Emissions of Sulfur Trioxide from Coal-Fired Power Plants
R.K. Srivastava and C.A. Miller
Office of Research and Development, National Risk Management Research Laboratory, Air Pollution Prevention and Control Division, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina
C. Erickson and R. Jambhekar
Babcock Borsig Power, Inc., Worcester, Massachusetts
ABSTRACT
Emissions of sulfur trioxide (SO3) are a key component of plume opacity and acid deposition. Consequently, these emissions need to be low enough to not cause opacity violations and acid deposition. Generally, a small fraction of sulfur (S) in coal is converted to SO3 in coal-fired combustion devices such as electric utility boilers. The emissions of SO3 from such a boiler depend on coal S content, combustion conditions, flue gas characteristics, and air pollution devices being used. It is well known that the catalyst used in the selective catalytic reduction (SCR) technology for nitrogen oxides control oxidizes a small fraction of sulfur dioxide in the flue gas to SO3. The extent of this oxidation depends on the catalyst formulation and SCR operating conditions. Gas-phase SO3 and sulfuric acid, on being quenched in plant equipment (e.g., air preheater and wet scrubber), result in fine acidic mist, which can cause increased plume opacity and undesirable emissions. Recently, such effects have been observed at plants firing high-S coal and equipped with SCR systems and wet scrubbers. This paper investigates the factors that affect acidic mist production in coal-fired electric utility
boilers and discusses approaches for mitigating emission of this mist.
INTRODUCTION
As understanding of the adverse effects of air pollution has grown, so also has the complexity of coal-fired power plant design and operation, especially with regard to air pollution control systems. Control of air pollutant emis- sions is not only a legal requirement but also is becoming a financial necessity, as salability of effluents and trading of emissions increase the direct monetary value of emis- sions control. The days when one must only consider the nuisance value of fly ash are long past.1
As plant complexity has increased, so have unex- pected consequences of changing segments of the total chemical process that occurs between fuel preparation and ultimate emissions. One of the more discernible ad- verse consequences is the formation and emission of sul- fur trioxide (SO3)/sulfuric acid (H2SO4), as highlighted by the recent and well-publicized experiences of a power plant in Ohio.2Although not directly subject to emission limits, SO3is important to consider during the design and operation of coal-fired utility boilers for a number of environmental and plant performance reasons.
The formation of SO3will occur during the combus- tion of sulfur (S)-bearing fuels such as coal and heavy fuel oils. Virtually all of the SO3converts to H2SO4as flue gas is cooled in the air preheater (APH). Relatively high con- centrations of SO3/H2SO4in the boiler, stack, or plume can cause adverse impacts to plant equipment and to the environment. Impacts on plant equipment can include corrosion, fouling, and plugging and may require addi- tional hardware or changes in operation to minimize SO3/ H2SO4concentrations and the resulting adverse impacts.
Health and Environmental Effects
Formation of visible H2SO4 droplets depends upon the concentration and dew point of H2SO4 and the IMPLICATIONS
Formation and emissions of SO3/H2SO4 can potentially lead to plant operation-related problems and environmental concerns. The emissions of SO3from a boiler depend on a number of interacting complex factors including coal sulfur content, combustion conditions, flue gas characteristics, and air pollution devices being used. Understanding the paramenters leading to excessive generation of SO3/H2SO4 emissions can assist in selection of practical control ap- proaches. This paper investigates the factors that affect SO3/H2SO4emissions from coal-fired electric utility boilers and discusses approaches for mitigating these emissions.
The paper is expected to be a state-of-the-art reference on the subject for regulators, utility industry personnel, and others stakeholders.
concentration of sub-micron particles upon which the acid can condense. As seen in Figure 1, H2SO4dew point is a function of water vapor and H2SO4concentration and increases as both these variables increase.3The curves in Figure 1 determine the fraction of H2SO4in the vapor and condensed phases for temperatures below the dew point.
For example, a 250 °F flue gas with 10% water vapor and 8 ppm H2SO4will have a vapor-phase acid gas concentra- tion of ⬃2 ppm, with the remaining 6 ppm in the con- densed phase.
If temperatures at the stack are low enough and water vapor and H2SO4 concentrations are high enough, the condensed-phase concentration of H2SO4can be at a level that results in the formation of a visible plume attached to the stack. Once the gases leave the stack, the rate of cooling and subsequent H2SO4 condensation competes with the plume dilution by entrainment of ambient air into the plume. Even in cases where stack conditions are such that the H2SO4is completely in the vapor phase at the stack exit, a detached plume may still form shortly downwind as temperatures drop below the dew point.
In most cases, and particularly in the case where H2SO4, water, and sub-micron solid particles are present, condensation is the dominant formation mechanism.
The opacity of these plumes is most strongly influenced by the concentrations of condensing species and sub- micron fly ash particles present in the stack gases. The size distribution of the sub-micron fly ash also can have a noticeable effect on plume opacity at low to moderate H2SO4concentrations, because these particles act as con- densation sites for the condensing vapor-phase H2SO4. The relationship between sub-micron particle concentra- tion, H2SO4concentration, and plume opacity is shown in Figure 2 for flue gas with a sub-micron particle mode at 0.15m diameter.4,5The gray area denotes a typical range
of sub-micron particle concentrations downstream of an electrostatic precipitator (ESP).6
The curves in Figure 2 illustrate that H2SO4concen- tration can have a strong impact on opacity when fine particle concentrations are present at realistic levels. Be- cause the condensing acid particles can nucleate to form particles even at very low levels of pre-existing sub- micron particles, further reduction of sub-micron solid particles may not significantly reduce plume opacity. As an example, a unit emitting 16 ppm of vapor-phase H2SO4and 5 mg/m3of sub-micron particles is predicted to result in a plume opacity of⬃35%. An opacity of 20%
can be achieved by reducing sub-micron PM concentra- tion to⬃1.6 mg/m3(i.e.,⬃70% reduction) or by reducing H2SO4 to ⬃8 ppm (a 50% reduction). For units with limited ability to further reduce PM emissions, control of SO3 (and, subsequently, H2SO4) may be the only viable option for meeting plume-opacity requirements. Because essentially all the SO3is converted to H2SO4at or below stack temperatures, Figure 2 also illustrates the impact of increases in SO3formation on plume opacity. A unit with 1 mg/m3 of sub-micron particles in the stack gases is predicted to experience doubling of plume opacity from
⬃5 to 10% when H2SO4concentrations increase from 5 to 10 ppm.
