Basic Inventory Control 5.0.113 serial key or number

vvEPA United States Industrial Environmental Research EPA-600/7-80-083 Environmental Protection Laboratory April 1980 Agency Research Triangle Park NC 27711 Sulfur Dioxide Oxidation in Scrubber Systems nteragency Energy/Environment R&D Program Report

---

------- RESEARCH REPORTING SERIES Research reports of the Office of Research and Development, U.S. Environmental Protection Agency, have been grouped into nine series. These nine broad cate- gories were established to facilitate further development and application of en- vironmental technology. Elimination of traditional grouping was consciously planned to foster technology transfer and a maximum interface in related fields. The nine series are: 1. Environmental Health Effects Research 2. Environmental Protection Technology 3. Ecological Research 4. Environmental Monitoring 5. Socioeconomic Environmental Studies 6. Scientific and Technical Assessment Reports (STAR) 7. Interagency Energy-Environment Research and Development 8. "Special" Reports 9. Miscellaneous Reports This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT - RESEARCH AND DEVELOPMENT series. Reports in this series result from the -effort funded under the 17-agency Federal Energy/Environment Research and Development Program. These studies plate to EPA's mission to protect the public health and welfare from adverse effects of pollutants associated with energy sys- tems. The goal of the Program is to pssure the rapid development of domestic/ energy supplies in an environmentally-compatible manner by providing the nec- essary environmental data and control technology. Investigations include analy- ses of the transport of energy-related pollutants and their health and ecological effects; assessments of, and development of, control technologies for energy systems; and integrated assessments of a wide'range of energy-related environ- mental issues. EPA REVIEW NOTICE This report has been reviewed by the participating Federal Agencies, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Government, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. This document is available to the public through the National Technical Informa- tion Service, Springfield, Virginia 22161.

---

------- EPA-600/7-80-083 April 1980 Sulfur Dioxide Oxidation in Scrubber Systems by J.L Hudson University of (Virginia Department of Chemical Engineering Charlottesville, Virginia 22901 Grant No. R805227 Program Element No. EHE624 EPA Project Officer: Robert H. Borgwardt Industrial Environmental Research Laboratory Office of Environmental Engineering and Technology Research Triangle Park, NC 27711 Prepared for U.S. ENVIRONMENTAL PROTECTION AGENCY Office of Research and Development Washington, DC 20460 I'ml

---

------- ABSTRACT This study relates the liquid phase oxidation of bisulfite and sulfite anions to sulfate to the conditions in lime/limestone scrubbing systems such as those used for removal of sulfur dioxide from power plant stack gases, Experiments were carried out for the oxidation in both calcium sulfite and sodium sulfite clear solutions and in calcium sulfite slurries. In the slurry studies, both the rate of chemical reaction and the rate of solid to liquid mass transfer were investigated, but with both slurries and clear solutions, the experiments were run so that gas to liquid transfer of oxygen was not a limiting resistance. A mathematical model for the dissolution of solid particles and liquid phase chemical reaction was developed in conjunc- tion with these experimental results, The oxidation was carried out over a pH range of 4.0 to 5.5 (although some experiments were at pH 6 and pH 11), at temperatures of 25°C to 50°C, and in the presence of iron and manganese catalysts and several organic acid inhibitors. Batch reactors were used for the slurry studies and for the slower clear solution oxidations while a flow reactor was employed for faster clear solution reactions. Although under special combinations of catalyst, pH, and organic acid, the order was as low as one or rose to two, the homogenous chemical oxidation rate of calcium sulfite oxidation is 1.5 prder over most of the range of conditions. The rate increases strongly with increasing pH. Of the organic acids studied, glycolic is the strongest inhibitor, followed by adipic and succinic. Citric and acetic acids are less inhibitory than the others. Both manganese and iron catalyze the reaction even in the presence of the organic acid inhibitors. Oxygen concentration varied over a large range; its order is 0.5 in the rate expression for the manganese catalyzed oxidation at higher pH. In a three phase slurry oxidation the overall reaction of sulfite to sulfate declines with increasing pH; this decrease is caused by the sharply reduced calcium sulfite solubility with increasing pH. This calcium sulfite solubility was determined in independent experiments. The results of the mathematical model for the dissolution and reaction of calcium sulfite particles compare well with results of the three phase slurry oxidations. ii

---

------- CONTENTS Abstract • i:L Figures v Tables • • • • viii 1. Introduction . . . 1 A. General Considerations * 1 B. Background Literature 3 2. Experiments • 32 A. Apparatus • 33 1. Batch Reactor 33 2. Flow Reactor , 37 B. Procedure 44 1. Rate from Concentration Measurements 44 2. Rate from pH Measurements 46 3. Rates from Temperature Measurements 46 C. Analysis of Calcium Sulfite 48 D. Equilibrium Relationship for S 44Species 50 E. lodometric Titration 50 F. Analysis of Data 52 1. Analysis of a Single Experiment. 52 2. Multiple Regression Analysis 53 3. Analysis of pH Method Data 61 4. Analysis of Flow-Thermal Method 64 3. Oxidation in Calcium Sulf ite Solutions 69 A. CaSO- Oxidation (No organic acids) 70 B. CaSO^ Oxidation in the Presence of Organic Acids.. 71 1. Succinic Acid 83 a. Effect of pH on the rate of oxidation.... 83 b. Effect of succinic acid concentration.... 85 2. Adipic Acid 90 3. Glycolic Acid 90 4. Comparison of Rates of Oxidation with Succinic Acid, Adipic Acid, Glycolic Acid, Citric Acid, and Acetic Acid 96 5. Effect of Catalysts—Manganese and Iron 107 a. Succinic Acid 107 b. Glycolic Acid Ill C. Reaction in Liquor from Penberthy Oxidation Runs at Shawnee 118 D. Dependence of Oxidation Rate on Stirring Speed and Oxygen Flow Rate 122 E. Rate of CaSO^ Oxidation at High Catalyst Concentra- tions by Flow-Thermal Method 133 iil

