to metal ion in the solution decreases. Thus, a high total degree of exchange can be obtained by a step-wise procedure using fresh solution each time since sodium is removed in each step. The prepared catalysts were dried at 250 O F for 48 h. An automatic tablet machine was used to prepare the Fe203/MnO/Zn0 catalyst. A mixture was made from powders of the three oxides in the proportion 52, 24, and 24% by weight, respectively. Starch was added to obtain 5 wt % binder. The mixture was pressed to 3/16 in. diameter by 3/16 in. long pellets. The Mn02/Pb0 catalyst was prepared by thoroughly mixing fine powders of Mn02 and PbO. The mixture consisted of 95% by weight of the former and 5% by weight of the latter. Plaster of paris was suspended in water to be used as a binder. The suspension and powder were mixed and stirred for 1 h. The mixture was dried overnight at 250 O F to remove the water. A solid mass of hard material containing 10 wt% binder on a dry basis was obtained by this procedure. The solid mass was crushed to irregular particles with a size of less than in. The particles were calcined at 930 OF for about 48 h. The CuO/Mn02/Sb203catalyst was prepared by the same procedure, with 57,38, and 5 wt % CuO, Mn02,and Sbz03, respectively. With regard to the catalysts that are commercially available or were supplied by manufacturers, the method of preparation was specified only for those containing molybdenum. The latter catalysts were prepared by impregnating 1/4-in.A120, granules with metal salts. The granules were dried prior to calcination at 930 OF. Experimental Conditions Experiments were performed on a uniform engineering basis in a fixed-bed reactor. Parameters such as space velocity, temperature, and gas composition were set to reflect the conditions at a coal-fired power plant. Table I11 sumarizes the conditions under which the experiments were conducted. Screening of the catalysts was started at a relatively low space velocity of 1500 h-l. The space velocity is defined as the actual volumetric gas flow rate through the reactor
Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984
811
P = Pressure gauge PC = Proportional controller TC = Thermocouple
i
Humidifier
I
I I I I
I I
I
I Bypass
Furnace
r-----
-w
---
IT
N2 (Purge)
I
0 2 "
20%
Cylinder A
NO" (Cal.) 750 ppm
I I
I
I
I
I
I
I
I
I Rotameters
'
I Reactor
I
I I
-++--_I
L--c+NO" 2800ppm
N02" 300ppm
Cylinder B Cylinder C
To NO, analyzer TO SO2 analyzer
S02/C02/N2 1.20 %I48% Cylinder D
line
I--
analyzer
To vent
Figure 1. Schematic of apparatus. Table 111. Experimental Conditions (Space Velocity, 1500 h-l: Temperature, 150-800 OF) flue gas compn: calcd from component cylinder compn and measd by instrum anal. (vol basis) exptl conditions 620-670 NO, (ppm) 720 0.05-0.10 N02/N0, (mol/mol) 0.10 2600-2800 2400-2800 SO2 (ppm) H2O (%I 70 co2 (%I 11.2 02 (700) 5 4.9-5.1b Nz (%) halance "11% at 200
OF,
13% at 150 OF. bSpot checks.