Data on direct adverse effects on human health are inconsistent, but studies indicate that aqueous acidic aerosols at typical and even elevated ambient concentra- tions generally have minimal effects on symptoms and mechanical lung function in young healthy adults. How- ever, there are studies that have shown changes in muco- ciliary clearance and modest bronchoconstriction in asth- matics exposed to elevated concentrations of 400g/m3 or more.7,8Of more concern than the effects of exposure Figure 1. Dew point of H2SO4as a function of H2SO4concentration
for different water vapor concentrations.3
Figure 2. Iso-opacity plot calculated for a bimodal distribution of coarse and sub-micron particles. The sub-micron particle mode is at 0.15m diameter.5The gray area represents typical sub-micron mass concentrations in the stack of a coal-fired utility boiler.6
to typical ambient levels of SO3/H2SO4 aerosol alone is exposure to this aerosol mixed with other ambient PM constituents, including elemental carbon and metals.9
In situations where SO3/H2SO4 aerosol combines with a sinking plume, the ambient concentrations near the stack can reach significantly higher levels than those normally experienced. In such cases where meteorologi- cal and operating conditions combine to form a near- stack acidic fog, damage to property and vegetation can occur if the conditions are sustained for an extended period. Anecdotal evidence of adverse health impacts, such as burning eyes, sore throats, and headaches, has also been reported in such cases.2
Impacts on Plant Hardware
The most common problem associated with elevated con- centrations of SO3is low-temperature (⬍300 °F [150 °C]) corrosion.10Once formed, SO3reacts easily with the mois- ture in combustion flue gases to produce H2SO4. Below its dew point, H2SO4condenses and will collect in relatively low-temperature areas of the flue gas path, such as the air heater, and corrode the contacted metal components. The dew point varies with H2SO4and water concentrations in the flue gas (see Figure 1), but is typically between 200 and 300 °F (95 and 150 °C, respectively), corresponding to low (8%) moisture, low (0.1 ppm) H2SO4conditions and high (15%) moisture, high (20 ppm) H2SO4conditions, respec- tively. At higher temperatures (above 1000 °F [540 °C]), SO3can cause corrosion in superheaters and reheaters.11 These high-temperature reactions are less common than the low-temperature condensation and corrosion reac- tions. In addition to forming corrosive H2SO4, SO3reacts to form sulfate (SO42⫺) particles, especially ammonium sulfate [(NH4)2SO4] and ammonium bisulfate (NH4HSO4), in boilers using ammonia (NH3), which can lead to foul- ing or plugging of APH passages.12Regular maintenance is conducted at plants to address the impacts described. This maintenance can include monitoring and replacement of materials, washing of APH passages, and other activities.
In some cases, modification or addition of hardware may be required to minimize corrosion or plugging. At plants using wet flue gas desulfurization (FGD) systems, the reduction in flue gas temperature can result in in- creased H2SO4 condensation and subsequent corrosion.
In such cases, the flue gas may require reheating to ensure it remains above the dew point, thereby resulting in in- creased operation costs.13
REGULATORY OVERVIEW
There are currently no U.S. regulations that directly limit emissions of SO3/H2SO4 aerosols from utility boilers.
Other regulatory programs may, however, require SO3 control, as discussed later. German emission standards
limit the combined sulfur dioxide (SO2) and SO3emission concentrations to a daily average of 50 mg/Nm3 and a 30-min average of 200 mg/Nm3. There is no separate emission limit for SO3alone.14
In the United States, SO3emissions are included as part of limits on opacity and particulate matter (PM) emissions, and H2SO4is listed as a hazardous air pollutant (HAP) under Title III of the Clean Air Act Amendments of 1990.15Under Title III, electric utility generating stations are exempt from the mass emission limits of 10 t (9.09 tonnes) per year of any single HAP or 25 t (22.7 tonnes) per year of combined HAPs set for other industrial sources. Thus, although H2SO4 is a listed HAP and is emitted in significant quantities by coal-fired utility gen- erating stations,16there is no regulatory requirement to control these emissions as HAPs. They must, however, be reported to the U.S. Environmental Protection Agency (EPA) under the Toxics Release Inventory (TRI) require- ments. The H2SO4is typically measured and reported as equivalent SO3.
EPA has defined primary PM as particles that are either emitted directly as a solid or liquid or are emitted as a vapor but condense or react upon cooling and dilution in the ambient air to form solid or liquid PM immediately after discharge from the stack.17 Secondary PM is com- posed of particles that form through chemical reactions in the ambient air well after dilution and condensation have occurred.18Under this definition, SO3or H2SO4that is in the vapor phase in the furnace or stack and either reacts to form SO42⫺particles or condenses to form liquid H2SO4 immediately after leaving the stack is considered part of primary PM emissions and is therefore subject to PM emission limits.
In some states or localities, plants may be required to control SO3to maintain opacity standards. Even though opacity measurements in the stack may be within regula- tory limits, the formation of a visible plume caused by the presence of SO3may result in plume opacities higher than allowable levels and a need to control SO3 emissions.
Most states have a 20% opacity limit for stationary sources, including coal-fired utility boilers, and many also have provisions for higher opacity levels during unit startup. There are some exceptions to the 20% limit, with a few states having opacity limits of 40% and at least one (West Virginia) with a 10% opacity limit.
The formation of a detached plume is indicated when in-stack opacity measurement values are lower than those in the downwind plume. While in-stack opacity measure- ments are made using continuous emission monitors, measurements of the downwind plume opacity are cov- ered under EPA Method 9.9The increase in opacity from in-stack measurements to downwind is often caused by the condensation of vapors such as H2SO4and can result
in regulatory violations and good neighbor policy con- cerns.
FORMATION OF SO3
Formation in the Boiler
The S in coal has inorganic and organic components. The inorganic component is predominantly pyrite, which can exist both as distinct particles (excluded) and bound in the coal matrix (included). The organic component is part of the various organic structures present in coal. The fractions of included and excluded pyrite and organic component delivered to a furnace are known to be depen- dent on the fineness to which the coal is ground. During the combustion of coal, virtually all the S gets oxidized to gaseous species such as SO2and SO3, with SO2being the principal oxide. Although detailed chemical mechanisms for oxidation of fuel S are not understood well at present, it is believed that this oxidation proceeds through rapid formation of SO2, occurring on a timescale comparable to that of fuel oxidation reactions. Because SO2formation is so rapid, its concentration can be estimated using equi- librium calculations.