---

------- 4. Oxidation in Slurries 139 A. Experiment 140 1. Solubility Studies 140 a. Theoretical Analysis 140 b. Experimental Results 141 2. Oxidation in Calcium Sulfite Slurries 143 3. Oxygen Flow Rate Studies 143 4. pH and Temperature Studies 146 5. Slurry Density Studies 153 6. Catalyzed Slurry Oxidation Studies 153 7. Liquid Phase Slurry Behavior 161 8. pH Behavior During Reaction 161 9. Slurry Reactions with and without pH Controller 161 10. Liquid phase Catalyst Studies 171 B. Mathematical Model 171 1. General Description of Model 171 2. Physical Description of Particles 174 a. Electron Micrograph 174 b. Particle Size Distribution 175 3. Derivation of the Spherical Model. 175 4. Solutions of the Spherical Model. 179 5. Derivation of the Flat Plate Model 192 5. Oxidation in Sodium Sulfite Solutions 194 A. Na-SO^ Oxidation . . . 195 1. Manganese Catalyst 195 B. ^280^ Oxidation with Succinic Acid Buffer 195 1. Iron Catalyst 195 2. Manganese Catalyst 208 3. Mixed Catalysts. . 217 C. Rate of Na.SO- Solution Oxidation by the Flow-Thermal Method 221 6. Conclusion 227 References . . . 233 iv

---

------- FIGURES Number Page 1-1 Effect of pH on the Relative Concentrations of Sulfur (4+) Species in Solution .4 1-2 Electron Micrograph of Calcium Sulfite Particles - Sulfite and Sulfate Particles . 8 1-3 Electron Micrograph of Calcium Sulfite Particle Individual Agglomerate 8 1-4 Comparison of Gladkii's Figure 6 with Figure 5 23 1-5 Comparison of Slurry Oxidations at 50°C. . . 24 1-6 Comparison of Gladkii's Figure 6 with Soviet Slurry Oxidation Rate Expression 26 1-7 Determination of Soviet Reaction Order .......... 27 1-8 Comparison of Slurry Oxidations at 40°C 29 2-1 Experimental Apparatus 34 2-2 Reaction Vessel 35 2-3 Tubular Flow Reactor 38 2-4 REPORT and CURVIT Programs 54 2-5 Element of Reacting Fluid 66 2-6 Sample Output for Flow Reactor 68 3-1 Rate of Calcium Slufite Oxidation; no added catalyst ... 72 3-2 Rate of Calcium Sulfite Oxidation; [Mn] = 0.6 ppm 76 3-3 Rate of Calcium Sulfite Oxidation; [Mn] = 0.53 ppm, [Fe] = 1.02 ppm . 77 3-4 Rate of Calcium Sulfite Oxidation; variable [Mn] 79 3-5 Rate of Calcium Sulfite Oxidation; variable [Mn], ([S ] = 0.0037 mol/£) 80 3-6 Effect of Temperature on the Rate Constant ........ 81 3-7 Effect of pH on Rate of Reaction 86 3-8 Concentration Sulfur (4+) Versus Time for Various Concentrations of Succinic Acid ..... 87 3-9 Concentration Sulfur (4+) Versus Time, Run 566, 0.2M Succinic Acid 88 3-10 The Inhibition of the Sulfite Oxidation Due to Succinic Acid 89 3-11 Initial Rate Versus Succinic Acid Concentration 91 3-12 The Inhibition of the Sulfite Oxidation Due to Adipic Acid. • • /-+ • (^ 3-13 Concentration S vs Time for Various Concentrations of Glycolic Acid, pH = 4.0 97 3-14 Inhibition of the Sulfite Oxidation due to Glycolic Acid . 98 3-15 The pH Effect on the Inhibition of the Sulfite Oxidation Due to Glycolic Acid 99 3-16 Effect of Glycolic Acid on Rate of Oxidation of CaSO , at T=50°C, pH=4.0, [S ]=0.01M ....... 102 3-17 Comparative Strength of Various Organic Acids as Inhibitors at 0.2M 103 v