divided by the bulk volume of catalyst charged in the reactor. The temperature range investigated was 150 to 800 O F . However, it has previously been pointed out that at 300 O F or below, catalytic oxidation cqn be an add-on process. Also, the homogeneous decomposition of NO2to NO becomes important at high temperatures. Equilibrium data indicate that complete oxidation of NO cannot be obtained above 400 O F . For instance, the equilibrium conversion for the homogeneous reaction is 92% at 500 O F . However, higher temperatures may be of interest since recent data indicate that it is possible to achieve higher than the equilibrium conversion value in the catalytic oxidation of NO (Arai et al., 1977). The composition of the simulated flue gas was chosen to reflect the conditions when burning high-sulfur coal. This was done because wet scrubbing processes under development for simultaneous removal of SO2 and NO, require a molar ratio of SO2to NO, of at least three. Water vapor contents of 11and 13 vol % were used for flue gas at 200 and 150 O F , respectively, since the moisture content increases when the gas is cooled by evaporation of water. A heat balance for spray cooling flue gas at 300 O F and 7
vol % H20 was used to calculate the appropriate moisture content at 200 and 150 OF. Description of Apparatus. The apparatus used in this study consisted of a manifold for mixing gases from cylinders to obtain a simulated flue gas, a vertical fixed-bed reactor for catalytic conversion of NO to NO,, an analysis system for measuring the concentrations of NO/NOZ,SOz, and 0, in the simulated flue gas, and a multi-point digital temperature indicator for monitoring thermocouple readings at various locations. A schematic of the experimental setup is shown in Figure 1. The simulated flue gas was obtained from four gas cylinders with the following approximate composition on a volume basis: cylinder A, 20% O,/balance N,; cylinder B, 2800 ppm NO/balance N,; cylinder C, 300 ppm N02/balance N,; cylinder D, 1.20% S02/48% C02/baiance N2. Each cylinder was connected to a calibrated rotameter equipped with a pressure gauge and a flow controller to maintain a constant flow rate at varying downstream pressures. Teflon tubing was used for gas flow at ambient temperature and stainless steel tubing was used for elevated temperatures. The stainless steel tubing was wrapped with heating tapes and insulation to maintain and control the temperature of the simulated flue gas. The gas from cylinder A flowed through a humidifier to provide water vapor. In most experiments, the humidifier was maintained at 154 O F so that the saturation level of the 02/Nz exit gas was 28 vol %. The gases from all four cylinders were mixed together in equal proportions to provide simulated flue gas with the composition listed in Table 111. The humidifier consisted of a 6-L insulated stainless steel vessel wrapped with heating tape and equipped with a thermocouple. The temperature of the water in the humidifier was controlled at a desired level (f0.5 "F)by means of a proportional temperature controller. The gas
812
Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984
Table IV. Percent Oxidation of NO to NO2 at a Space Velocity of cat. 150 O F b 200 O F c 300 O F __ 1.7 7.6 Cr3+-zeolite _0 0 Fe2+-zeolite _0 Cuz+-zeolite 0 __ __ 72.7d Fez03/MnO/Zn0 _18.2 11.7 MnOz/PbO __ 0 2.9 CuO/Mn02/Sb203 -_ 15.6 8.d Crz03-A1203e Ni0-A1203 47.2 57.6 63.5 -CoO-Alz03 _53.9d 37.0 6.6 53.4 Fe203/Mo03-A1203 -_ 0 CuO/M003-A1203 46.0 53.4 42.2 Biz03/Mo03-A120, 34.1 V,05/MoO-A1203 35.3 53.0 _0 0 Pt-AlZOB
__
1500 h-'" 400OF 0 0 0
600 O 0
500 O F 0 0 47.7 -_
__
1.0 0 12.4 0 0
F
800 OF
---
0 0.9
1.8 -_ 0 -_ 0 0 0 0 12.0
__
700 O
55.2
0
0 4.3 0 10.6
F
__
-_ -_
0 7.1 -0 43.6
12% CO,; 7% H,O; 5% 0,; 2400 to 2800 ppm SO,;620 to 670 ppm NO,; balance N2(volume basis). 13% H20. 11% H,O. dOptimum temperature determined a t a space velocity of 15 000 h-l. e 500 ppm NO,. f2lOOppm SO,.
Table V. Percent Oxidation of NO to NOz At a Space Velocity of 15000 h-Io cat. 150 OFb 200 O F c 300 O F 400 O _Fe203/MnO/Zn0 0 15.3 9.2 CoO-Al,03 3.9 11.0 4.5 0 a
F
500 OF
12% CO,; 7% H,O; 5% 0,;2800 ppm SO,; 630 to 650 ppm NO,; balance N, (volume basis).
