United States Environmental Protection Agency Industrial Environmental Research Laboratory Research Triangle Park NC 27711 Research and Development EPA-600/S7-84-049 May 1984 <>EFV\ Project Summary Fundamental Study of Sulfate Aerosol Formation, Condensation, and Growth Shui-Chow Yung, Cumbum N. Rangaraj, Benjamin L. Hancock, Daniel Ugale, and Seymour Calvert A theoretical and experimental pro- gram was performed to study the for- mation and growth of sulfate particles. Existing theoretical models on acid par- ticle formation and growth were re- viewed and evaluated. The formation and growth of sulfate particles during slow cooling, rapid cooling, and dilution cooling of flue gas were experimentally determined and compared with theo- ries. The experimental results show that the temperature at which the self- nucleation of sulfuric acid occurs is lower than the acid dew point tempera- ture. Thus, if the flue gas is slowly cooled to between dew point and nu- cleation temperature, it is possible to force the sulfuric acid to condense out on surfaces, rather than forming fine particles. The theories, experimental methods, and results are described in this report. This Project Summary was developed by SPA's Industrial Environmental Re- search Laboratory, Research Triangle Park, NC, to announce key findings of the research project that Is fully docu- mented In a separate report of the same title (see Project Report ordering infor- mation at back). Introduction Primary sulfates are significant pollutants, contributing to the formation of acid rain, reduced atmospheric visibility, and human respirable diseases. To illustrate the mag- nitude of the sulfate problem, more than one-third of the airborne respirable particles in the Eastern U.S. are in the form of sulfate. A major source of atmospheric sulfate par- ticles is the chemical reaction of SO2 and par- ticles in the ambient air. However, control of secondary sulfates through direct control of S02 emissions has not generally resulted in a decrease in atmospheric sulfate. There- fore, it is worthwhile to consider control techniques for primary sulfates. Existing particle control systems do not ef- fectively remove condensible aerosols be- cause the aerosol precursors are often in the vapor state when they pass through the con- trol device. Although the vapors usually will condense in a wet scrubber, they often form ultrafine particles which are very difficult to capture. The first step toward developing accep- table technology for reducing condensible aerosol emissions is to obtain an adequate data base and understand the mechanisms involved: the condensation, formation, and growth of sulfate aerosols in a simulated flue gas environment. Under contract to the U.S. Environmental Protection Agency, A.P.T., Inc., performed a detailed theoretical and experimental study of sutfate particle formation and growth. The theories, experimental methods,and results are described in this report. Objectives This research was a theoretical and ex- perimental study of sulfate aerosol formation and growth under conditions that exist in in- dustrial smoke stacks and the near-stack plume. The objectives were to develop the fundamental data and mathematical models necessary to design emission control stra- tegies and control devices for sulfate con- densation aerosols. ------- Approach The general approach was first to define the mechanisms by which S02 and S03 are converted to sulfate particles and then use this knowledge to develop optimum control. S02 and S03 can convert to sulfate particles by: 1. Condensation of sulfuric acid vapor (or water-vapor-associated S03) to form sulfuric acid drops. 2. Condensation of sulfuric acid vapor on pre-existing particles such as fly ash, liquid drops, and acid drops. 3. Sorption of S02 by liquid or solid par- ticles, followed by oxidation of S02 to sulfate. 4. Gas-phase oxidation of S0a to S03 which is subsequently condensed by mechanisms 1 and 2. 5. Chemical reaction of acid vapor with solid or liquid particles. This study emphasizes sulfate formation by mechanisms 1 and 2. Mechanism 4 is not likely to happen in industrial smoke stacks, and mechanisms 3 and 5 are system depen- dent and difficult to generalize. Sulfate formation due to condensation is a physical process in which the gas tem- perature must be below the acid dew point. Gas can be cooled by heat transfer to the surroundings, quenching (such as by in- troducing water sprays), and mixing with cold gas. All three cooling processes could occur in industrial smoke stacks and in near- stack plumes. Therefore, sulfate formation in typical flue gas mixtures during slow cool- ing, rapid quenching, and dilution cooling was experimentally studied. The results ob- tained in this study plus published data were then used to verify sulfate formation models and develop emission control technologies. Experiments Apparatus Figure 1 shows the experimental system design, which was basically the same for all experiments. It consisted of a flue gas simulator for supplying acid-laden gas for various cooling apparatus arrangements. Major components of the flue gas simulator included an acid vapor generator, a steam generator, a fly ash particle generator, and a S02 gas cylinder. Sulfuric acid and water vapors were gen- erated by evaporating dilute sulfuric acid and water at controlled rates. Fly ash particles were produced by re-dispersion. S02 was metered into the flue gas simulator from the gas cylinder. Room air was used as the car- rier gas because the amount of oxygen and the presence of CO2 and nitrogen oxides in the flue gas have no effect on sulfate parti- cle formation in the stack. Air Double Pipe Heat Exchanger -Water Packed Bed I Air Gas Slow Cooling Experiments Dilution Chamber -Air Heater or Cooler Rapid Cooling Experiments "Gas Dilution Cooling Experiments Figure 1. Experimental system design. The following cooling apparatus was used in the experiments: 1. Slow cooling — double-pipe heat ex- changer. 