---

------- FIGURES (continued) 3-18 Comparative Strength of Various Organic Acids as Inhibitors at 0.01M 104 3-19 Sulfur (4+) Concentration vs Time for Various Mn Concentrations HQ 3-20 Effect of Mn on Oxidation Rate+of Succinic Acid Buffered CaSO Solutions for [S ]=0.01M, T=50°C, pH=5.0 ... 113 3-21 Concentration Sulfur (4+) vs Time, 1000 ppm Glycolic Acid, Various Concentrations of Mn 114 3-22 Effect of Mn Addition to Glycolic Acid Inhibited CaSO Oxidation; rate at [S ] = 0.01M 116 3-23 Effect of Mn Addition to Glycolic Acid Inhibited CaSO, Oxidation; Rate at [S ] = 0.01 M 117 3-24 Stirring Speed = 215 rpm '.'.'. 123 3-25 Stirring Sp,eed = 1800 rpm *'-.'' 124 3-26 Rate at [S ] = 0.01 M vs Stirring Speed . .' .' .' .' . .' .' .' 128 3-27 1.4 Order Rate Constant vs Mn Concentration in CaSO., Filtrate Oxidations ....... 132 3-28 Representative T vs t Results for CaSO- Oxidation. .'.'.'.' 134 3-29 Effect of [Mn] on CaS03 Oxidation Rate 136 4-1 Solubility of Calcium Sulfite at 40°C. . 142 4-2 Effect of Slurry Density on Oxidation Rate ........ 144 4-3 Typical Slurry Reaction Curve 145 4-4 Effect of Mn Catalyst on Slurry Oxidation Rate ...... 147 4-5 Slurry Oxidation, pH = 4.5 ' 143 4-6 Slurry Oxidation, pH = 4.7 "'.'.'.'.'.'. 149 4-7 Slurry Oxidation, pH = 5.0 .............. . . 150 4-8 Effect of pH on Slurry Oxidation Rates • • • ^ 4-9 Slurry Oxidation, pH = 4.5 ' 154 4-10 Slurry Oxidation, pH = 4.5 '.'.'.'.'.'.'. 155 4-11 Catalyzed and Uncatalyzed Oxidations at 50°C 158 4-12 Effect of Catalyst on pH Behavior (40°C) ......... 159 4-13 Effect of Catalyst on pH Behavior (50°C) ......... 160 4-14 Slurry Oxidation, 200 ppm Mn, pH = 5.0 '. 164 4-15 Slurry Oxidation, 200 ppm Mn, pH = 4.5 165 4-16 Slurry Oxidation, 200 ppm, pH = 5.0 166 4-17 Slurry Oxidation, pH = 5.0 . . '.'.'. 167 4-18 pH Behavior During Slurry Oxidations; Effect of Slurry Density . . . \ _ 168 4-19 pH Behavior During Slurry Oxidations; Effect of Initial . • PH 169 4-20 pH Behavior During Slurry Oxidations; Highly Catalyzed Runs 170 4-21 Slurry Oxidation with No Mn Added 172 4-22 Liquid Phase Catalyst Behavior During Slurry Oxidations. . 173 4-23 B-30 Program and Sample Calculation 181 4-24 Determination of ke from Highly Catalyzed Slurry Data (2000 ppm Mn add^d) 187 4-25 Comparison of Total S from Model to Experimental Slurry Oxidation Data for No Added Mn 188 VI

---

------- FIGURES (continued) 4-26 Comparison of Model to Experimental Data for 6.66 ppm Mn Added. . 189 4-27 Comparison of Model to Experimental Data for 200 ppm Mn Added 19° 4-28 Comparison of Liquid S from Model to Experimental Slurry Oxidation Data for No M£+ Added 191 4-29 Computer Predicted Liquid [S ] Profiles 193 5-1 Dependence of Rate on Sulfite Concentration 197 5-2 Dependence of Rate on Manganese Concentration 198 5-3 Sulfur S Concentration vs Time for Iron Catalyzed Na2S03 Oxidations 20° 5-4 Rate at [S ] = 0.028 mol/£ v£ Concentration Iron Added for Sodium Sulfite Oxidations 203 5-5 Sulfur (4+) Concentration vs Time for Iron Catalyzed Oxidations at Various Temperatures Using Na^O- . . . 204 5-6 Activation Energy Determination for Fe Catalyzed Na2Su3 Oxidations ' • • 206 5-7 Rate at [S ] = 0.028 M vs Concentration Fe 210 5-8 Sulfur (4+) Concentration vs Time for Manganese Catalyzed Oxidations of Na9SO« Solutions 213 5-9 Rate at [S ] = 0.02S M vs Concentration Mn 215 5-10 Reduced Sulfur (4+) Concentration vs Time for Comparison of Oxidation Reactions 218 5-11 Comparison of Mn Added Only to Oxygen Side or Only to Sulfur Side 226 vii

---

------- LIST OF TABLES Number 1-1 Fly Ash Analysis 6 1-2 Limestone Analysis 9 1-3 Comparison of Literature of Sulfite Dioxide Oxidation. . . 15 2-1 Analysis of Calcium Sulfite Solid 49 3-1 Rate of Calcium Sulfite Oxidation; No added catalyst (PH - 4.6) 73 3-2 Rate of Calcium Sulfite Oxidation; [Mn] = 0.6 ppm 74 3-3 Rate of Calcium Sulfite Oxidation; Added Mn and Fe . . . . 75 3-4 Rate of Calcium Sulfite Oxidation; variable [Mn] 78 3-5 Effect of Temperature on Rate Constant 82 3-6 Effect of pH on Succinic Acid Buffered CaSO» Solutions . . 84 3-7 Effect of Succinic Acid on Oxidation Rate. 92 3-8 Results of Order and Rate Determination Analysis of Data for Calcium Sulfite Oxidation with Varying Concentra- tions of Adipic Acid 93 3-9 Results of Order and Rate Determination Analysis of Data for Calcium Sulfite Oxidations with Varying Concen- trations of Glycolic Acid Added 95 3-10 Effect of Glycolic Acid on Oxidation Rate of Calcium Sulfite 100 3-11 Results of Order and Rate Determination Analysis of Data for Calcium Sulfite Oxidations with Varying Concen- trations of Glycolic Acid Added 101 3-12 Effect of Organic Acids on Oxidation Rate of Calcium Sulfite 105 3-13 Catalyst Impurity Levels in CaSO., Solutions 108 3-14 Effect of Mn Impurity on Oxidation Rate 109 3-15 Effect of Mn Concentration on Oxidation Rate of Succinic Acid Buffered CaS03 Solutions 112 3-16 Effect of Manganese Addition to Glycolic Acid Inhibited Runs . 115 3-17 Oxidation of Sulfite Solutions 119 3-18 Supplemental Analysis of Shawnee Sample 121 3-19 Effect of Stirring Speed - No added catalyst 125 3-20 Effect of Stirring Speed on Reaction Rate Constant - 0.5 ppm Mn. . 127 3-21 Effect of Oxygen Flow Rate on Oxidation in Unsaturated Solutions . 129 3-22 Results of Filtrate Oxidations; pH,- = 4.5, T = 40°C. . . . 131 3-23 Effect of High Mn Concentration on CaSO.. Oxidation Rate . . 137 3-24 Effect of pH on the Rate of Mn Catalyzed CaS03 Oxidation. . 138 4-1 Results of Studies Showing Effect of Initial pH on Oxidation 152 viii