was dispersed by a '/2-in. stainless tube with a fine frit at the end. The humidifier was also equipped with a pressure gauge, an extra thermocouple connected to the digital indicator, and an external glass tube to measure the liquid level. The reactor consisted of a 25 in. long X 1in. i.d. vertical quartz tube. The catalysts were supported on a coarse frit quartz disk and the simulated flue gas entered the reactor at the top. In most experiments, the catalyst bed had a depth of 10 in. Ball joints were used to connect the reactor inlet and outlet with the stainless steel tubing. A Solv-Seal joint was attached to the top of the reactor to allow the latter to be charged with catalyst. The reactor was surrounded by a tube furnace and the temperature was controlled by a proportional controller using a signal from a thermocouple attached to the reactor wall. There was an additional thermocouple located in a small well ' / 4 in. above the quartz support disk and connected to the digital indicator. Nitrogen was used to purge the system of any impurities, and another cylinder containing 750 ppm of NO in N2 was used to calibrate the NO analyzer. A Thermo Electron Corporation (TECO) chemiluminescentinstrument, Model 10A, was used to measure NO concentration. This instrument has a built-in thermal converter to decompose NOz to NO. The thermal converter can be used to measure NO, (NO plus NOz) or bypassed to measure NO only. For SOz,an International Biophysics Corporation (IBC) Celeso instrument wm used. This analyzer has a fuel cell detedor which is specific for SOz. A Beckman oxygen analyzer based on paramagnetism was used for spot checks of the oxygen concentration in the flue gas. The SOz and oxygen analyzers were preceded by U-tube cold traps immersed in ice water to remove moisture from the flue gas. Experimental Procedure. Blank runs were performed without catalyst in the reactor to determine if any oxidation occurred in the absence of catalyst. This step is important in order to establish that it is the catalysts that are responsible for the oxidation. The ratio of NOz to NO was found to be the same when flue gas in the temperature range of 150 to 800 OF was passed through the empty reactor as when the gas was bypassed around the reactor. The following procedure was used to perform a test run (experiment) with a catalyst sample. Simulated flue gas was passed through the system while bypassing the reactor.
600 O
--
__
0
0
* 13%
F
700 OF
__
0
H20. 11% HzO.
The temperature of the humidifier was set to obtain the desired moisture content by assuming complete saturation of the gas leaving the humidifier. The temperature of the stainless steel tubing was set at about 30 O F above the water dew point using the system of heating tapes and Variacs. When the readings of NO, and SOz concentrations in the flue gas attained constant levels as indicated by chart recorders, the flue gas was passed through the reactor by switching a three-way valve. The temperature of the catalyst bed was controlled at a desired value ( f l OF) by using a proportional temperature controller. Steady state was considered to be achieved when the NO, and SOz concentrations in the outlet gas from the reactor approached the inlet values within a deviation of 5%. The exposure time before steady state was reached varied from about 1 h up to almost 10 h depending on temperature, type of catalyst, and space velocity. All analysis instruments were calibrated at least once a day.
