Mass Transfer within Electrostatic Precipitators: In ... - ACS Publications

Apr 26, 2006 - Mass Transfer within Electrostatic. Precipitators: In-Flight Adsorption of Mercury by Charged Suspended. Particulates. HEREK L. CLACK*...
0 downloads 0 Views 170KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 3617-3622

Mass Transfer within Electrostatic Precipitators: In-Flight Adsorption of Mercury by Charged Suspended Particulates HEREK L. CLACK* Department of Mechanical, Materials and Aerospace Engineering, Illinois Institute of Technology, 10 West 32nd Street, Chicago, Illinois 60616

Electrostatic precipitation is the dominant method of particulate control used for coal combustion, and varying degrees of mercury capture and transformation have been reported across ESPs. Nevertheless, the fate of gasphase mercury within an ESP remains poorly understood. The present analysis focuses on the gas-particle mass transfer that occurs within a charged aerosol in an ESP. As a necessary step in gas-phase mercury adsorption or transformation, gas-particle mass transfersparticularly in configurations other than fixed bedsshas received far less attention than studies of adsorption kinetics. Our previous analysis showed that only a small fraction of gas-phase mercury entering an ESP is likely to be adsorbed by collected particulate matter on the plate electrodes. The present simplified analysis provides insight into gas-particle mass transfer within an ESP under two limiting conditions: laminar and turbulent fluid flows. The analysis reveals that during the process of particulate collection, gas-particle mass transfer can be quite high, easily exceeding the mass transfer to ESP plate electrodes in most cases. Decreasing particle size, increasing particle mass loading, and increasing temperature all result in increased gas-particle mass transfer. The analysis predicts significantly greater gasparticle mass transfer in the laminar limit than in the turbulent limit; however, the differences become negligible under conditions where other factors, such as total mass of suspended particulates, are the controlling mass transfer parameters. Results are compared to selected pilotand full-scale sorbent injection data.

Introduction Coal combustion for electric power generation was responsible for the release of 45.6 tons of mercury into the environment in 2001 (1). Compared to incinerators, mercury emissions from coal-fired power plants (CFPPs) are more difficult to control because mercury concentrations are much lower, mercury speciation is much more uncertain, and electric power generation by its nature strongly discourages activities that diminish or disrupt a facility’s generating capacity. In 2002, coal-fired power plants (CFPPs) represented roughly one-half of the total generating capacity in the United States. Of this capacity, one-third predates the original Clean Air Act (1970) and another third is at least 25 years old (2). The wide age distribution of CFPPs suggests that (1) original * Corresponding author e-mail: [email protected]. 10.1021/es050246+ CCC: $33.50 Published on Web 04/26/2006