---

------- LIST OF TABLES (continued) 4-2 Effect of Slurry Load on Oxidation Rate at 40°C 156 4-3 Effect of Catalyst on Slurry Oxidation Rate 157 4-4 Experiments Made in Mn Catalyst Study 162 4-5 Experimental Conditions Used in Slurry Oxidations Shown in Figures 4-5 to 4-7, 4-9, 4-10, 4-14 to 4-17 163 4-6 Size Distribution of Calcium Sulfite Particles (Coulter Method) 176 5-1 Na SO Oxidation • 196 5-2 Results of Sodium Sulfite Oxidation with Iron Catalyst Added 202 5-3 Results of Iron Catalyzed Reactions at Various Temperatures Using Sodium Sulfite 205 5-4 Results of Sodium Sulfite Oxidations with Varying Initial Sulfite Concentration and 5 ppm Iron Added 207 5-5 MULTREG Results for Iron Catalyzed Na2S03 Oxidations. ... 209 5-6 Results of Sodium Sulfite Oxidations with Manganese Added . 211 5-7 Regression Analysis of Manganese Catalyzed Na^O^ Oxidations 2l6 5-8 Results of Sodium Sulfite Oxidations with Iron and Manganese Added. 5-9 Multiple Regression Results for Mixed Catalyst Sodium Sulfite Oxidations 22° 5-10 Effect of [Mn] on Na2SOa Oxidation Rate 223 5-11 Effect of pH on Na2S03 Dxidation Rate 224 ix

---

------- SECTION 1 I. Introduction A. General Considerations B. Background Literature INTRODUCTION A. General Considerations In the removal of sulfur oxides from stack gases by any of the scrubbing methods, a fraction of the sulfur compounds is oxidized to sulfates. This oxidation occurs whether the scrubbing agent is a slurry (lime/limestone systems) or a clear solution (double alkali or other such systems). Further- more, the oxidation occurs not only in the absorber, but also in other sec- tions of the system, such as in the hold tanks. In lime/limestone scrubbing systems S02 oxidation is important for sev- eral reasons. (The species actually taking part in the reaction is either a bisulfite or sulfite ion depending on the pH of the solution; however, it is convenient and it is common practice to refer to sulfur dioxide or calcium sulfite oxidation.) The oxidation in the scrubbing system can increase the degree of supersaturation of calcium sulfate, leading to an increase in the rate of gypsum scale formation (183, 30). Thus, it is important to limit calcium sulfite oxidation in some systems. On the other hand, calcium sulfate is preferable to calcium sulfite from the standpoint of solids disposal since the sulfate has a higher settling velocity than the sulfite and the sulfate also has a higher compaction and a lower chemical oxygen demand (31-36). The latter is important from water pollution considerations. Therefore, in some

---

------- limestone scrubbing systems it may be desirable to promote or inhibit oxi- dation. Oxidation in lime/limestone systems, both in the scrubber and in the hold tanks, is a very complicated process. The rate of oxidation can depend on chemical kinetics since the rates of liquid phase reactions of sulfite ion and bisulfite ion vary with concentration, with the presence of catalysts or inhibitors, and with the type of oxidizing agents, viz., oxygen or nitrogen oxides. (It should be noted that very small concentrations of some catalysts can influence the reaction greatly and that, therefore, many reactions which are supposedly being run without catalysts are in fact being catalyzed by impurities present.) The rate of oxidation is also influenced by phase and chemical equilibria. There are three phases present in the scrubbing system, solid, liquid, and gas. Both the phase equilibria and chemical equilibria are strongly dependent on pH. For example, lowering the pH in the presence of calicium sulfite increases the solubility and, therefore, increases the con- centration of dissolved sulfur containing species. This can increase the rate of oxidation under some conditions, particularly, say, in a hold tank. How- ever, a lowering of the pH of scrubbing liquor in the absorber converts sul- fite ion to bisulfite ion; this would lower the rate of oxidation since sulfite ion is oxidized much more quickly than bisulfite ion in a clear liquor (but not in a slurry where solubility effects are important). Finally, the rate of oxidation can be influenced or controlled by mass transfer, viz., by the rate of transfer of oxygen to the liquor and by the rate of dissolution of solid particles. The complex sulfur dioxide oxidation process was broken down into steps and each step studied separately. The oxidation of calcium sulfite and sodium sulfite clear solutions was inventigated as a function of manganese and iron