2. Rapid quenching — packed-bed column. 3. Dilution cooling — parallel-stream dilu- tion chamber. Measurement Methods Gas samples at the cooling apparatus inlet and outlet were obtained simultaneously to determine: 1. Inlet acid vapor, water vapor, SO2, and fly ash particle concentration. 2. The amount of acid vapor condensed on existing particles. 3. Concentration of newly formed acid particles. 4. The amount of acid vapor condensed on walls. 5. Particle size distribution. Figure 2 shows the sulfate sampling sys- tem. Train "A" (used at the inlet) consisted of a cascade impactor (or quartz filter) followed by a condensation coil, a quartz filter, three impingers, and gas metering and moving instruments. Train "B" (used at the outlet) consisted of an impactor (with final filter removed) connected in series with a screen diffusion battery, a quartz filter, a condensation coil, another quartz filter, three impingers, and gas metering and moving in- struments. The instruments upstream of the condensation coil were heated to and main- tained at gas temperature. The water-cooled condensation coil was maintained at be- tween 60 and 90 °C when SO2 was injected. The acid collected by the impactor, diffu- sion battery, and filter was analyzed by ex- traction and titration. Acid collected by the condensation coil and impinger was deter- mined by washing followed by titration. Results Rapid Quenching Rapid quenching experiments simulated the formation of sulfate particles in a scrub- ber. The quencher used in this study was a randomly packed bed of Berl saddles. The effects of acid vapor concentration (10 to 100 ppmV), water vapor concentration (5 to 15% by volume), SO2 concentration (0 to 1.3 g/m3), and quench water temperature (20 to 60°C) on nucleated acid particle size distribu- tion were determined experimentally. The results are: 1. Higher water temperature and lower acid and water vapor concentrations resulted in higher concentration of fine acid particles. Low water temperature decreased the gas temperature and caused additional water vapor conden- sation on the nucleated particles. Decreasing the acid and water vapor concentration reduced the particle growth. 2. The presence of SO2 had no effect on acid-bearing particle nucleation. ^ 3. In the presence of fly ash particles, • acid-bearing particles were larger ------- Nozzle \ Impactor or Filter Heated i Condenser i 1 [UL i i lilr _ V 1 T SV-* & To Impingers and -EPA Method5 Sampling Train Constant Temperature Water Bath Nozzle \ Train "A" I Casacade Impactor uiiiuxionouiieiy , To Impingers 4 IYI IVI IVl PVl L*J ! Ur^ i M-i_an Samolina 1 l^j 1^1 1^1 ixxi q . |TS [ \\T^j ~"-i""-u niter : t leaf Enclosure Train "B" ±2] \Matar Path Figure 2. Acid concentration and particle size distribution sampling system. because of condensation of acid vapor on the particles. Rapid quenching caused local high supersaturation of acid vapor; therefore, many very small acid particles were nucleated. 4. Downstream from the quencher, par- ticle growth due to agglomeration was negligible. Dilution Cooling Dilution cooling experiments simulated the formation of sulfate particles in the near- stack plume. The cooler was a concentric parallel-stream dilution chamber with dilution air forming a sheath around the flue gas. The effects of dilution air temperature, water and acid vapor concentrations, and the presence of S02 and fly ash particles on acid nuclea- tion were studied. The results are: 1. The concentration of fine acid particles decreased with increasing dilution air temperature, decreasing acid vapor concentration, and increasing water vapor concentration. 2. S02 had no effect on acid particle nucleation. 3. The concentration of fine acid particles decreased in the presence of fly ash particles. Slow Cooling Slow cooling experiments were relevant to the acid particle nucleation which could occur in industrial smoke stacks, dry pollu- tion control devices such as electrostatic precipitators, and baghouses. The experi- ments were performed on a co-current, double-pipe heat exchanger with flue gas flowing in the center pipe. Acid particle samples were taken where nucleation was observed to start. The experimental findings are: 1. Over 80% of the acid vapor was con- densed on preexisting fly ash particles. 2. Higher acid vapor and lower water vapor concentrations led to the forma- tion of many ultrafine acid particles. 3. SO2 had no effect on acid particle nucleation. 