The primary reaction that results in SO3formation in flames is
SO2⫹ O ⫹ M 3 SO3⫹ M (1)
This recombination reaction proceeds rapidly near the combustion zone in the presence of super-equilibrium concentrations of oxygen (O) atoms. The consumption of SO3near flames occurs primarily via
SO3⫹ HO23 HOSO2⫹ O2 (2)
and
SO3⫹ O 3 SO2⫹ O2 (3)
A few other reactions are considered to result in formation and destruction of SO3near the flame zone, but the effect of these on SO3concentrations is relatively minor. Table 1 summarizes the reactions involving SO3in the combus- tion zone.19
To understand SO3 formation, calculations were made using the S chemistry described previously with gas compositions and temperature-time history characteristic of a coal-fired boiler.20This history is shown in Figure 3.
Initial gas-phase compositions are assumed to result from complete combustion (with 10 and 30% excess air) of a Montana sub-bituminous and a Western Kentucky bitu- minous coal. The sub-bituminous coal is composed of 53.26% (by weight) carbon (C), 3.35% hydrogen (H),
0.87% nitrogen (N), 0.78% S, 9.34% ash, 21.23% water, and 11.16% O. The bituminous coal contains 69.79% C, 4.79% H, 1.34% N, 2.95% S, 7.47% ash, 5% water, and 8.65% O. The initial compositions resulting from com- plete combustion of the coals are then equilibrated at 1647 °C (1900 K) to provide the input composition for the chemical kinetic calculations based on the S chemistry.
Note that the H2-O2-carbon monoxide (CO) subset of the hydrocarbon mechanism given in ref 21 is used in these calculations. This subset includes 25 reactions involving 11 species.
Figures 4 and 5 plot SO3mole fraction and the ratio SO3mole fraction/SO2mole fraction versus time, respec- tively. Because reaction 1 is exothermic, little SO3 is formed initially at high temperatures in the furnace. As the flue gas cools, SO3is produced, and the rate of pro- duction is high in the convective region of the boiler (see Figures 3 and 4). SO3production via gas-phase S kinetics is complete before the flue gas enters the economizer. The results also reflect that SO3production increases with coal Table 1. The reactions involving SO3in the combustion zone. The rate constants are listed in the form k⫽ ATexp(⫺EA/RT ) with units in mol-cm-sec.
Reaction A  EA/R
SO3⫹ H ⫽ HOSO ⫹ O 2.5e05 2.92 25,300
SO3⫹ O ⫽ SO2⫹ O2 2.0e12 0 10,000
SO3⫹ SO ⫽ SO2⫹ SO2 1.0e12 0 5000
SO2⫹ O ⫹ M ⫽ SO3⫹ M 9.2e10 0 1200
SO2⫹ OH ⫽ SO3⫹ H 4.9e02 2.69 12,000
HOSO2⫽ SO3⫹ H 1.4e18 ⫺2.91 27,600
HOSO2⫹ O ⫽ SO3⫹ OH 5.0e12 0 0
HOSO2⫹ OH ⫽ SO3⫹ H2O 1.0e12 0 0
HOSO2⫹ O2⫽ SO3⫹ HO2 7.8e11 0 330
Figure 3. Temperature-time history for a coal-fired power plant.20
S content and furnace excess air level. This is because increases in the S content of coal or furnace excess air level result in a corresponding increase in the concentra- tion of SO2 and/or O, which, in turn, results in greater production of SO3via reaction 1. Finally, Figure 5 reflects that SO2conversion rate (i.e., SO3/SO2molar ratio) for the modeled coals ranges between 0.1 and 0.65%, approxi- mately, and is relatively independent of coal type at a specific excess air level. Because boilers are operated with between 10 and 30% excess air, a 0.65% conversion of coal S to SO3 should provide a conservative estimate of SO3production in the furnace.
In addition to the SO3formation in the furnace dis- cussed above, additional formation takes place in the temperature range of 1100 – 800 °F (593– 427 °C) found in the economizer region of the boiler. This formation re- sults from oxidation of SO2 via molecular oxygen (O2) catalyzed by iron oxides present in both ash and tube surfaces.22This oxidation mechanism depends on several operating and design parameters, including SO2concen- tration, ash content and composition, convective pass surface area, gas and tube surface temperature distribu- tions, and excess air level. Because the impact of these parameters depends on site-specific factors (e.g., cleanli- ness of tube surfaces), it is difficult to characterize the
extent of SO3formation caused by catalytic oxidation of SO2. A laboratory study found that conversion of SO2to SO3in the presence of fly ash ranged between ⬃10 and 27% for temperatures between 500 and 700 °C, and es- sentially increased linearly with the iron oxide content of the ash.23These conversion rates are substantially higher than those found in coal-fired boilers, for which data suggest furnace/economizer conversion to be ⬃0.8 to 1.6% for bituminous and 0.05– 0.1% for sub-bituminous coals.24 The above laboratory study results indicate that temperatures and residence times greater than, and car- bon ash contents lower than, those typically found in coal-fired utility boilers are required to achieve high oxi- dation rates.
Formation of SO3in SCR Reactors
The SCR technology is increasingly being used at power plants to control emissions of oxides of nitrogen (NOx). In the SCR process, NH3is injected into the flue gas within a temperature range of ⬃600–750 °F (315–400 °C), up- stream of a catalyst. Subsequently, as the flue gas contacts the SCR catalyst, NOx, which predominantly is nitric ox- ide (NO) in combustion devices, is chemically reduced to molecular nitrogen (N2). In the most commonly used SCR Figure 4. SO3produced during coal combustion.
process layout, known as hot-side SCR, the catalyst is located between the economizer and the APH.
It is well known that the catalyst used in the SCR technology oxidizes a small fraction of SO2in the flue gas to SO3:25–29
SO2⫹ 1⁄2O23 SO3 (4)
The extent of this oxidation depends on the catalyst for- mulation and SCR operating conditions. Generally, this oxidation can range from 0.25 to 0.5% of SO2in bitumi- nous and from 0.75 to 1.25% in low-S sub-bituminous coal applications. To examine what this oxidation means, consider an SCR application with an SO2oxidation guar- antee of 0.5%. Also assume that the concentrations of SO2 and SO3in the flue gas at the inlet of the SCR reactor are 2000 and 20 ppm, respectively. Then, based on oxidation across the catalyst, the concentrations of SO2and SO3at the exit of the SCR reactor will be 0.995⫻ 2000 ⫽ 1990 ppm and 2000⫺ 1990 ⫹ 20 ⫽ 30 ppm, respectively. Thus, the SO3 loading in the flue gas at the exit of the SCR reactor will be 50% more than that at the inlet. This
example illustrates that the level of oxidation across the catalyst can have a significant impact on SO3concentra- tion.