Results and Discussion Data on the oxidation of NO to NOz at a space velocity of 1500 h-l are presented in Table IV. Several catalysts exhibited optimum performance at a temperature of 200 OF. Catalysts exhibiting greater than 50% oxidation of NO to NOz at this temperature in order of increasing performance are Vz05/Mo0-A1z03, Biz03/Mo03-A1,03, FezO3/MoO3-AlZO3,CoO-Al,03, Ni0-A1,03, and Fe203/ MnO/ZnO. The latter catalyst exhibited 72.7% oxidation at 200 O F . This catalyst and the CoO-Alz03 catalyst were tested at a space velocity of 15000 h-' to determine the optimum temperature. The results of these tests are summarized in Table V. A high space velocity was used because of the high degree of adsorption of NO, exhibited by these catalysts. For example, at a space velocity of 1500 h-' the Co0-AlzO3 catalyst was exposed to simulated flue gas at a temperature of 200 OF for more than 7 h before any NO, was detected in the outlet gas from the reactor. The effect of space velocity on the oxidation of NO at a temperature of 200 O F using CoO-Alz03 or Fe203/ MnO/ZnO catalyst is shown in Figure 2. The latter catalyst exhibited 50% or greater oxidation at space velocities below 3500 h-l. The results summarized in Table IV indicate that two of the catalysts exhibited optimum performance at a temperature of 600 OF. The Cuz+-zeolite catalyst and the
Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984
813
100
t
Simulated flue gas (composition given in Table Ill) Temperature: 2OO0F
c
E i
P
50
.-c5 0
0"
0
100
-
Simulated flue gas (composition given in Table Ill) Space velocity: 1,500 hr-I Temperature: 2OO0F
c
E
$
-
---
50.-
C
.-w
-
Bi203/Mo03-A1203
0
0"
0-
5
10
Pt-AlZO3catalyst exhibited 55.2 and 43.6% oxidation of NO,respectively, at this temperature. It appears that at temperatures of 600 O F and above, SOz is catalytically oxidized to SO3 because when the outlet gas was bubbled through water, a mist was formed which was due to the presence of H#04. Although the SO3 in the outlet gas was not determined quantitatively, no SO2was detected in the outlet gas from the reactor when using the Pt-A1203 catalyst a t 600 and 700 O F . Three experiments were performed with the Cuz+zeolite catalyst using a gas stream with the following composition on a volume basis: NO, 640 ppm; Oz,5%; N2, balance. The results are listed in Table VI. These experiments were performed in order to determine the effect of deleting SOz, CO2, and HzO from the gas stream used in the test runs and to obtain data for comparison with data available from the literature. The equilibrium conversion for the
15
20
1
I
25
30
homogeneous oxidation of NO is also given in Table VI for comparison purposes. The equilibrium data indicate that complete oxidation of NO cannot be obtained above 400 O F . A comparison of the data in Tables IV and VI for the Cu2+-zeolitecatalyst indicates that there is less oxidation of NO when SOz, COP,and HzO are present in the gas stream. The data obtained by others agree with the data obtained from the present study for the Cu2+-zeolite catalyst assuming that 60% of the available sites on the prepared zeolite are occupied by Cuz+(Arai et al., 1977). Four of the catalysts exhibiting high activity were selected for relatively long-term runs to study possible deactivation effects. These catalysts were Fez03/MnO/ ZnO, NiO-AlZO3, Bi20,/Mo03-A1203, and Cu2+-zeolite. The catalysta were exposed to simulated flue gas at a space velocity of 1500 h-' for more than 20 h; the temperature
Ind. Eng. Chem. Process Des. Dev. 1984, 23, 814-819
814 T a b l e VI. P e r c e n t O x i d a t i o n of V e l o c i t y of 1500 h-’“
cat. Cu2+-zeolite equilibrium conversion for homogeneous reaction
NO
400
to
O F
60 100
NOz At a 500 O 76 92
F
and modification of the catalyst operating conditions. The appropriate option(s) dependb) upon the cause(s) of catalyst deactivation. Although it may be difficult to find a catalyst that yields nearly complete oxidation of the NO at a space velocity of 1500 h-l or greater, it should be possible to achieve 80 to 90% oxidation with a relatively modest effort. Complete oxidation of NO to NOz may not be required for high NO, removal because Nz03can take part in the absorption of NO,. If N203is absorbed at the same rate as Nz04,it should be possible to obtain more than 90% removal of NO, in a wet scrubber if at least 50% of the NO is oxidized in an upstream catalytic reactor.
Space
600 OF 70 79
“5% 0,; 63C-640 ppm NO; balance N2 (volume basis).