---

------- catalysts, pH, temperature, and organic acid inhibitors. The rate of dissolu- tion of calcium sulfite particles in a stirred liquid was determined and this information, along with the kinetic rate results, were used as a basis for a mathematical model for the dissolution and reaction of calcium sulfite parti- cles . Measurements were then made on the rate of oxidation in calcium sulfite slurries and the results compared to predictions of the model. B. Background Literature Atmospheric sulfur dioxide affects our lives in many ways. It is a health hazard (9, 10), corroder of structures (9.4) and equipment, visibility impairer (106), and habitat modifier (88, 166). Man's activities of smelting and burning fossil fuels among others have significantly altered the world sulfur budget (97), threatened his health (10, 57), and now require control (60, 61). Control methods include: wet scrubbing, dry sorbents (furnace injection) (144, 161), catalytic oxidation (5, 27, 37, 38, 73, 121, 123, 146, 164), dry filters (and baghouse add-ons) (1), and fuel precleaning (111). Today, wet scrubbing is the most widespread method (59) for cleaning S02 from stationary sources. When the S02 enters any scrubbing liquor, the fol- lowing equilibria are set up (154): S02 + H20^=^S02-mH20^=^HSO; + H+^=^So|" + H+ low *-pH — —*" high These relationships are shown graphically in Figure 1-1. Water-only scrubbing is ineffective. Removing the H+ in the last two equilibria above effectively shifts the concentrations to the right, thereby sequestering the sulfur dioxide as bisulfite on sulfite anions. All of the scrubbing methods

---

------- 1.0. 0.0- 0.6— Hole Fraction 0.4- 0.2-H 0.0. HS03 Conditions 40°C 0.02 M Na SO •'""'3 1 T pH T 8 S03 2- 10 Figure 1-1. Effect of pH on the Relati.ve Concentrations of Sulfur (4+) Species in Solution.

---

------- turn on this principle of tying up hydronium ion by the addition of many kinds of alkalis (eg. CaC03, CaO, NaOH, NH3). The resulting sulfite/bisulfite solu- tion can be discarded (these are "throwaway" processes) or regenerated (132, 177). Overall, the scrubbing process consists of: Gas-Liquid mass transfer from the stack gas, equilibria and reaction in the scrubbing mixture, modula- tion by Solid-Liquid transfer of absorbents (and possible Liquid-Solid trans- fer of products). The Gas-Liquid mass transfer is accomplished in several ways (90): spray tower, venturi (89, 98, 175), packed bed (159) and turbulent contact (moving ball) absorber (113). The efficiency of the G-L contacting step has been the subject of much study to assess major resistances (120, 178) and control them (142). Since S02 is chemically transformed as it enters (102) the liquid surface (from bubbles (104) or surface films (11, 46, 58)), the rates are best treated as absorption with reaction and described by parallel approaches viz., film theory (47, 85, 157), penetration theory (40, 41), or surface renewal (51, 52, 143). From coal fired boilers, stack gas entering the scrubber carries (53) '4 to 7% 02; various oxides of nitrogen; C02, CO, and organic residues; as well as inorganic residues entrained as fly ash (including chloride (29, 114) and metals (105) as well as inert particulates (64, 77). Some scrubbing methods remove particulates (69) as well as sulfur oxides, but these particulates (inorganic and other coal debris) influence the scrubbing process as catalysts and inhibitors of sulfite oxidation. See Table 1-1. The sulfur oxides rapidly (23) form a potentially reactive mixture des- cribed by the S02 dissociation equilibrium (99). The first ionization con- stant being 1.74 x 10"2 (167), the second 6.24 x 10~8 (189). The equilibria have been studied for S02/water systems (137, 169, 170) and scrubbing systems (124, 144).

---

------- TABLE 1-1. FLY ASH ANALYSIS Chemical Analysis Component Weight Percent A1203 20-30 Fe203 12-23 CaO 2-7 0.5-1.5 6

---

------- The most common absorbents (solids) are lime, limestone and dolomite found as crystalline aggregates (see Figures 1-3 and 1-2) and solid mixtures (78). Their solubility has been determined (65, 82) and noted in scrubber operation (29, 70). These soluble minerals provide the carbonates and hydrox- ides to fix the S02 in solutions. An analysis (43, 79) is given in Table 1-2. 2- The dissolved sulfur species HSCL and S03 are subject to oxidation in 2_ solution by dissolved oxygen producing SO^ . This gypsum is produced in lime/ limestone throwaway processes and can lead to supersaturation in all calcium-based processes causing scaling (29, 56, 77). Scale formation can be controlled by additive controlled oxidation (below 20%) (29) and gas pre- treatments. The oxidation occurs along the trajectory of the S02 as it enters the absorbing solution (127). Hydrodynamic conditions (151, 168) and mass transfer to the solution (169, 170) determine the region where reaction occurs. The limit of this region, called the reaction plane, determines what theory best describes the rate of absorption (80, 120). For instance, when the film is large (equal or greater than the diameter of the small particles of solid), dissolution can occur in the reacting zone, further increasing reaction and enhancing absorption (138). One approach to scaling control is limitation of oxidation by the addi- tion of an inhibitor (50, 77, 147). Many organic materials (especially those bearing hydroxyl groups) inhibit the oxidation. This action is not likely to be true catalysis (20, 50, 140), hence the inhibitors may be subject to con- sumption and would require replacement. Factors to consider with regard to inhibitors are: i Inhibitor must be water-soluble. Its actions may be modeled well enough to limit oxidation in part of the system.

---

------- Figure 1-2. Electron Micrograph of Calcium Sulfite Particles Sulfite and Sulfate Particles Figure 1-3. Electron Micrograph of Calcium Sulfite Particle Individual Agglomerate

---

------- TABLE 1-2. LIMESTONE ANALYSIS Chemical Analysis Component Weight Percent A12°3 Fe203 MnO CaO Cl Cu Cr As Hg 6.01 0.19 0.06 55.5 0.004 0.00044 0.00011 0.0002 6xlO"6

---

------- Products formed from inhibitors may also retard (or promote) the oxidation. The contacting method in the scrubber changes the effectiveness of inhibitors (43). Fly ash may contain phenolic organic residues active as inhibitors. Recently organic acids used to enhance limestone solubility were identi- fied as rate retarders. The most extensive set of experiments dealing with the organic acid effect of both the sodium and the calcium sulfite oxidation are reported by Hatfield, Kim and Mullins (81). It was reported that in the sodium system, organic acids promoted the sulfite oxidation with the order being: adipic > glycolic > no acid. However, in the calcium system organic acids were found to reduce the oxidation with the order being: glycolic > adipic > no acid. The results dealing with the calcium system are comparable to the results of the present work, dealing with the comparative inhibitory strength of the organic acids. The relative strength was: glycolic > adipic > succinic > no acid. There are large discrepancies between the current results and those of Hatfield et al. for the rate of oxidation of the sulfite. Hatfield et al. reported that the sulfite oxidation goes toward completion in several hours, but the results of this work show that the sulfite goes to sulfate in a matter of minutes (usually less than 20 minutes even with inhibitors added). Hatfield et al. added CaCO (calcium carbonate) to the reaction mixture for pH control and used an oxygen flow rate of only 20 ml/min. The present exper- iments used the addition of NaOH for pH control and a flow rate of oxygen at 3 A/min to maintain 0^ saturation in the reactor. This suggests that while the CaC03 or the NaOH may have an effect on the sulfite oxidation, the experiments by Hatfield et al. were oxygen limited (controlled by the oxygen flow rate 10

---

------- rather than the sulfite oxidation kinetics). These workers also determined the rate of reaction to increase with pH, which is in agreement with the current findings. In a recent work Altwicker (8) found hydroquinone caused a changing reduction in the oxidation rate which he attributed to inhibitor in the re- action plane being consumed quicker than it could be replenished, casting doubt on the sulfite method for measuring interfacial area in contactors. The scaling problem is greatest in the scrubber and mist eliminator. By 2- inducing oxidation elsewhere in the system, supersaturation in SO, will not happen. For this reason various approaches to forced oxidation in the holding tank beneath the scrubber have been studied (29, 71). For example: forced oxidation of the calcium sulfite reaction product in both lime and limestone FGD systems has been successfully demonstrated in 10 megawatt prototype units at the Shawnee Test Facility. The oxidized gypsum product (74) results in less disposal volume and settles by an order-of-magnitude faster than the unoxidized material. It filters to better than 80% solids and handles like moist soil compared with the unoxidized material which filters only to about 50 to 60% solids and is thixotropic (83). Improved settling is treated elsewhere as well (110, 124, 132, 161), and good sulfate removal has been discussed (29). The problem of scrubber scaling is eliminated in sodium-based liquors (56, 77, 110, 132, 161), but the added cost of sodium absorbents (49, 69, 110) necessitates regeneration of the "alkali values" of the sodium absorbent usually with a calcium base (lime/limestone) that is discarded. Recovery oriented, sulfur-concentrating processes are usually gaseous (27, 128, 165), but some processes (55, 92) use a liquor, operating by electro- lytic regeneration (92). The most prominent (68, 77, 103) two loop process is 11

---

------- the double alkali process (dual alkali), featuring (56, 96, 110) efficient, higher pH scrubbing with sodium liquor at low L/G rates, reduced scaling, and thorough limestone utilization, but may have disposal problems. Other companies which have utilized liquid phase oxidation in wet scrub- ber flue gas desulfurization processes include Babcock and Wilcox (magnesia-base wet scrubbing) (13), Wellman-Lord, Stone and Webster/Ionics (63), Aerojet-General Corporation (zinc oxide process) (6,92), and Monsanto (CALSOX system) (87). There are many other systems including ones based on magnesia (100, 158). Many models for the physical and chemical events underlying these removal processes have emerged beginning with the equilibria among the solution ionics (133) and a generalization of the oxidation in solution (109, 153). The real challenge is in formulating (47) the complex equilibria, dissolution rates, absorption rates (40), and reaction rates in a way suitable for computing concentration profiles, removal rates and efficiences, and absorbent effi- ciencies in the scrubber (114) and holding tank (175). Any description requires the reaction kinetics of sulfite oxidation in solution. Indeed, this topic has been the subject of multifaceted research for 132 years yielding sure findings only slowly. The oxidation is a free radical chain process subject to photochemical initiation, the quantum efficiency of which is 5 x 104 (16). The impurity level of the work that produced this figure raises doubt about its validity. Further work shows no influence of UV on the oxidation rate in aerosols of low pH (95). The most careful photochemical study concluded that there is no simple relation between the light absorbed and the rate (112). Experiments using various surfaces and particulate additives (86) found no variation with contact area of catalyst. This well establishes the reac- 12

---

------- tion as a case of homogeneous catalysis. Other workers (20) who exceeded solubility limits of catalyst suggested the subsequent particles might have provided sites for heterogeneous catalysis, but this is not the case (18). The effect of impurities is dramatic. The sensitivity of the reaction rate to stray metal ions is the hallmark of the reaction, but inorganic anions (188) as well as organic molecules affect the rate (although in the opposite way). In fact, the bane of much experimental work was rubber in the apparatus (48, 115, 140, 171). Pure gum rubber stops the reaction (112) giving evidence that the sulfur vulcanizing chemicals are not the cause. The degree of inhi- bition by hydrocarbons is so sensitive that oxidation has been suggested as a semiquantitative analysis for their presence (152). This sensitivity is an important consideration for operations in rubber-lined slurry handling equip- ment, such as those at Shawnee Valley. -12 The acceleration of the rate by metals is extreme: 10 M copper ion added during sulfite oxidation increased the rate (171). Many metals have been studied: CQ 7, 21, 42, 44, 54, 95, 107, 108, 109, 130, 126, 149, 163, 185, 187 r 7, 17, 21, 66, 84, 95, 109, 150, 140, 172 L»U Mn 21, 42, 48, 75, 86, 93, 95, 179 Fe 21, 39, 84, 86, 93, 95, 140 21, 42, 179 Mg 21, 39 Ni 21, 93, 140 Al Zn 21' 179 Ca 179 xra + 21, 45, 95, 115, 140, 152 JNI14 13

---

------- General comparisons have tested their relative effectiveness: Co, Ni > Cu > Fe ref 140 Mn > Co, Ni > Fe, Cu > Mg, Zn, Na, NH4+ ref 21 Mn » Zn, Mg > Ca ref 179 Mn > Cu > Fe > Co, NH > Na > uncat. ref 95 329 199 167 49 49 4 3 = Relative strength Mn -\. 7000 x Mg ref 42 The most effective catalysts are clearly manganese and cobalt with iron giving a notable effect. The metal showing the least effect is sodium (21, 91, 95). For this reason, the spectator ion is nearly always chosen as sodium. Am- monium, a constituent in many natural systems, increases the rate of oxida- tion. The probable cause is not catalysis, but rather its effect on pH may merely shift the equilibrium toward sulfite which is most rapidly oxidized. Several workers (20, 45, 136, 150, 152) have considered rionmetallic catalysis and mechanistic interactions of sulfur compounds. The very few experiments have no salient results. One curious observation is that sulfate formed by the reaction has less effect than initially added sulfate upon the reaction rate (111, 134). No satisfactory answer exists (42, 66, 141), but control of the position of the sulfur species equilibrium is one candidate. An early controversy concerned bisulfite oxidation rates increasing with dilution (17, 115, 119). The spur- ious effect was due to decreased oxygen mass transfer at higher concentrations (140). The oxidation in all pH ranges has been studied with wide concentration changes of sulfur species and catalyst and some slight changes of oxygen concentration. Many of the important results in the literature are compared Table 1-3. Although no firm results stand out, the many contrasts serve to 14 on

---

------- TABLE 1-3. COMPARISON OF THE LITERATURE OF SULFUR DIOXIDE OXIDATION Reference lloalhcr & Toodeve8'1 Coughanowr & Krause"8 Walter1" Powell135 Catipovlc1''2 Chen 6. Barion & Llnek & 108 Alper7 Relnders & Fc"et^ Phillips i Johnson DeWaal & Ok on 54 Uessellngh & LI nek & Tvrdlk109 Tek^ Saulckl f. Onda129 Vagi i I not187 Year 1934 1965 1972 1973 1974 1972 1966 1970 1973 1925 1941 1959 1966 1970 1971 1973 1973 1972 1962 S Species Cone, H Order II SO 5-50xlO-4 1 3 2.5xlO"5 0 SO.'H.O 1.7xlO"3 0 1 1.9x10-3 o S02'H20 0.009 0 HSOj" 0.03-.09 1 IISO." .001-. 01 0 SO 2' 0.009 1 S032" .009-. 03 3/2 S032" 0.04-0.4 3/2 S032" 0.25-0.8 0 SO 2" 0.8 „ 0.5-1.0 S032" 0.06 1 SO.2" 0.01-0.05 1 SO ,2~ .001-1.0 1 3 0 S032" 0.8 1 S032" 0.8 SOj2" 0.3-0.8 0 S032" 0.8 0 S032" 0.1-0.3 SO 2" 0.3-1 S032" 0.008 1 °2 Cone, H Order 2xlO"3 I 0 3-6xlO"4 0 4-8x10-4 0 1.2xlO'3 0 1.2xlO"3 l.lxlO-3 1-4x10-4 0 0-0.003 0 2-llxlO"4 1-2 ,3-2xlO"3 2 1-8x10-3 1 l.lxlO'3 l.lxlO"3 1-7,10-* \ l.lxlO"3 1 1-11x10"* 2 6.4xlO"4 1 SxlO-* 2 l.lxlO"3 2 .3-3xlO"3 1 2 2-llxlO"4 2— 1 l-15xlO"4 1 Cat Ident Cone, M Order Mn 3xlO"6-8xlO-5 2 2.7x10-4 2 Mn 1.8xlO-3-0. 18 2 Mn 1.8-9x10-5 2 MB 0.036-0.82 0.7 - Mn 6.6-19.9X10"6 l Co 10"7-3xlO"6 1/2 Cu ID'4 i Co 0-10° 1/2 C" 1°~3_5 .3 t Co 4x10 -10 Cu 10"5-10"3 1 Cu 10-9xl0'4 1 Cu .5 Cu.Veraene 10 Co 10"3 Co 4x10 * 1 Co 5xlO"6-10"3 1 Co 8.8xlO"6-10"3 1 Co 5xlO"7-10 6 1/2 Co 10"9-5xl06 .7 Co 0.5-7.0x10 1 Contacting Method 6 1 i r r 1 n« stirring II «. R "T" stirrinB stirring stirring H & R II & R stirring stirring Pack col stirring stirring low turb high turb Wet wall Met wall Stirring Wet Wall Wet Wall Wet Wall Pack col bubbling Temp 40°c l-s°r 25°C 25°C 25°C 25°C 25°C 25°C 25°C 35°C 25"c 10°C 25UC 25°C 30°C 30°C 30°C 25°C ^°r 30°C 25°C 30°C 20°C pll 7 5 - - 1-2 1-4 1.1 -10 - 8-9 9.2 B.I 10 R.7 9.2 7-9 8.5 7-9 7-9 10 8.5 8-9 k - m O7i in-7 ™ole .03-22.1x10 j gec 1.3-16x10-1° f-i^ n i i it ,n-lmole 0.3-2.46x10 ^ se- 1.66xlO'5sec"1 ,~-R m 0.053s*c-l l "C •> 7 nr^in-5 mole 3.2-66x10 j"sec 1200 1/g mole sec 19.3 sec l - - 0.013 sec * 2.5x10° I/mole sec - 50 — 104sec"1 - 1-5x10* sec"1 1.43x10 I/mole sec 2.3xl06 I/mole sec - -i.Ar52^— s o l-sec +.4 l+1.46xlO"7 Cs S 0 Comment A impurity likely Ea 32^4^. obs. mass trans ll.il"01* inert apparat u«, Dl)l gave great care to purity of mat'ls. Ea 18.7 kcal/mole Ea 17.5 kcal/mole Ea 18.3 kcal/mole no effect due to stirring st hip.h speed stirring changed 02 order NH^ present no effect of added acid inh drops rate to 10"s Versene likely effect on mech Cat solubility exceeded Ea 12 kcal/mole Ea 10.53 kcal/mole absorp- tion Indep of hydrodynamics 0 order change sharp Ea 15 kcal/mole added SO, : no effect Ea 12.4 \cal/mole 0- order switches with Cat Cone 0, order change pradual

---

------- define the main kinetic questions. The only problem comes from the varying conditions in the experiments which render some data uncomparable. The order in sulfur is usually zero or one, but this depends on the pH range. For low pH the HS03" order seems to be 0: at high pH the S032" order may be one. There are also changing order results. The oxygen order is the most controversial and has the most complicated fluctuations. The order has been observed to shift from zero to two as the sulfite concentration changes (12), switch between one and two as the catalyst concentration increases (148), and to vary as its own concentration changes (129). Whether the change is gradual (129) or abrupt (109) is a question. Catalyst dependence is stronger than that of the reactants, but the value is unsettled. Mn and Co produce orders of magnitude change in the reaction rate by variations of only a few ppm in their concentrations (48). Mg, con- versely, produces a sluggish, but steady, increase in the oxidation rate as its concentration changes over orders of magnitude (42). Experiments on the pH are scattered widely in the literature. The treat- ment by Fuller and Crist is worthy of note (66). (Linek and Tvrdik also treat the subject well). (109). Because the effect of pH is so closely linked with the sulfurous solution equilibrium, no clear chemical trends emerge. At extremely high pH ( >12) , however, the rate is depressed (19, 115, 190). This behavior is sometimes due to reduction in oxygen solubility (118). Never- theless there appears to be a genuine chemical effect, for the rate reduction occurs as well for oxygen already in solution (101). The energetics of the reaction are not as fully examined as the kinetics. All the works report operating temperatures between 20°C and 40°C. The low pH activation energy is around 20 kcal/mol, at high pH it is near 15 kcal/mol. 16

---

------- The published rates vary widely in first and second order kinetic equa- tions. The values appearing in Table 1-3 are for conditions of low pH where half-lives of a day are common as well as instances of high PH reacting to half concentration in 0.01 sec and less. In the low pH reaction, the stirring also causes discrepancies. Schultz and Gaden (155) report a decreasing reaction rate with increased stirrer speed. This finding touched off studies (131) that have still not resolved the ambiguity. Also at low pH no true (86) induction period has been ob- served, although start up lags due to physical causes do occur (152). At high pH an induction period up to 2 sec is possible (48). The above questions all touch on the molecular activity making up the reaction mechanism. And, just as these specific questions remain open, the final elucidation of the oxidation mechanism remains to be done. It is gener- ally accepted (66) that the oxidation proceeds by a free radical chain- with or without metal participation. Many aqueous sulfur radicals possibly in- volved are characterized (190). The early suggestions by Haber (76) and Titoff (171) have given way to the proposed mechanism of Abel (2, 3, 4) and Backstrom (14, 15, 16). A few points of interest to all mechanisms are: the oxygen atom transferred does not come from the solvent (186), the uncatalyzed reaction occurs by initiation other than by stray metal ions (66), the acti- vation energies in the catalyzed reaction agree closely with those for some metal-ligand substitutions of the catalyst (108). The results of some recent oxidation studies in slurries have given better definition to the oxidation problems. Ramachandran and Sharma proposed a model for gas absorption in a three phase slurry system. An instantaneous chemical reaction was assumed to occur in the liquid boundary layer surrounding the solid particles. Two cases were 17

---

------- examined in their model. The first case assumed that solid dissolution was unimportant in the rate of gas absorption. The second case assumed that solid dissolution into the liquid phase was an important factor. Results from the second case showed the rate of gas absorption to be proportional to the square root of the concentration of the solids present. Bjerle, Bengtsson, and Farnkvist (28) conducted an experimental examina- tion of CaC03 slurry oxidation in a laminar jet absorber at 25°C and 45°C and pH 8.5. In their weak (2%) slurry they found the S02 mass transfer coeffi- cient to be nearly that in an otherwise identical clear solution (clear kS02 = 1