In general, for a given catalyst material experiencing the same flue gas conditions, the oxidation rate of SO2to SO3(or conversion rate) is inversely proportional to area velocity (AV), which is simply the ratio of flue gas volu- metric flow rate to geometric catalyst surface area. This implies that the conversion rate is proportional to catalyst volume (and, hence, geometric surface area) and gas res- idence time in the catalyst. Thus, the conversion rate,, can be expressed as
⫽ K/AV (5)
where the constant of proportionality, K, is a function of catalyst material and design as well as flue gas proper- ties;30that is,
K⫽ f(catalyst material, catalyst design,
flue gas properties) (6) Figure 5. SO3conversion during coal combustion.
A systematic study of oxidation of SO2to SO3over hon- eycomb SCR catalysts has been reported in ref 31. The findings of this study are as follows:
(1) The conversion rate depends primarily on the vanadium content of the catalyst and, therefore, can be controlled by adjusting this content;
(2) The oxidation reaction is considerably slower than diffusion of SO2 within the pores of the catalyst. Therefore, the entire volume of the cat- alyst is active in oxidation of SO2to SO3in con- trast to reduction of NOx to N2, which, being diffusion-limited, occurs mainly at the catalyst surface. The rate of oxidation is linearly propor- tional to catalyst wall thickness. Accordingly, re- ducing the wall thickness should not affect NOx reduction but should reduce SO2to SO3conver- sion;
(3) The reaction rate is of variable order in SO2con- centration, increases with temperature, is inde- pendent of concentrations of O2 and water in practical applications, is strongly inhibited by NH3and is slightly enhanced by NOx.
SO3 DEPLETION OR CONVERSION TO H2SO4 Processes in APHs
Utility boilers use APHs to transfer heat from hot flue gas exiting the economizer to combustion air flowing into the boiler. These APHs are available in rotary regenerative and tubular designs, with the former used more widely. Typ- ically, the flue gas temperature at the APH inlet is between 600 and 700 °F (316 and 371 °C) and⬃300 °F (149 °C) at the exit. SO3is hygroscopic and, therefore, absorbs vapor- phase moisture at temperatures above its dew point to form H2SO4vapor.22This process occurs in the APH. The extent of conversion of SO3 to H2SO4 depends on the temperature distribution in the APH and flue gas moisture content. However, virtually all SO3converts to H2SO4at temperatures of 400 °F or less. If local metal temperatures in the APH flow passages drop below the acid dew point, some H2SO4condenses on these surfaces as liquid drop- lets (aerosol). This rate of condensation is dependent on the wall temperature and H2SO4concentration in the flue gas.
In a regenerative APH, where flow passages are peri- odically exposed to hot flue gas and relatively cold incom- ing combustion air, evaporation of condensed H2SO4oc- curs on exposure to air. The rate of this evaporation is dependent on the moisture content of air and metal sur- face temperature of the APH flow passage.32,33
Results of an extensive field test program indicate that⬃40% of the flue gas SO3present at the regenerative APH inlet is removed in the APH by the condensation- evaporation mechanism discussed above.34 The tubular
APH design does not have surfaces that are periodically exposed to combustion air; except for small leakages in welds and seal, no mixing of the flue gas and combustion air occurs. Therefore, the resultant H2SO4formed is not removed by evaporation into the combustion air and passes directly out of the APH. Modeling of the acid con- densation phenomenon in tubular APHs is described in ref 35.
In addition to the condensation-evaporation of SO3 discussed above, an additional conversion process takes place in boilers using SCR. In such applications, NH3 is injected as a reagent in the SCR process. A minor fraction, 2–5 ppmv, of injected NH3slips past the SCR catalyst and does not react with NOx. This fraction of NH3, known as NH3slip, reacts with SO3downstream of the SCR reactor and forms (NH4)2SO4 and NH4HSO4salts and, thereby, results in removal of SO3 from the flue gas.12 This salt formation can be detrimental to the APH performance if APH passages become plugged and pressure loss across the APH results in forcing offline washing. The amount and type of salt formed will depend on the amount of NH3 slip; based on typical concentrations of SO3and H2O in flue gas, the amount of NH3slip is the limiting factor in salt-formation reactions.
Localized Condensation of H2SO4in the Duct between APH and PM Control
Condensation of H2SO4 also can take place in the duct between the APH and the PM control device.35,36 For example, because of the rotating heat transfer element employed in the regenerative APH, gas flow stratification across the flow cross-section at the APH inlet is enhanced in the APH. As a result, strong transverse variations in gas temperature and H2SO4vapor concentration can exist in the gas leaving the APH. These variations can lead to localized condensation of H2SO4. Based on operating con- ditions, localized condensation also can occur in units using tubular APHs. For a given combination of flue gas H2SO4concentration and moisture content, Figure 1 can be used to determine if condensation is occurring.
Fly Ash Adsorption and Removal of H2SO4in PM Control Equipment
In addition to the processes described, some H2SO4gets adsorbed on fly ash in the APH and downstream equip- ment. The rate of this adsorption depends on the temper- ature of the flue gas, concentration of H2SO4, and fly ash properties, in particular alkalinity.23,37 The adsorption may increase rapidly as the flue gas reaches the cold end of the APH and may continue in the duct between the APH and the PM control equipment. The adsorbed H2SO4 gets removed with the fly ash in the PM control device (ESP or baghouse).
The firing of sub-bituminous coals, which generally have S contents on the order of 0.5%, results in fly ash with a relatively high amount of alkali (20 –30% by weight). Such alkaline ash adsorbs virtually all H2SO4in the flue gas.38In such cases, SO3injection is required for improving ESP performance. In contrast, based on fly ash properties and temperature, the majority of SO3in flue gas of a bituminous coal-fired boiler may or may not be adsorbed in fly ash.30 Finally, hot-side ESPs operate at high-enough temperatures where little adsorption of H2SO4occurs.
As mentioned above, adsorption of H2SO4on fly ash depends on the alkalinity of the fly ash and the concen- tration of H2SO4. Figure 6 illustrates the general adsorp- tion characteristics of fly ash as a function of the molar ratio of alkali content of fly ash and SO3concentration at the inlet of the APH. The alkali content of fly ash is defined as the molar sum of magnesium oxide (MgO) and calcium oxide (CaO) in the ash. This figure represents a correlation of data from Babcock Borsig Power field-test- ing programs concerning SO3production and capture in flue gas systems.