was 200 OF for the first three catalysts and 550 O F for the Cu2+-zeolite. The first three catalysts exhibited a decreasing activity after 13 to 15 h as shown in Figure 3, and the Cuz+-zeolitewas deactivated much faster. The causes for this deactivation have not been determined. The experimental results are very encouraging. A Fe20,/MnO/Zn0 catalyst yielded over 70% oxidation of NO to NO2 in simulated flue gas at a temperature of 200 OF and a space velocity of 1500 h-l. The optimum temperature for most of the catalysts tested appears to be around 200 O F . This temperature is very suitable for an add-on process since particulates can be removed in a dry collection device ahead of the catalyst bed. The flue gas can then be spray-cooled to 200 O F with water. After catalytic oxidation of NO to NOz, the flue gas can be sent to a limestone scrubber for removal of SO2and NO2. The scrubber will cool the flue gas to its adiabatic saturation temperature of about 125 OF. The experimental results indicate that species such as SO2and H20, which are present in the flue gas, lower the oxidation level considerably. This effect has previously been reported in the literature and may be explained by adsorption of SO2,HzO,etc., on the catalyst surface so that there is a decrease in the number of available active sites. The experimental results have revealed an additional problem with regard to catalytic oxidation of NO in flue gas. After about 14 h of run time, the catalyst activity starts to decrease. Preliminary investigations have revealed that formation of nitrates occurs on the catalyst surface. Additional work is needed to solve the problem of catalyst deactivation as well as to find a more active catalyst. The options for eliminating catalyst deactivation include selection of a catalyst that does not form nitrate and nitrite salts or sulfate and sulfite salts,modification of the catalyst carrier, modification of the catalyst preparation method,
Acknowledgment
The authors are grateful to Dr. Robert A. Ference of the Climax Molybdenum Company for supplying the molybdenum-containing catalysts and to Mr. Masahi Y amada of Tanaka Kikinzoku Kogyo K. K., Tokyo, Japan, for supplying the platinum-containing catalyst. Registry No. Fe203, 1309-37-1;MnO, 1344-43-0;ZnO, 131413-2; MnOz, 1313-13-9; PbO, 1317-36-8; CuO, 1317-38-0; Sbz03, 1309-64-4; Cr2O3, 1308-38-9; NiO, 1313-99-1; COO, 1307-96-6; Moo3, 1313-27-5;BizO3, 1304-76-3;Vz05, 1314-62-1;NO, 1010243-9; NO,, 11104-93-1;Cr, 7440-47-3; Fe, 7439-89-6 Cu, 7440-50-8; Pt, 7440-06-4.
Literature Cited Arai, H.; Tominaga, H.; Tsuchiya, J. “Proceedings, 6th International Congress on Cataiysis”; Chemical Society, Letchworth, England, 1977; p 997. Hattori, H.; Kawai, M.; Egashlra, S.; Sato, G.; Owaki. N.; Hanada, M.; Kuroda, R.; Kutsukake, M.; Annaka, T. Kogal Hakua Shobo 1977, 12(1), 62. Kotera, T.; Ozasa, M.; Takano, T. Japanese Patent 76 44 558, 1976. McKee, R. H. U.S. Patent 1319322, 1921. Miyadera, T.; Kawai, M.; Hirasawa, S.; Mlyajima, K.; Oyama, M.; Kklo, N.; Yamaraki, M.; Hattori, H.; Kobayashi, H. 32nd Spring Term Annual Meeting of the Japan Chemical Society, Tokyo, Japan, April 1975; Japan Chemical Society: Tokyo, Japan, 1975; Paper 2005 (Japanese). Seiji. A. Kogai Hakua Shobo 1975, 10(1), 32. Takayasu, M.; An-nen, Y.; Morita, Y. Mntyo Kyokai-Shi 1975, 54(11), 930. Takayasu, M.; An-nen, Y.; Morita, Y. Was& Daigaku Rikogaku Kenkyusho Hokoku 1976, 72, 17 (English). Takeyama, F.; Masuda, K.; Takahashi, A. Japanese Patent 7562859, 1975.
Received for review April 22, 1983 Accepted January 25, 1984
This work wm supported by the Tennessee Valley Authority under Contract No. TV-52401A.
Prediction of Point Efficiencies on Sieve Trays. 1. Binary Systems Hong Chan and James R. Fair’ Department of Chemical Engineering, The University of Texas, Austin, Texas 78712
A new model for the prediction of mass transfer efficiency on crossflow sieve trays has been developed. The model is based on twwesistance concepts, takes into account axial dispersion of the flowing froth or spray, and is supported by a large bank of performance data on commercial scale sieve tray columns. The model fits the data sample (143 points) with an average absolute deviation of 6.27%, a great improvement over the only other general model available, that of the AIChE Research Committee. For the same data sample, the latter model gives a fit of 22.9% average absolute deviation.
One of the remaining uncertainties in the design of a sieve tray distillation column is the specification of mass transfer efficiency. Methods for predicting vapor and liquid capacity as well as pressure drop seem fairly well 0196-4305/84/1123-0814$01.50/0
in hand and lead to reliable values for design. Other hydraulic parameters, such as entrainment and weeping flow rates, have been treated previously in the literature, and while published methods for their prediction are not 0
1 9 8 4 American Chemical Society