 2006 American Chemical Society

CFPP designs reflected the technology available when they were built, and (2) current CFPPs may differ substantially from their original design as a result of modifications and repairs. For these reasons, an “elegant solution” to mercury emissions control is highly desirable. Of the two primary forms of mercury produced by CFPPs, the oxidized form (Hg2+, consisting of HgCl2 and to a lesser extent HgO) is water-soluble and highly condensable (3, 4). As a result, Hg2+ species are readily removed by condensation onto fly ash, bottom ash, and ductwork surfaces and by absorption within wet flue gas desulfurization (WFGD) processes. By comparison, the elemental form of mercury (Hg0) exhibits low condensability and negligible water solubility and therefore passes largely unaffected through the exhaust train. The relative proportions of Hg0 and Hg2+ in the exhaust train vary greatly, ranging from 90/10 to 20/80 (5-7), depending on the rank and origin of coal, boiler design, and sampling location in the exhaust train. Thus, proposed near-term solutions for capturing mercury collaterally with SO2 in a WFGD are somewhat limited by the uncertainty in the Hg2+ fraction in the flue gas at any given time, unless a supplemental mercury oxidation step is added. In addition, results by Ghorishi and Chang (8) demonstrated that Hg2+ can be reduced to Hg0 depending on the chemistry of the scrubber fluid. Much of the mercury emissions control research has focused on characterizing mercury adsorption isotherms and developing the homogeneous and heterogeneous kinetic mechanisms of mercury oxidation (4, 9-17). Lab-scale mercury adsorption and heterogeneous oxidation studies very frequently use fixed bed reactors, where the mercuryladen gas passes through a stationary bed of solid material and undergoes adsorption or oxidation therein. Far fewer studies have examined entrained flow or in-flight adsorption or oxidation of mercury, where the powdered material is suspended in the flowing gaseous reactants. The distinction between in-flight and fixed bed configurations is important because only 10% of CFPPs in the United States currently have fabric filters (25) and can directly extrapolate fixed bed laboratory results to their full-scale operation. Retrofitting the remaining CFPP population with filter retrofits can present significant challenges. For the two-thirds of CFPPs with ESPs installed, a retrofitted fabric filter constitutes a redundant particulate control process. Such retrofits may raise secondary issues such as the need to install induced draft (ID) fans to overcome the increased pressure drop, which in turn may draw a sufficiently deep vacuum that upstream ductwork may need to be reinforced. Older facilities are often space-constrained such that the difficulty of retrofitting projects grows rapidly with project scale. Of the relatively few lab-scale studies of in-flight mercury oxidation or adsorption, Scala (18, 19) modeled Hg2+ adsorption by entrained powdered activated carbon (PAC), finding that high PAC loading (10 g/m3) and long residence times (∼10 s) were needed to achieve 50% removal. Biswas and co-workers (20, 21) studied UV photocatalytic Hg0 oxidation by entrained TiO2 particles, an alternative to direct UV photooxidation of Hg0 such as that studied by Granite and Pennline (22). Previous results from our research group (23, 24), along with those from Gullett and co-workers (10, 17), have reported the effects of thermofluid properties, including turbulence intensity, on in-flight mercury adsorption, and transformation. The prevalence of ESPs at CFPPs constitutes a widely available platform for in-flight mercury capture or oxidation. Anecdotal evidence does not provide a clear inference. In a VOL. 40, NO. 11, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3617

FIGURE 1. Schematics of one-half of a particle-laden channel flow between two ESP plate electrodes, laminar (left) and turbulent (right) limits. (Drawings are not to scale. Some nomenclature common to both configurations are presented only in the figure at the left.) pilot-scale study, Rostam-Abadi et al. (26) injected PAC upstream of an ESP at a rate ranging from 6 to 24 lbm/ MMacf (pounds per million actual cubic feet, equivalent to 0.1-0.4 g/m3). They observed no mercury capture upstream of the ESP but as much as 73% was captured within the ESP. By contrast, Public Services Co./ADA Technologies (27) not only observed significant mercury adsorption (48%) after PAC injection and upstream of the ESP but also most of this figure (62%) was on the fly ash in the flue gas. A common assumption is that Hg adsorption in an ESP takes place on the collected particulate matter on the plate electrodes. A recent analysis (28) using the heat/mass transfer analogy showed that typical ESP conditions could not achieve the mass transfer rates needed to achieve the 70% mercury capture within ESPs observed in field data. The objective of the present analysis is to better understand gas-particle mass transfer within ESPs associated with CFPPs, as sufficiently detailed experimental data are not available on which a fundamental understanding can be based. Complicating this effort is the wide variability in size of ESPs installed at CFPPs. Specific collection area (SCA, surface area per unit volumetric flow rate) of ESPs associated with CFPPs in the United States varies from roughly 100 ft2/ kacfm to more than 700 ft2/kacfm, with no one size class representing more than 20% of the total fleet (38). Because of this large variability, the present study considers a simple model at two bounding conditions: a laminar, minimally dispersive flow and a turbulent, highly dispersive flow. Considering the differences in gas-particle mass transfer at these two extremes will shed light on the potential for mercury capture that exists within ESPs of all sizes.

Theory and Numerical Method The numerical model considers a particle-laden channel flow between two plate electrodes of an ESP. Figure 1 presents schematics of both limiting conditions: laminar, minimally dispersive flow and turbulent, highly dispersive flow. In practice, flows within ESPs are turbulent (33, 36), the result of fluid dynamic instabilities (i.e., high Reynolds numbers) and electrohydrodynamically induced secondary flows (i.e., “corona wind” produced by high electric fields). Because turbulence in an ESP can originate from either phenomenon, both phenomena are excluded from the laminar limiting case by assuming reduced values of both Reynolds number and electric field strength. The electrodes are of length L and have a spacing of H. Symmetry allows consideration of only one-half of the channel (0 < y < H/2 in Figure 1). Under both laminar and turbulent conditions, plug flow is assumed (U ) Uinlet * f(y)). At the channel entrance, the gas phase consists of 4 ppbv of elemental mercury (Hg0) uniformly distributed in air; thermodynamic and fluid dynamic properties are taken as those of air, an ideal gas of specified temperature and 3618