Aerosol Formation in Wet FGD Systems As the H2SO4vapor-containing flue gas passes through a wet FGD system, it is rapidly cooled below the acid dew point. Because the rate of this cooling is greater than the rate of absorption of H2SO4vapor in the scrubber solu- tion, the dew-point crossover results in H2SO4mist with sub-micron droplets.30,39In a wet scrubber, generally, the mass transfer between this mist and the scrubbing liquid occurs through inertial impaction, gravitational settling, Brownian diffusion, diffusiophoresis, and thermophoresis.
For sub-micron mist droplets with diameters less than
⬃0.05 m, Brownian diffusion is known to be the primary component of mass transfer.40In practical situations, de- pending on the relative velocity between the sub-micron mist droplets and the collecting liquid form (droplets, wetted walls, liquid sheets, and bubbles), the Brownian motion-induced mass transfer is often not adequate to result in high-efficiency capture of sub-micron droplets.
Therefore, generally, larger droplets in the mist can be removed in the scrubber, but a significant portion of the sub-micron droplets are not removed and are emitted from the stack. This explains why wet FGD systems are relatively inefficient in removing SO3/H2SO4. In general,
⬃50% of the H2SO4 entering the scrubber may be re- moved in the scrubber.
MEASUREMENT OF SO3
The controlled condensation system (CCS) has been used since the 1960s for the measurement of SO3and H2SO4in flue gas streams.41,42The general arrangements of the CCS and thimble holder are shown in Figures 7 and 8, respec- tively. The measurement system consists of a quartz-lined heated probe that draws gas through a quartz thimble for the removal of PM. The probe and thimble are heated to avoid condensation of SO3vapor in the gas sample. The gases enter a temperature-controlled condenser where the SO3 is condensed on the wall to be collected and mea- sured after the sample run via a deionized water rinse.
The measurement techniques for SO3and H2SO4re- cently have been under review for accuracy and improve- ment for TRI reporting. Recently, a review and verification project has been undertaken by EPRI to qualify the CCS method and compare field data with available
Figure 6. Fly ash SO3capture rate downstream of furnace.
prediction methods used to estimate SO3 emissions for TRI reporting.37 The EPRI project compared laboratory data with field test data for a variety of coals with a range of S and fly ash chemical compositions. The study further investigated the SO3removal rates and efficiencies of the plant equipment used for field verification.
The EPRI project has concluded that alkaline ash will produce a bias in the SO3 measurement. The bias is a result of SO3removal in the thimble holder by the alka- line ash collected on the quartz thimble. This bias is seen as significant for low-S, high-alkaline ash fuels (Powder River Basin coals). The CCS was found to have a bias of 20 –25% low readings for high S, acidic ash fuels and greater than 40% for Powder River Basin coals. A possible solution to the bias is utilized in Europe.43 The test method uses a system similar to the CCS; however, the quartz thimble is replaced with a small tubular ESP. The tubular ESP removes the ash to the sidewalls away from the gas stream; in contrast, the quartz thimble filters the flue gas through the collected ash. Furthermore, the Eu- ropean system can distinguish between gaseous and aero- sol SO3using deionized water procedures on various col- lection plates. The European SO3method system has been used extensively in the United States on a variety of coals with repeatable and reliable results.
MITIGATION OF SO3 EMISSIONS
The mitigation of SO3has been an active area of research for many years.13,23,44,45This research has resulted in the development, refinement, and implementation of differ- ent techniques and methods for the successful mitigation of SO3 in the flue gases of fossil-fuel–fired boilers. Re- cently, a guide that summarizes SO3mitigation options and their respective success in either full-scale or pilot testing has been written.10 The mitigation options are described below.
Alkali Addition into the Furnace
The injection of alkali in the flue gas stream has been a method of SO3mitigation for almost 30 years.44The lo- cation in the flue gas stream and the delivery method of the alkali have been studied and tested depending on operating and site conditions.
The addition of alkali into the furnace recently has been proven to be effective at the full scale.30,39However, the method of delivery and the effectiveness of the alkali used varies from site to site depending on specific condi- tions. The addition of limestone to coal before pulveriza- tion for the control of SCR catalyst arsenic poisoning has been shown to be an effective method of furnace SO3 control. Recent commissioning data46have shown at least Figure 7. CCS sampling train.
a 50% reduction in furnace SO3from limestone addition.
The injection of alkaline sorbents (Ca- and Mg-based slur- ries) has been shown to be effective at controlling SO3 emissions of the furnace.39 The slurry injection method has successfully obtained SO3furnace conversion reduc- tions of 40 – 80% but has been found to be sensitive to injection location and elevation. The effects of the MgO sorbent slurries on SCR catalyst activity are currently un- der investigation. To date, no studies have measured the potential benefit of SCR catalyst arsenic poisoning control by MgO sorbent injection. The addition or injection of alkali in the furnace does not influence the conversion rate of the SCR catalyst.
Alkali Injection after the Furnace
The injection of other alkaline materials after the furnace/
economizer exit has been used for the control of SO3for both APH corrosion and stack emissions.10The primary sorbents used are compounds such as hydrated lime, lime- stone, MgO, and sodium carbonate. The selection of a sorbent for a given site will depend on economic factors, such as availability and required SO3removal rates. Alkali injection has successfully removed between 40 and 90%
of the SO3in the flue gas at various plants depending on the injected material and injection rate.10 The sorbents are introduced into the flue gas as either a dry powder or mixed with water to form slurry before injection. The location of injection in the flue gas stream varies. Plants with APH cold-end corrosion problems may elect to inject the sorbent before the APH; however, adequate APH cleaning equipment is required with this configuration.
In general, the common injection location for SO3 con- trol is between the APH and the ESP. The injection of sorbent before the ESP must, however, consider the effect on particulate control. The ESP will have higher inlet mass loading, and the fly ash will have different resistivity characteristics. It has been reported that dry injection of hydrated lime has resulted in strong sparking and lower operating currents in the ESP during pilot-scale testing.10
Ammonia Injection before the ESP
The injection of NH3after the APH and before the ESP has been shown to be⬍90% effective in the removal of SO3in full-scale application.10 This method of mitigation re- sults in the formation of (NH4)2SO4and NH4HSO4salts in the ESP, depending on the NH3 and SO3 concentration Figure 8. CCS thimble holder.
ratios. The formation of NH4HSO4is expected when NH3 to SO3molar ratios are less than 1; this formation tends to decrease the ESP particle loading caused by fly ash agglomeration. NH3injection before the ESP is used for fly ash conditioning to increase ESP performance caused by the agglomeration effects. With NH3to SO3molar ratios between 1 and 2, increased formation of (NH4)2SO4 is expected, with an increase in the particle loading of the ESP. The injection of NH3results in the adsorption of NH3 by the fly ash. Because the fly ash will contain most of the injected NH3, the concentration may exceed acceptance limits for ash salability. Additional treatment of fly ash holding ponds and basins may be required if large amounts (⬃30 ppmv) of SO3are being removed. The use of NH3for SO3mitigation is practical on units equipped with SCRs, because NH3 is used as the reagent and is readily available on site.