4. Much acid vapor was condensed on walls upstream of the nucleation start- ing point. Comparison Between Theory and Experiments Nucleation Temperature The nucleation of sulfuric acid particles in- volves two condensible species, water and su If uric acid. For a given water and acid vapor concentration combination, the tem- perature at which nucleation of sulfuric acid particles starts can be predicted from ther- modynamics and classical kinetic theory. In this study, the simplified equation used to predict the nucleation temperature was de- fined as the temperature which gives a nucleation rate of l/cm3-s. The calculations indicate that nucleation, once started, goes to completion quickly. Acid vapor is depleted in a very short time. The predicted nucleation temperature in- creases with increasing water vapor and acid vapor concentrations as shown in Figure 3. This means that less cooling of flue gas is needed to initiate the formation of acid par- ticles if the acid vapor and water vapor con- centrations are high. Table 1 shows the predicted and measured nucleation temperatures. The measured acid particle nucleation temperatures varied around the predictions. Considering the un- certainties in vapor pressure and surface ten- sion data, the agreement between theory and experiment is considered to be good. Slow Cooling A mathematical model was developed to characterize the nucleation, condensation, and growth of surfuric acid particles in a flow system. The model accounts for gas cool- ing, vapor loss to the wall, homogeneous binary nucleation, heterogeneous condensa- tion on existing particles, diffusional loss of nucleated particles, and coagulation. A com- puter program, written to apply the model to slow cooling experiments: (1) uses a quasi-steady state approach, (2) follows a small element of fluid through the pipe, and (3) computes changes in temperature, vapor concentration, and particle concentration due to the above mentioned mechanisms in successive increments of distance along the pipe. The measured gas cooling rate and wall condensation of acid vapor were compared to those predicted from heat and mass transfer. Because of entrance effects, the measured cooling rate and wall condensa- tion of acid vapor were much higher than calculated. Capture Strategy Implications The flue gas dew point temperature has been measured and predicted by numerous investigators. Although there are differences in dew point temperature given by different investigators, they are all much higher than the acid particle self nucleation temperature found in this study and predicted from classical nucleation theory (Figure 4). Thus, if the flue gas is slowly cooled to between dew point and nucleation temperatures, it is possible to force the sulfuric acid to con- dense out on surfaces, such as on fly ash particles and walls, rather than form fine par- ticles. The experimental results obtained in this study show that this control method is feasible. Avoiding rapid cooling of the flue gas minimizes acid particle nucleation. When flue gas much be quenched, the quench water should be as cold as possible. It would also help if additional water vapor, such as waste steam, is introduced into the gas stream (as in F/C, flux-force/condensation). Conclusions Conclusions from this study are: 1. The temperatures at which the self- nucleation of sulfuric acid particles starts can be predicted from theory. ------- Table 1. Predicted and Measured Nucleation Temperatures Inlet Run No. 116/01 116/02 116/04 116/05 117/01 117/02 117/03 118/01 118/02 119/01 119/02 Acid Vapor Cone., ppm 83.4 49.9 52.1 61.9 29.0 40.2 35.5 22.1 27.6 22.0 31.4 Water Vapor Cone., Vol. % 10 5.5 4.7 14.5 10.1 5.0 13.8 10.1 14.0 10.1 4.8 Acid Vapor Cone., ppm 65.2 39.6 46 60.4 11.6 19.9 26.1 19.3 17.3 4.6 17.7 At Nucleation Point Water Vapor Cone., Vol. % 10 2.8 2.9 14.5 4.9 2.5 12.2 9.6 11.6 10.6 3.1 Gas Temperature °C 106 96 99 109 104 86 103 96 103 91 101 Predicted Nucleation Temperature °C 105 97 98 113 95 91 105 98 102 99 90 120 110 o 3 ! I 100 I 90 Percent by Volume Water Vapor 80 Figure 3. Predicted nucleation temperature. and they agree with our experimental results. Thus, the sulfuric acid nuclea- tion temperature for flue gas can be predicted, if the gas pressure, water vapor concentration, and S03 concen- tration are known. 2. Avoiding rapid cooling of flue gas minimizes the formation of fine acid particles. 3. While the presence of S02 has no ef- fect on acid nucleation, fly ash particles 160 150 140 130 3 120 5 QJ 100 90 80 Verhog and Banchero Lisle and Sensenbaugh Dew Point Temperature Predicted Self-Nucleation Temperature i 0 10 20 30 40 50 60 7O 80 90 100 Concentration, ppmv have great effect. Most of the acid vapor condenses on ash particles if the gas is cooled slowly. 4. Increasing the water vapor concentra- tion in the gas stream increases the acid particle diameter by additional condensation or solution-induced par- ticle growth. 0 10 20 30 40 50 100 Sulfuric Acid Vapor Concentration, ppm Figure 4. Dew point and acid nucleation temperatures. ------- S-C Yung. C. N. Rangaraj. B. L Hancock. D. Ugale, andS. Calvertare withA.P.T., Inc., San Diego, GA 92109. Leslie E. Sparks is the EPA Project Officer (see below). The complete report, entitled "Fundamental Study of Sulfate Aerosol Formation, Condensation, and Growth," (Order No. PB 84-179 886; Cost: $ 17.50, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield, VA 22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Industrial Environmental Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC 27711 ------- United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 PS 00005£S> U S ENVIR PROTECTION AGENCY KEGIUN 5 LiHRAKf d$0 5 CHICAGU IL U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/957 ------- |