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 11, 2006

pressure, for such a dilute mixture. The simplified Hg0-air mixture decouples mass transfer effects from mercury chemical kinetics in the analysis of mercury removal within ESPs. This approach enables the study and optimization of the fundamental mass transfer mechanisms, unimpeded by the ongoing, but as yet incomplete, development of mercury kinetic mechanisms. The particulate phase is spherical in shape, monodisperse, and acts as a perfect Hg0 sink on whose surface Hg0 concentration is identically zero. The model considers neither electrical nor physical particle interactions, nor does it consider practical phenomena such as sneakage (particle-laden flow escaping through leaks) or rapping reentrainment (resuspension of collected particles during periodic cleaning of collection electrodes) that degrade ESP performance in practice (36). Laminar Limit - No Particle Dispersion. In the laminar limit, the absence of particle dispersion yields particle motion that is governed by Coulombic and fluid drag forces, both of which are invariant in time but are functions of particle diameter dp. Thus, in the laminar limit, particles in a monodisperse particle-laden flow all exhibit a constant, uniform velocity U(u,v) ) (Uinlet, - Ues), where constant fluid drag determines Uinlet and the balance between Coulombic and fluid drag forces (both constant) determines Ues. This uniform motion of the particles relative to the gas and toward the plate electrode creates particle-laden and particle-free regions of the channel, separated by a boundary oriented at an angle R relative to the centerline of the channel, as depicted in Figure 1, left. Because gas-particle mass transfer only occurs in the particle-laden region, a differential fluid volume ∆V that enters the channel at axial location x ) 0 and at time t ) 0, flowing in a stream tube at elevation yst above the plate electrode, undergoes gas-particle mass transfer only until t ) (H/2 - yst)/Ues or, alternatively, until x ) (H/2 - yst)Uinlet/ Ues. During this time, the particle mass loading (particle mass per unit volume) within the particle-laden region remains spatially uniform because of the uniform motion of the particles relative to the gas phase. Thus, a determination of overall gas-particle mass transfer within the channel involves (1) calculating the convective mass transfer rate for a single particle; (2) extending (1) to an ensemble of particles in a particulate-laden unit volume ∆V using the particle mass loading; and (3) evaluating the total gas-particle mass transfer occurring within ∆V while in the particle-laden region. Because (3) is a function of the stream tube elevation yst of ∆V, the total gas-particle mass-transfer at the exit of the channel is also a function of yst. It should be noted that while the present analysis assumes laminar flow as a limiting case, this is an idealization; it is widely accepted that flows within ESPs are turbulent (33, 36). Turbulent Limit - Infinite Turbulent Diffusivity and Particle Dispersion. The Deutsch-Anderson equation (36)

is widely used to predict ESP performance. Inherent in the Deutsch-Anderson equation is an assumption of sufficiently vigorous turbulence (hydrodynamic, electrohydrodynamic (i.e., “corona wind”), or both) that its dispersive effects maintain uniform property gradients in the direction normal to the bulk flow (y-direction in Figure 1). This condition effectively implies infinite turbulent diffusivity. As a result, concentrations of both particles and gas-phase mercury vary only in the x-direction, allowing an analysis that is spatially 1-D based on a control volume ∆V that extends across the half-height of the channel. Analysis of gas-particle mass transfer within ∆V, however, remains virtually unchanged from the approach used in the laminar limit because of the short transient response time of the particles. A separate analysis shows that, of the particle sizes considered here, even the largest particle sizes having the longest transient response times regain kinematic steady state within fractions of a millisecond. Consequently, the viscous drag and Coulombic forces are nearly always in equilibrium, despite the turbulent fluctuations in fluid velocity, and the particles maintain a virtually constant slip velocity as a result. Equation Development. The Fro¨ssling equation (29) (eq 1) correlates mass transfer to spheres at the low Reynolds numbers typically associated with small particles

Shd ) 2 + 0.552Red1/2Sc1/3

(1)

where Shd is the mean Sherwood number (analogous to the Nusselt number in heat transfer), Red is the Reynolds number based on particle diameter, and Sc is the Schmidt number (ratio of kinematic viscosity ν to molecular diffusivity Dab). The particle Reynolds number Red is based on the particle diameter dp and the magnitude of the particle slip velocity relative to the gas, which in this case equals Ues. The Reynolds analogy relates the mean convective mass transfer coefficient hm to Shd by eq 2

hm )

ShdDab dp

(2)

where Dab is the binary diffusion coefficient, and dp is the particle diameter. Dab is calculated using an expression (eq 3) developed by Fuller et al. (30)

Dab )

(

)

1.858e-27T3/2 1 1 + 2 M M Pσab ΩD a b

1/2

(3)

in which P is pressure [atmospheres], T is temperature [K], Mx is molecular weight of species x [g/gmol], σab is the average collision diameter for species a and b [m], and ΩD is the collision integral [-]. Values for σ and ΩD originate from the Lennard-Jones 6-12 potential, obtained by Svehla (31) and presented by Hines and Maddox (32). Calculated values of Dab are for the Hg0-air system. With the ability to calculate hm, the convective mass transfer rate to each spherical particle is given by the Reynolds analogy to Newton’s law of cooling (eq 4)

m ˘ Hg(t) ) hmApF(CV(t) - Cs)

(4)

where m ˘ Hg is the convective mass transfer rate of Hg0 from the surrounding gas to the surface of the particle, Ap is the surface area of the particle (Ap ) 4π(dp/2)2), F is the bulk gas density, and CV(t) and Cs are the concentrations of Hg0 in the differential volume ∆V and at the particle surface, respectively (Cs ≡ 0, based on assumptions of infinite sorbent reactivity and capacity). For an ensemble of N monodisperse particles in a differential volume ∆V, the corresponding gas-particle

mass transfer rate M ˙ Hg is obtained by replacing the single particle surface area Ap with the collective area of the ensemble N‚Ap where N equals the product of particle number density NDp [#/m3] and ∆V (eq 5).

M ˙ Hg(t) ) hmNDp∆V‚4π(dp/2)2F(CV(t) - 0)

(5)

Whereas NDp is a dichotomous function of time in the laminar limit (eq 6), in the turbulent limit it is a continuous function of time (eq 7) following a modified form of the DeutschAnderson equation based on the geometry in Figure 1, right

NDp ≡ NDp,inlet for t
1 µm) particles, Crawley (33) gives the maximum number of elementary charges retained nmax (eq 13)

dp2EL nmax ) 4KEe

(13)

where EL is the field strength at the surface of a particle necessary to induce the spontaneous emission of an ion. The value of EL is different for negatively (3 × 104 stV/cm) versus positively (7 × 105 stV/cm) charged particles; the present analysis assumes negatively charged particles and a relative permittivity of unity. The value of KE, a unit conversion factor, is unity if the remaining parameters in eq 13 are in cgs units, as is the case presently. The variable e is the electron charge as defined previously. The present model calculates nmax based on particle diameter but sets the particle charge n (eq 10) at a fraction of this value based on a user-specified particle charging efficiency ηcharge (n ) nmax‚ηcharge, 0.0 < ηcharge < 1.0). Other analyses of gas-particle mass transfer within ESPs (41) model particle charging as a function of the electric field. More advanced analyses of particle collection within ESPs have shown that the Deutsch-Anderson equation likely oversimplifies particle motion under turbulent conditions. Leonard, Mitchner, and Self (37) modeled ESP performance using finite values of turbulent diffusivity and found that particle number density was nonuniform normal to the plate electrodes. Using a Lagrangian simulation of ESP operation, Soldati et al. (39) found that their results differed from Deutsch-Anderson-derived results only when particulate mass loadings were very dilute. Although these examples demonstrate improved modeling of actual ESP operation, the Deutsch-Anderson equation remains a valid limiting condition for the present analysis.

Results and Discussion Table 1 presents the ESP design and operating parameters associated with the results in Figures 2 and 3 for the laminar limit and in Figures 4 and 5 for the turbulent limit. Under both laminar and turbulent limiting conditions, residual mercury fractions exhibit an initial decay, after which they maintain constant values referred to here as the terminal residual mercury fractions. Laminar Limit Results. Figure 2 reveals increasing gasparticle mass transfer with increasing particle mass loading, in agreement with pilot- and full-scale studies (e.g. refs 26 and 34). Unlike such studies, however, the present results 3620

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 11, 2006

FIGURE 3. Stream-wise evolution of mean residual Hg0 fraction at the laminar limit, as a function of particle diameter dp (MLp ) 0.1 g/m3).

FIGURE 4. Stream-wise evolution of residual Hg0 fraction and residual particulate matter fraction (circles) at the turbulent limit, as a function of MLp (dp ) 10 µm).

FIGURE 5. Stream-wise evolution of residual Hg0 fraction and residual particulate matter fraction (circles) at the turbulent limit, as a function of dp (MLp ) 0.1 g/m3). are parametrized by particle mass loading rather than carbonto-mercury (C:Hg) ratio to focus on the details of gas-particle mass transfer. Thus, the present results are relevant to both PAC and fly ash. Mean residual mercury fractions of 0.576 (about 42% capture) are possible for a particle mass loading MLp of only 0.1 g/m3 (Figure 2). By comparison, MLp is typically less than 0.1 g/m3 for injected sorbents in pilot- and full-scale tests, and MLp for native fly ash varies between 1 and 10 g/m3. Increasing MLp from 0.1 to 1 g/m3 decreases

the mean residual mercury fraction from 0.576 to 0.077, corresponding to mercury capture of 92%. At 1 g/m3, gasparticle mass transfer is more likely representative of Hg0 adsorption or heterogeneous oxidation on native fly ash. Note the expanded x-axis; the low Uinlet required for laminar flow enables complete particle removal in a small fraction of the channel length L. Figure 3 shows the impact of particle diameter on gasparticle mass transfer. Suspensions of larger particles exhibit higher electrostatic drift velocities and shorter residence times but higher Reynolds numbers Red and convective mass transfer coefficients hm. Suspensions of larger particles also have lower surface-to-volume ratios and thus present less surface total area for a given mass loading. Figure 3 shows the combined effects of these phenomena. For MLp ) 0.1 g/m3, suspensions of larger particles exhibit lower gas-particle mass transfer, resulting in higher mean residual mercury fractions. The higher electrostatic drift velocities attained by the larger particles cause the onset of the terminal residual mercury fraction to occur at lower values of x/L. Additional results under laminar limiting conditions are presented in the Supporting Information for particulate matter (PM) collection efficiency, the distribution of residual mercury fraction across the half-channel, and the effects of gas temperature. Turbulent Limit Results. At the turbulent limit, turbulent dispersion produced by hydrodynamic instabilities, corona wind, or both ensure well-mixed conditions and uniform concentration profiles for particulate matter (CPM ) CPM(x)) and Hg0 (CHg ) CHg(x)) in the transverse (y) direction. Gasparticle mass transfer results presented in Figure 4 at the turbulent limit appear qualitatively similar to those presented in Figure 3 under laminar limiting conditions. The initial decay rate of the residual mercury fraction is approximately the same for the laminar and turbulent limits. However, because the assumed gas velocity at the turbulent limit is 10 times that of the laminar limit, the onset of the terminal residual mercury fraction occurs farther downstream (higher values of x/L in Figure 4) under turbulent limiting conditions. Along with the residual mercury fractions, Figure 4 also presents simultaneous residual particulate matter fractions (CPM/CPM,0). Note that for a specified particle diameter, eq 15 predicts a single CPM/CPM,0 curve regardless of the initial particle mass loading. For the 10-µm particles assumed in Figure 4, collection is completed relatively rapidly, within the first 40% of the 10 m-long channel. Although the turbulence responsible for the well-mixed conditions leads to highly convoluted particle paths during particulate collection, the higher electric field (200 kV/m compared to 20 kV/m in the laminar limit) somewhat compensates by inducing ten times higher drift velocities. Figure 5 presents gas-particle mass transfer results at the turbulent limit as a function of particle diameter, analogous to Figure 3 at the laminar limit. As expected, for a specified particle mass loading (0.1 g/m3), smaller particles not only undergo greater gas-particle mass transfer and achieve lower residual mercury fractions (CHg/CHg,0) but also exhibit the slowest decay in residual particulate mass fraction (CPM/ CPM,0). By comparison, the 40-µm particles are collected very rapidly and have little time available for gas-particle mass transfer, resulting in CHg/CHg,0 ≈ 1, a result that is nearly indistinguishable from the upper boundary of the graph (Figure 5). Because they are collected so rapidly, the 40-µm CPM/CPM,0 results in Figure 5 are omitted to avoid confusion with the 1-µm CHg/CHg,0 results, which themselves are nearly indistinguishable from the y-axis as presented. For all but the 1-µm particles in Figure 5, the onset of the terminal residual mercury fraction is not an abrupt transition as occurred under laminar limit conditions but is instead continuous, reflecting the continuous, exponential decrease

in particle mass loading dictated by eq 15 for well-mixed conditions. For the 1-µm particles, gas-particle mass transfer is high enough that Hg0 removal is complete during particle collection. A three-way comparison of results for varying dp, MLp, and laminar vs turbulent conditions is presented in the Supporting Information. Comparison to Adsorption on Channel Walls. A previous analysis (28) showed that mercury adsorption on particulatecovered plate electrodes within an ESP is severely mass transfer limited. For representative conditions (U0 ) 1 m/s, H ) 0.5 m, L ) 10 m, T ) 500 K, P ) 1 atm) in a channel flow between two plate electrodes treated as perfect mercury sinks, only about 12% of the mercury was lost to the electrodes based on mass transfer considerations. In practice, the finite reactivity and adsorption capacity of the dust cake collected on the plate electrodes would yield poorer results. By comparison, gas-particle mass transfer (1 g/m3, 10-µm particles) achieved roughly 75% removal at the turbulent limit and 90% removal at the laminar limit. Thus, mass transfer considerations suggest that most of the mercury capture observed across ESPs takes place on the suspended particulate matter rather than the dust cake collected on the plate electrodes. Comparison to Full-Scale Sorbent Injection Results. Rostam-Abadi et al. (26) injected PAC upstream of an ESP on a small (12 MW) coal-fired boiler burning bituminous Illinois coal. Injecting various PACs (dp ) 6-18 µm) at a rate of 6 lbm/MMacf (0.1 g/m3), they observed mercury capture ranging from 20 to 73% at a temperature of 472 K. Such scatter is typical of mercury measurements in CFPPs. By comparison, the present analysis at the same particle mass loading (0.1 g/m3), 500 K, and for 10-µm particles predicts 13% (turbulent limit) and 42% (laminar limit) mercury capture based solely on gas-particle mass transfer considerations. Better agreement would require knowing the geometry (e.g., H, L), operating conditions (e.g., E, U0), and particle charging efficiency ηcharge of the ESP used in (26). This also applies to estimating the mercury uptake by the dust cake on the ESP plate electrodes. It is important to note that the present analysis, the sorbent injection results of Rostam-Abadi et al. (26), and other fullscale tests of mercury capture across ESPs without sorbent injection (25) contradict past assertions (12) by other researchers that no adsorption of volatile trace metals occurs within an ESP. Those assertions were largely based on nowdated experimental results that assumed all mercury in flue gas was Hg0 and that did not have the benefit of more recent findings showing the dependency of mercury species on coal rank and origin.

Supporting Information Available Additional results presented in three additional figures. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) U.S. EPA. Toxic Release Inventory; 2001. (2) National Research Council. Interim Report of the Committee on Changes in New Source Review Programs for Stationary Sources of Air Pollutants; National Academies Press: Washington, DC, 2005. (3) Krishnan, S. V., Gullett, B. K. Jozewicz, W. Sorption of Elemental Mercury by Activated Carbons. Environ. Sci. Technol. 1994, 28, 1506-1512. (4) Ghorishi, B., Gullett, B. K. Sorption of mercury species by activated carbons and calcium-based sorbents: effect of temperature, mercury concentration and acid gases. Waste Manage. Res. 1998, 16, 582-593. (5) Meij, R. Trace element behavior in coal-fired power plants. Fuel Process. Technol. 1994, 39, 199-217. (6) Prestbo, R. M.; Bloom, N. S. Mercury speciation adsorption (MESA) method for combustion flue gas; methodology, artifacts, VOL. 40, NO. 11, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3621

(7)

(8) (9)

(10) (11) (12) (13)

(14) (15)

(16)

(17)

(18) (19) (20)

(21)

(22) (23)

(24)

intercomparison, and atmospheric implications. Water, Air, Soil Pollut. 1995, 80, 145-158. Broderick, T.; Haythornthwaite, S.; Bell, W.; Selegue, T.; Perry, M. Determination of Dry Carbon-Based Sorbent Injection for Mercury Control in Utility ESP and Baghouses. Presented at the 91st Air and Waste Management Association Conference and Exhibition, San Diego, CA, June, 1998. Chang, J. C. S.; Ghorishi, S. B. Simulation and Evaluation of Elemental Mercury Concentration Increase in Flue Gas Across a Wet Scrubber. Environ. Sci. Technol. 2003, 37, 5763-5766. Zhuang, Y.; Thompson, J. S.; Zygarlicke, C. J.; Pavlish, J. H. Development of a Mercury Transformation Model in Coal Combustion Flue Gas. Environ. Sci. Technol, 2004, 38, 58035808. Serre, S. D.; Gullett, B. K.; Ghorishi, S. B. Entrained-Flow Adsorption of Mercury Using Activated Carbon. J. Air Waste Manage. Assoc. 2001, 51, 733-741. Krishnan, S. V.; Gullett, B. K.; Jozewicz, W. Mercury Control in Municipal Waste Combustors and Coal-fired Utilities. Environ. Prog. 1997, 16, 47-53. Korpiel, J. A.; Vidic, R. D. Effect of Sulfur Impregnation Method on Activated Carbon Uptake of Gas-phase Mercury. Environ. Sci. Technol. 1997, 31, 2319-2325. Wu, B.; Peterson, T. W.; Shadman, F.; Senior, C. L.; Morency, J. R.; Huggins, F. E.; Huffman, G. P. Interactions between vaporphase mercury compounds and coal char in synthetic flue gas. Fuel Process. Technol. 2000, 63, 93-107. Liu, W.; Vidic, R. D.; Brown, T. D. Impact of Flue Gas Conditions on Mercury Uptake by Sulfur-impregnated Activated Carbon. Environ. Sci. Technol. 2000, 34, 154-159. Niksa, S.; Fujiwara, N.; Fujita, Y.; Tomura, K.; Moritomi, H.; Tuji, T.; Takasu, S. A Mechanism for Mercury Oxidation in CoalDerived Exhausts. J. Air Waste Manage. Assoc. 2002, 52, 894901. Lee, T. G.; Hedrick, E.; Biswas, P. Hg Reactions in the Presence of Chlorine Species: Homogeneous Gas Phase and Heterogeneous Gas-Solid Phase. J. Air Waste Manage. Assoc. 2002, 52, 1316-1323. Ghorishi, S. B.; Keeney, R. M.; Serre, S. D.; Gullett, B. K.; Jozewicz, W. S. Development of a Cl-Impregnated Activated Carbon for Entrained- Flow Capture of Elemental Mercury. Environ. Sci. Technol. 2002, 36, 4454-4459. Scala, F. Simulation of Mercury Capture by Activated Carbon Injection in Incinerator Flue Gas. 1. In-Duct Removal. Environ. Sci. Technol. 2001, 35, 4367-4372. Scala, F. Modeling mercury capture in coal-fired power plant flue gas. Ind. Eng. Chem. Res. 2004, 43, 2575-2589. Wu, C.-Y.; Lee, T.-G.; Tyree, G.; Arar, E.; Biswas, P. Capture of Mercury in Combustion Systems by In Situ-Generated Titania Particles with UV Irradiation. Environ. Eng. Sci. 1998, 15, 137148. Rodriquez, S.; Almquist, C.; Lee, T.-G.; Furuuchi, M.; Hedrick, E.; Biswas, P. A Mechanistic Model for Mercury Capture with In Situ-Generated Titania Particles: Role of Water Vapor. J. Air Waste Manage. Assoc. 2004, 54, 149-156. Granite, E. J.; Pennline, H. W. Photochemical Removal of Mercury from Flue Gas. Ind. Eng. Chem. Res. 2002, 41, 5470-5476. Clack, H. L.; Narasimhan, U.; Serre, S. D. Direct Comparison of Fixed Bed and In-Flight Trace Gas Adsorption for Mercury Control. In Proceedings of the 98th Air and Waste Management Association Conference and Exhibition; San Diego, CA, June, 22-26, 2003. Clack, H. L.; Serre, S. D.; Narasimhan, U. The Relative Influence of Turbulence and Turbulent Mixing on the Adsorption of

3622

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 11, 2006

(25)

(26)

(27)

(28)

(29) (30)

(31) (32) (33) (34)

(35)

(36)

(37)

(38)

(39)

(40)

(41)

Mercury within a Gas-Sorbent Suspension. In Proceedings of the 2003 DOE-EPRI-U.S. EPA-A&WMA Power Plant Air Pollutant Control Symposium; Washington, DC, May 19-22, 2003. Brown, T. D.; Smith, D. N.; Hargis, R. A., Jr.; O’Dowd, W. J. Mercury Measurement and Its Control: What We Know, Have Learned, and Need to Further Investigate. J. Air Waste Manage. Assoc. 1999, 49, 1-97. Rostam-Abadi, M.; Chang, R.; Chen, S.; Lizzio, T.; Richardson, C.; Sjostrom, S. Demonstration of Sorbent Injection Process for Illinois Coal Mercury; Final Technical Report; Illinois Clean Coal Institute Project Number 00-1/2.2D-1, 2001. Public Service Company of Colorado/ADA Technologies, Inc. Investigation and Demonstration of Dry Carbon-Based Injection for Mercury Control; Final Report Under Phase I DOE/FETC Mega PRDA Program (period of performance September 1995 to July 1997), 1997. Clack, H. L. Mass Transfer within Electrostatic Precipitators: Trace Gas Adsorption by Sorbent-covered Plate Electrodes. J. Air Waste Manage. Assoc., 2005, in press. Frossling, N. Uber die Verdunstung fallender Tropfen. Beitr. Geophys. 1938, 52, p 170. Fuller, E. N.; Schettler, P. D.; Giddings, J. C. A New Method for Prediction of Binary Gas-Phase Diffusion Coefficients. Ind. Eng. Chem. 1966, 58, 19. Svehla, R. A. NASA Technical Report R-132; Lewis Research Center: Cleveland, OH, 1962. Hines, A. L.; Maddox, R. N. Mass Transfer: Fundamentals and Applications; Prentice-Hall: New York, 1985. Crowley, J. M. Fundamentals of Applied Electrostatics; Wiley: New York, 1986. Sjostrum, S.; Haythornthwaite, S.; Ruhl, J.; Smith, J. Investigation and Demonstration of Dry Carbon-Based Sorbent Injection for Mercury Control; Quarterly Report DE-AC22-95PC95256-07; U.S. Department of Energy: Morgantown, WV, 1997. Pavlish, J. H.; Sondreal, E. A.; Mann, M. D.; Olson, E. S.; Galbreath, K. C.; Laudal, D. L.; Benson, S. A. Status review of mercury control options for coal-fired power plants. Fuel Process. Technol. 2003, 82, 89-165. Air Pollution Engineering Manual/Air & Waste Management Association; Davis, W. T., Ed.; John Wiley & Sons: New York, 2000; pp 87-99. Leonard, G.; Mitchner, M.; Self, S. A. Particle Transport in electrostatic precipitator performance. Atmos. Environ. 1980, 14, 1289-1299. Bustard, C. J.; Durham, M. D. Results from ADA-ES/DOE Mercury Control Program at Wisconsin Electric Pleasant Prairie Power Plant. Presented to the Wisconsin Department of Natural Resources, February 25, 2002. Soldati, A.; Casal, M.; Andreussi, P.; Banerjee, S. Lagrangian Simulation of Turbulent Particle Dispersion in Electric Precipitators. AIChE J. 1997, 43, 1403-1413. Ohadi, M. M.; Nelson, D. A.; Zia, S. Heat transfer enhancement of laminar and turbulent pipe flow via corona discharge. Int. J. Heat Mass Transfer 1991, 34, 1175-1187. Clack, H. L. Particle Size Distribution Effects on Gas-Particle Mass Transfer within Electrostatic Precipitators. Environ. Sci. Technol. 2006, in press.

Received for review February 6, 2005. Revised manuscript received February 8, 2006. Accepted March 8, 2006. ES050246+