Fuel Switching and Blending
The firing of sub-bituminous coals typically results in low SO3formation and emission rates. Consequently, switch- ing from firing bituminous to sub-bituminous coals can be a mitigation option. However, equipment and fuel cost factors often make such a change impossible. Many boiler systems do not have the capacity and the equipment to accommodate the firing of sub-bituminous coal without major modifications that make fuel switch economically unacceptable. Coal availability and costs also can con- strain fuel switching. One possible solution is the blend- ing of bituminous and sub-bituminous coals to create a blend that has fuel and ash characteristics favorable for SO3emissions. This strategy is currently used for control- ling SO2 emissions. When sub-bituminous coals are blended with bituminous coals, the overall S content of the fuel is reduced, resulting in a reduction of the SO2 concentration in the flue gas as well as the conversion of SO2to SO3in the furnace and SCR and, thereby, resulting in an overall reduction in SO3. Also, the sub-bituminous coal ash contains large percentages of alkaline materials that further assist in the capture of SO3in the APH and ESP.
Wet ESPs
Similar to dry ESPs, wet ESPs (WESPs) operate in a three- step process involving (1) charging of the entering parti- cles; (2) collection of the particles on the surface of an oppositely charged surface; and (3) cleaning the collec- tion surface. Both technologies employ separate charging and collection systems. However, the collecting surface in WESPs is cleaned with a liquid, in contrast to mechanical cleaning in dry ESPs. Consequently, the two technologies differ in the nature of particles that can be removed, the overall efficiency of removal, and the design and
maintenance parameters.47 While dry ESPs are typically limited to power levels of 100 –500 W per 1000 cfm of flue gas, WESPs can operate with power levels as high as 2000 W per 1000 cfm. Because of wet cleansing of the collec- tion system, PM does not accumulate in the collection electrodes; this mitigates particle re-entrainment. Based on these factors, WESPs can collect sub-micron particles and acid mist very efficiently. WESPs can be configured for vertical or horizontal gas flows in tubular or plate designs. Tubular designs offer smaller footprints and, in general, are more efficient than the plate type.
WESPs can be integrated easily with a wet scrubber.
In fact, integration of the WESP within the wet scrubber is a design option with many attractive features including a compact footprint; the ability to integrate the handling of the wash water and solids from the WESP with scrubber slurry, thereby avoiding the need for separate tank and blowdown system; and the ability to collect the fine H2SO4mist, which typically escapes the scrubber because of its very small droplet size.47
In 1986, the first commercial WESP application on a U.S. power plant took place when AES Deepwater, a 155-MW cogeneration plant firing petroleum coke as the primary fuel, was equipped with a WESP. With the WESP in operation, the plume opacity at the plant is generally 10% or less.48
Recently, an upflow tubular design WESP has been retrofit at Northern States Power Company’s Sherco Sta- tion in a wet scrubber/WESP configuration. Two more power plant applications are underway presently: (1) a 5000-cfm slipstream at Bruce Mansfield Station; and (2) a plate-type WESP for integration with Powerspan’s ECO technology to be demonstrated at First Energy’s R.E.
Burger plant.
Tests at the Sherco Station (WESP retrofit to the outlet section of the wet scrubber) allowed the scrubber to main- tain a 70% SO2reduction while keeping particulate emis- sions at 0.01 lb/106 Btu and opacity under 10%. Full conversion of all scrubber modules at the plant with WESPs is now underway.49 The WESP at the Mansfield Station is achieving greater than 95% removal of SO3and PM2.5and stack flow with near 0 opacity.50
Changing the Operation of APH
Increasing the heat transfer (or cooling of the flue gas) in the APH would appear to be a potentially viable strategy for removing some of the SO3/H2SO4in the APH. This, in turn, would lead to increased condensation of H2SO4in the APH and also to improved plant efficiency. However, the dew point at the APH outlet is ⬃230 °F (110 °C), thereby limiting the SO3 removal to ⬃90%.30 On the other hand, the potential for corrosion in the APH and downstream duct would increase with increased H2SO4
condensation. Consequently, more frequent soot blowing may be required to control this corrosion. These factors would need to be considered when deciding to change the APH operation to mitigate SO3emissions. Data reflect that
⬃25% increase in H2SO4condensation may be possible by lowering the flue gas temperature at the APH exit by⬃40
°F (22 °C).
SUMMARY
Formation and emissions of SO3/H2SO4 potentially can lead to plant operation-related problems and environ- mental concerns. The formation of SO3is complex, de- pending upon fuel, operating parameters, and plant con- figuration, and understanding the parameters leading to excessive generation of SO3and subsequent formation of H2SO4 can assist in selections of practical control ap- proaches. Elevated SO3concentrations can lead to corro- sion, the formation of sulfite scale, and fouling and plug- ging of low-temperature plant components, and can add to the particle loading to control equipment. In some cases, elevated SO3concentrations can lead to the forma- tion of visible plumes at the stack exit or shortly down- stream of the stack, resulting in noncompliance with local regulations.
SO3can be formed in the boiler during the combus- tion of S-bearing fuels or downstream, particularly in SCR reactors. SO3/H2SO4 can condense on low-temperature components and gets adsorbed by fly ash. Adsorption is much greater in the high alkaline ashes resulting from sub-bituminous coal firing. SO3/H2SO4traditionally has been measured using extractive controlled condensation methods.
If needed, SO3/H2SO4emission can be mitigated us- ing a variety of methods. Injection of alkali materials into the furnace, either with the fuel or in slurry form, has resulted in reductions of up to 80%. Post-furnace injec- tion of alkali materials can achieve up to 90% reductions but can increase particle loadings and ash resistivity char- acteristics. NH3 injection can also reduce SO3/H2SO4by
⬍90% and may result in increased particle loading to the downstream collection systems. In plants with adequate operational and equipment flexibility, fuel switching and blending can be used to reduce formation and emissions of SO3. WESPs are also an option for control of SO3/ H2SO4, and a variety of designs have been successfully demonstrated for collection of acid mists and opacity control.
REFERENCES
1. Combustion Engineering; Fryling, G.R., Ed.; Combustion Engineering Inc.: New York, 1966.
2. Hawthorne, M. Plant Spewing Acid Clouds; Columbus Dispatch, August 3, 2001, p1C.
3. Verhoff, F.H.; Banchero, J.T. Predicting Dew Points of Flue Gases;
Chem. Eng. Prog. 1974, 70 (8), 71-72.
4. Damle, A.S.; Ensor, D.S.; Sparks, L.E. Prediction of the Opacity of Detached Plumes Formed by Condensation of Vapors; Atmos. Environ.
1984, 18, 435-444.
5. Damle, A.S.; Ensor, D.S.; Sparks, L.E. Options for Controlling Conden- sation Aerosols to Meet Opacity Standards; J. Air Pollut. Control Assoc.
1987, 37, 925-933.
6. Markowski, G.R.; Ensor, D.S.; Hooper, R.G.; Carr, R.C. A Submicron Aerosol Mode in Flue Gas from a Pulverized Coal Utility Boiler; Envi- ron. Sci. Technol. 1980, 14 (11), 1400-1402.
7. Leikauf, G.D.; Spektor, D.M.; Albert, R.E.; Lippmann, M. Dose- Dependent Effects of Submicrometer Sulfuric Acid Aerosol on Particle Clearance from Ciliated Human Lung Airways; Am. Ind. Hyg. Assoc. J.
1984, 45, 285-292.
8. Utell, M.J.; Frampton, M.W.; Morrow, P.E. Air Pollution and Asthma:
Clinical Studies with Sulfuric Acid Aerosols; Allergy Proc. 1991, 12, 385-388.
9. Method 9 —Visual Determination of the Opacity of Emissions from Sta- tionary Sources; 40 CFR 60, Appendix A; Government Printing Office: Washington, DC, 1996. Available at: www.epa.gov/ttn/emc/
promgate/m-09.pdf (accessed April 2004).
10. Peterson, J.; Jones, A.F. SO3Mitigation Guide; EPRI Report Number TR-104424; EPRI: Palo Alto, CA, October 1994.
11. Singer, J.G. Combustion—Fossil Power Systems, 3rd ed.; Combustion Engineering Inc.: New York, 1981.
12. Burke, J.M.; Johnson, K.L. Ammonium Sulfate and Bisulfate Formation in Air Preheaters; EPA-600/7– 82-025a; Industrial Environmental Research Laboratory, U.S. Environmental Protection Agency: Research Triangle Park, NC, 1982.
13. Zetlmeisl, M.J.; McCarthy, K.J.; Laurence, D.F. Investigation of Some Factors Affecting Nonflame Generation and Suppression of SO3; In- dust. Eng. Chem. Prod. Res. Dev. 1983, 22, 710-716.
14. Miller, C.A., U.S. Environmental Protection Agency. Personal commu- nication with J. Bock of Siemens AG, Redwitz, Germany, June 18, 2002.
15. Clean Air Act Amendments of 1990; Public Law 101–549; Government Printing Office: Washington, DC, November 15, 1990. Available at:
www.epa.gov/epahome/laws.htm (accessed April 2004).
16. U.S. Environmental Protection Agency. Study of Hazardous Air Pollut- ant Emissions from Electric Utility Steam Generating Units—Final Report, Volume 1; EPA-453/R-98 – 049; Office of Air Quality Planning and Standards: Research Triangle Park, NC, 1998.
17. U.S. Environmental Protection Agency. Air Quality Criteria for Particu- late Matter; EPA-600/P-95/001 (NTIS PB96 –168224); National Center for Environmental Assessment: Research Triangle Park, NC, April 1996.
18. U.S. Environmental Protection Agency. Consolidated Emissions Re- porting; Fed. Reg. 2002, 67, 39602-39616.
19. Alzueta, M.U.; Bilbao, R.; Glarborg, P. Inhibition and Sensitization of Fuel Oxidation by SO2; Combust. Flame 2001, 127, 2234-2251.
20. Senior, C.L.; Sarofim, A.F.; Zeng, T.; Helble, J.J.; Mamani-Paco, R. Fuel Process. Technol. 2000, 63, 197-213.
21. Warnatz, J.; Maas, U.; Dibble, R.W. Combustion; Springer-Verlag: New York, 1996; pp 67-71.
22. Hardman, R.; Stacy, R.; Dismukes, E. Estimating Sulfuric Acid Aerosol Emissions from Coal-Fired Power Plants. Presented at the DOE-FETC Conference on Formation, Distribution, Impact, and Fate of Sulfur Trioxide in Utility Flue Gas Streams, Pittsburgh, PA, March 1998.
23. Marier, P.; Dibbs, H.P. The Catalytic Conversion of SO2to SO3by Fly Ash and the Capture of SO2and SO3by CaO and MgO; Thermochim.
Acta 1974, 8, 155-165.
24. Monroe, L. An Updated Method for Estimating Total Sulfuric Acid Emis- sions from Stationary Power Plants; Research and Environmental Affairs Department, Southern Company Generation and Electricity Market- ing: Birmingham, AL, 2001.
25. Nova, I.; dall’Acqua, L.; Lietti, L.; Giamello, E.; Forzatti, P. Study of Thermal Deactivation of a De-NOxCommercial Catalyst; Appl. Catal.
B: Environ. 2001, 35, 31-42.
26. Forzatti, P.; Nova, I.; Beretta, A. Catalytic Properties in DeNOxand SO2-SO3Reactions; Catal. Today 2000, 56, 431-441.
27. Dunn, J.; Koppula, P.R.; Stenger, H.G.; Wachs, I. Oxidation of Sulfur Dioxide to Sulfur Trioxide over Supported Vanadia Catalysts; Appl.
Catal. B: Environ. 1998, 19, 103-117.
28. Dunn, J.; Stenger, H.G.; Wachs, I. Oxidation of Sulfur Dioxide over Supported Vanadia Catalysts: Molecular Structure-Reactivity Relation- ships and Reaction Kinetics; Catal. Today 1999, 51, 301-318.
29. Dunn, J.; Stenger, H.G.; Wachs, I. Molecular Structure—Reactivity Relationships for the Oxidation of Sulfur Dioxide over Supported Metal Oxide Catalysts; Catal.Today 1999, 53, 543-556.
30. Gutberlet, H.; Hartenstein, H.; Licata, A. SO2Conversion Rate of DeNOx Catalysts—Effects on Downstream Plant Components and Remedial Measures. Presented at PowerGen 1999, New Orleans, LA, 1999.
31. Svachula, J.; Alemany, L.J.; Ferlazzo, N.; Forzatti, P.; Tronconi, E.;
Bregani, F. Oxidation of SO2 to SO3 over Honeycomb DeNOxing Catalysts; Ind. Eng. Chem. Res. 1993, 32, 826-834.
32. Levy, E.K. Effect of Boiler Operations on Sulfuric Acid Emissions.
Presented at the DOE-FETC Conference on Formation, Distribution, Impact, and Fate of Sulfur Trioxide in Utility Flue Gas Streams, Pitts- burgh, PA, March 1998.
33. Sarunac, N.; Levy, E. Factors Affecting Sulfuric Acid Emissions from Boilers. Presented at the EPRI/DOE/EPA Air Pollution Control Sympo- sium: The Mega Symposium, Atlanta, GA, August 16 –20, 1999.
34. Devito, M.S.; Oda, R.L. Flue Gas SO3Stratification at ESP Inlets. Pre- sented at the DOE-FETC Conference on Formation, Distribution, Im- pact, and Fate of Sulfur Trioxide in Utility Flue Gas Streams, Pitts- burgh, PA, March 1998.
35. Yilmaz, A.; Pechulis, M.; D’Agostini, M.D.; Levy, E.; Sarunac, N.;
Maust, P. Cold End Acid Condensation in Tubular Air Preheaters. In Proceedings of 1995 Joint Power Generation Conference, PWR-Vol. 28;
American Society of Mechanical Engineers: New York, NY, 1995; Vol.
3, pp 473-480.
36. D’Agostini, M.; Levy, E.; Wilson, R. A Method for Calculating Acid Deposition Rates in Regenerative Air Preheaters. In Heat Transfer Equip- ment Fundamentals, Design Application and Operating Problems; Sath, R., Ed.; ASME-HTD: New York, NY, 1990; Vol. 108, pp 78-87.
37. Blythe, G.; Galloway, C.; Rhudy, R. Flue Gas Sulfuric Acid Measure- ment Method Improvements. Presented at the U.S. EPA/DOE/EPRI Combined Power Plant Air Pollution Control Symposium: The Mega Symposium; Chicago, IL, August 20 –23, 2001.
38. Gutberlet, H. Pleasant Prairie Power Plant SO3Measurements at Different Test Locations; Riley Power Inc. Internal Report by E. On Engineering, Report Number 01– 01-100962; Riley Power: Worcester, MA, January 2002.
39. Blythe, G.; McMillan, R.; Davis, J.; Lamison, M.; Beeghly, J.; Benson, L.; Goetz, E.; Rhudy, R. Furnace Injection of Alkaline Sorbents for Sulfuric Acid Control. Presented at the U.S. EPA/DOE/EPRI Combined Power Plant Air Pollution Control Symposium: The Mega Symposium, Chicago, IL, August 20 –23, 2001.
40. Air & Waste Management Association. Air Pollution Engineering Man- ual, Ed. 2; Davis, W., Ed.; Wiley & Sons: New York, NY, 2000.
41. Lisle, E.S.; Sensenbaugh, J.D. The Determination of Sulfur Trioxide and Acid Dew Point in Flue Gases; Combustion 1965, 36, 12-16.
42. Dismukes, E.B. Conditioning of Fly Ash with Sulfur Trioxide and Ammo- nia; EPA-600/2–75-015 (NTIS PB-247231); U.S. Environmental Protec- tion Agency, Office of Energy, Minerals, and Industry: Research Tri- angle Park, NC, 1975.
43. Erickson, C., Babcock Borsig Power Inc. Personal communication with H. Gutberlet of E. On Engineering, Gelsenkrichen, Germany, July 7, 2002.
44. Reese, J.T.; Jonakin, J.; Caracristi, V.Z. Prevention of Residual Oil Combustion Problems by Use of Low Excess Air and Magnesium Additive; J. Eng. Power 1965, Transactions of ASME, Series A, 87, 229- 236.
45. Radway, J.E.; Exley, L.M. A Practical Review of the Cause and Control of Cold End Corrosion and Acidic Stack Emissions in Oil-Fired Boilers.
Presented at the Winter Meeting of the ASME, Atlanta, GA, December 6 –10, 1975.
46. Hutcheson, L.; Martin, M.; Ake, T.; Erickson, C.A.; Nystedt, P.; McFar- land, E.; Favor, C. Selective Catalytic Reduction System Performance at Duke Energy’s Cliffside Unit 5 Power Station. Presented at the EPRI 2002 Workshop on Selective Catalytic Reduction, Atlanta, GA, Octo- ber 22–23, 2002.
47. Altman, R.; Offen, G.; Buckley, W.; Ray, I. Wet Electrostatic Precipita- tion Demonstrating Promise for Fine Particulate Control-Part I; Power Eng. 2001 Jan, 37-39.
48. Kumar, K.S.; Mansour, A. Wet ESP for Controlling Sulfuric Acid Plume Following an SCR System. Presented at the ICAC Forum 2002, Hous- ton, TX, February 12–13, 2002.
49. Altman, R. Wet Electrostatic Precipitation Demonstrating Multiple Pollutant Control in Industrial Applications Holds Promise for Coal- Fired Utility Emission Reduction of Acid Mist, PM2.5, and Mercury.
Presented at the U.S. EPA/DOE/EPRI Combined Power Plant Air Pol- lution Control Symposium: The Mega Symposium; Chicago, IL, Au- gust 20 –23, 2001.
50. Srivastava, R., U.S. Environmental Protection Agency. Personal com- munication with J. Reynolds of Croll-Reynolds Co. Inc., Westfield, NJ, August 8, 2002.
About the Authors
R.K. Srivastava and C.A. Miller are at the Office of Research and Development, National Risk Management Research Laboratory, Research Triangle Park, NC. C. Erickson and R.
Jambhekar are at Babcock Borsig Power, Inc., Worcester, MA. Address correspondence to: R.K. Srivastava, Office of Research and Development, National Risk Management Research Laboratory, Air Pollution Prevention and Control Division, U.S. Environmental Protection Agency, Research Triangle Park, NC, 27711; fax: (919) 541-0554; e-mail:
