Estimates of Increased Black Carbon Emissions from Electrostatic

Jun 4, 2012 - carbon on submicrometer fly ash, each corresponding to its respective collection ... and values for fly ash (FA) from coal combustion ty...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/est

Estimates of Increased Black Carbon Emissions from Electrostatic Precipitators during Powdered Activated Carbon Injection for Mercury Emissions Control Herek L. Clack* Department of Mechanical, Materials and Aerospace Engineering, Illinois Institute of Technology, Chicago, Illinois, United States ABSTRACT: The behavior of mercury sorbents within electrostatic precipitators (ESPs) is not well-understood, despite a decade or more of full-scale testing. Recent laboratory results suggest that powdered activated carbon exhibits somewhat different collection behavior than fly ash in an ESP and particulate filters located at the outlet of ESPs have shown evidence of powdered activated carbon penetration during full-scale tests of sorbent injection for mercury emissions control. The present analysis considers a range of assumed differential ESP collection efficiencies for powdered activated carbon as compared to fly ash. Estimated emission rates of submicrometer powdered activated carbon are compared to estimated emission rates of particulate carbon on submicrometer fly ash, each corresponding to its respective collection efficiency. To the extent that any emitted powdered activated carbon exhibits size and optical characteristics similar to black carbon, such emissions could effectively constitute an increase in black carbon emissions from coal-based stationary power generation. The results reveal that even for the low injection rates associated with chemically impregnated carbons, submicrometer particulate carbon emissions can easily double if the submicrometer fraction of the native fly ash has a low carbon content. Increasing sorbent injection rates, larger collection efficiency differentials as compared to fly ash, and decreasing sorbent particle size all lead to increases in the estimated submicrometer particulate carbon emissions.

■

scale PAC injection testing at the Brayton,6 Meramec,7 Monroe,8 and Pleasant Prairie9 sites did not negatively impact stack opacity. However, testing of 18 different sorbents at Conesville10 resulted in increased ESP sparking, decreased ESP power, or increased opacity in most cases. At Stanton Unit 1,11 particulate filters used in conjunction with gas sampling at the ESP outlet were darkened, although these may have reflected load changes. At Limestone Unit 112 roughly half of the particulate loading measurements (EPA Method 17) taken during PAC injection exceeded the normal range of baseline measurements taken without PAC injection. At the Lausche site,13 observed opacity increases were highly dependent on particle size and injection rate. Injection of PAC with a mass median diameter (MMD) of 20 μm yielded a constant opacity of 5% for injection rates up to 8 lb/MMacf (pounds of PAC injected per million actual cubic feet of flue gas). However, opacity nearly doubled to 9% when MMD was reduced to 5 μm (at 2.5 lb/MMacf), and more than tripled to 15−16% when MMD was reduced to 1 μm (at 1.5 lb/MMacf). The penetration of injected PAC through ESPs is problematic to the extent that (1) submicrometer PAC emitted into the atmosphere behaves like combustion-derived black carbon (BC), (2) compared to BC emissions into the atmosphere due

INTRODUCTION Among the various strategies for reducing mercury emissions from coal combustion, injection of powdered sorbents has been extensively tested and demonstrated at full-scale. Although most such tests have been conducted in the U.S., international momentum is growing as well. In 2009, the Governing Council of the United Nations Environment Programme (UNEP) approved Decision 25/5, mandating the pursuit of a global, legally binding instrument for reducing mercury emissions into the environment. A large majority of sorbent injection tests have involved injection of mercury sorbents upstream of an electrostatic precipitator (ESP). This reflects the dominance of ESPs in use as particulate control devices at coal-fired power plants: Approximately 70% of plants in the U.S.,1 95% of plants in India,2 55% of plants in Russia3 and 88−90% of plants in China4 use ESPs for particulate matter (PM) control. Despite the predominance of ESPs installed at coal-burning power stations, the challenging experimental environment they present has prevented detailed, systematic examination of how mercury sorbents behave. ESP collection efficiency is known to be highly dependent on the resistivity of the particulate matter (PM). Whereas optimum values of resistivity range from 108−1013 ohm-cm and values for fly ash (FA) from coal combustion typically fall in or near this range, PAC has a value of ∼1 ohm-cm,5 indicating lower ESP collection efficiencies than those for FA, though the degree of difference has not been determined. Fullscale sorbent injection test results are mixed on the issue. Full© 2012 American Chemical Society

Received: Revised: Accepted: Published: 7327

January 31, 2012 May 30, 2012 June 4, 2012 June 4, 2012 dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

Article

to unburned carbon on fly ash, emissions of submicrometer PAC constitutes a significant increase, and (3) the climate forcing potential (CFP) of emitted submicrometer particulate carbon (both PAC and on fly ash) is either significantly positive or significantly negative compared to the CFPs of other constituents emitted from the same source. In this regard, there is an obvious convergence: coal combustion represents the potential for large emissions of both mercury and CO2, and in the case of PAC injection, the potential exists that efforts to reduce emissions of the former will offset reductions of the latter. The present analysis seeks to estimate the relative importance of PAC collection efficiency within and PAC penetration through an ESP and the resultant impacts on emitted submicrometer particulate carbon from coal combustion.

about 74 440 short tons, which is reasonably less than the NEI estimate of 94 000 short tons for the filterable PM2.5 emissions from all U.S. CFPPs in 2005.25 ESP performance is considered in terms of an overall PM collection efficiency (both fly ash and PAC), a collection efficiency differential between the supermicrometer and submicrometer PM (both fly ash and PAC), and finally a collection efficiency differential between the submicrometer fly ash and the submicrometer PAC. The overall PM collection efficiency for the ESP was assumed to be 99.5%, however, older ESPs in the U.S. have collection efficiencies as low as 90%;26 similarly, two-thirds of ESPs in the countries comprising the Newly Independent States of the former Soviet Union (NIS) exhibit collection efficiencies as low as 88%.3 While reasonable values for the overall ESP collection efficiency exist, submicrometer collection efficiencies for ESPs are highly variable and uncertain. While electrostatic particle collection theory predicts particle collection efficiency decreasing from 55% for 1 μm particles to 35% for 0.3 μm particles,27 lab-scale results differ widely. Kim and Lee28 and Zhuang et al.29 reported ESP submicrometer collection efficiencies ranging from 60% to 80%, Sung et al.30 reported higher performance (82−90%), whereas Brocilio et al.31 reported lower (1−30%). In measurements taken from full-scale ESPs at Australian CFPPs, Nelson et al. reported ESP submicrometer collection efficiencies of 88−98%.32 The present analysis assumes a collection efficiency differential between supermicrometer PM and submicrometer PM, where PM includes both PAC and FA, of 20% thus yielding a submicrometer PM collection efficiency of 79.5%. This value is lower than the full-scale results of Nelson et al.32 but at the upper end of the range of reported lab-scale results. Although the substantial uncertainty in ESP submicrometer PM collection efficiency might seem intractable, when presented in terms of percentage increases in submicrometer particulate carbon emissions, the results of the analysis are actually independent of submicrometer PM collection efficiency because it affects both submicrometer PAC and submicrometer fly ash equally. The analysis assumes only negative values for the collection efficiency differential between the submicrometer FA and the submicrometer PAC (i.e., lower ESP performance for PAC), based on the knowledge that PM resistivity is a key factor in ESP performance, and that the value for PAC (∼1 ohm-cm)5 falls well outside of the optimum range (108 − 1013 ohm-cm).33 Although it is assumed that injected PAC will agglomerate with FA particles during collection on the ESP electrodes and subsequent resuspension, and although it is known that the degree of climate forcing exerted by particulate carbon depends on to what degree it is externally or internally mixed within an agglomerate,34 the present analysis only considers the increase in particulate carbon emissions due to additional PAC escaping an ESP and does not consider second-order effects such as the behavior of fly ash-PAC agglomerates and the efficiency with which they are collected in an ESP. Although wet scrubbers may be present downstream of an ESP, the analysis assumes that the particle capture dynamics in a wet scrubber are sufficiently similar between FA and PAC that the same collection efficiency applies for both, thus yielding the same percent increase in submicrometer particulate carbon emissions with or without a wet scrubber, although absolute emissions would be lower for an ESP followed by a wet scrubber than for an ESP alone.

■

MATERIALS AND METHODS The analysis uses the 2005 U.S. electric power generation from coal (1990.5 × 109 kWh)14 to infer the associated flue gas volume, based on an assumed flue gas production of 323 m3/ GJ.15 Applying PAC injection at representative injection rates (lbs/MMacf) to 70% of this flue gas volume then gives a total mass of PAC injected upstream of ESPs nationwide, 3.5% of which is submicrometer in size based on measured particle size distributions of conventional PAC.16 Basing the analysis on an assumption that all ESP-equipped coal-fired power plants (CFPPs) would choose PAC injection admittedly trades the precision of modeling a specific configuration or scenario for broader results of a more general nature. However, this is justified based on the large variability in CFPP configurations and operating conditions, as well as the high degree of uncertainty in the BC content of emitted fly ash (see additional discussion in the Results section). It is acknowledged that PAC injection is but one of several options available to unit operators intending to implement mercury emissions reductions. To estimate emissions of particulate carbon associated with submicrometer fly ash (FA) emitted from ESPs, the analysis assumes a 2005 U.S. coal consumption for electric power generation of 1.0375 × 109 short tons17 having 10% ash content.18 After combustion, the ash either deposits on surfaces as slag, or falls as bottom ash, or is entrained in the flue gas as supermicrometer FA particles, or nucleates as flue gas temperatures cool to form submicrometer FA particles. The fraction of coal ash that, after combustion, forms submicrometer FA typically is less than a few percent and is sensitive to coal composition and combustion temperature.19−21 Neville et al.22 determined using laminar drop tube combustion experiments that the percentage of ash in coal ultimately converted into submicrometer PM increased from 0.1% at 1800 K to 20% at 2800 K. Similarly, Buhre et al.23 reported that submicrometer FA on average constituted 0.25% of the ash in coal combusted in a laboratory drop tube at 1400 °C and under 21% oxygen. Zhang et al.24 found similar results, reporting that as combustion temperature increased, the percentage of coal ash converted into submicrometer FA increased from approximately 0.2% to 1.5%. The present analysis assumes a value of 0.5% for the fraction of coal ash ultimately forming submicrometer FA after combustion. This value, when combined with the 2005 U.S. coal consumption for power generation, 10% ash content, and 79.5% ESP collection efficiency for submicrometer PM, yields submicrometer PM emissions from coal-fired power plants equipped with ESPs of 7328

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

■

Article

RESULTS The percentage increase in submicrometer particulate carbon emissions chiefly depends on the PAC injection rate, which is controlled, and the ESP collection efficiency differential between submicrometer fly ash (FA) and submicrometer PAC, which is currently unknown. As noted above, presenting the results of the analysis in terms of percentage increases in submicrometer particulate carbon emissions renders them independent of parameters that impact both fly ash and PAC equally, such as assumed overall ESP collection efficiency and the assumed collection efficiency of submicrometer PM (both FA and PAC). Figure 1 shows estimated changes (increases) in Figure 2. Estimated percentage increases in submicrometer particulate carbon emissions from ESPs during brominated PAC injection. Assumes 40% carbon content in the submicrometer fly ash.35.

Figure 3. Estimated percentage increases in submicrometer particulate carbon emissions from ESPs during the injection of finely ground, brominated PAC. Assumes 40% carbon content in the submicrometer fly ash.35.

BC content. The top panel of Figure 1, assuming a 40% BC content in the submicrometer FA, presents estimates of submicrometer particulate carbon emission increases during injection of conventional PAC. For the most optimistic scenario of no differential in ESP collection efficiency between submicrometer FA and submicrometer PAC, submicrometer particulate carbon emissions increase from 3 to 19%. Higher injection rates and increasing collection efficiency differential produce larger increases up to the worst case scenario, that is, the highest sorbent injection rate and highest collection efficiency differential, yielding an estimated 65% increase in submicrometer particulate carbon emissions. By comparison, the bottom panel of Figure 1 presents estimates of submicrometer particulate carbon emissions increases, also for injection of conventional PAC, assuming only 0.6% BC content in the submicrometer FA (ref 34 and references therein). Clearly, the assumption of much lower BC content in the submicrometer FA leads to PAC injection having a much greater impact on estimated increases in submicrometer particulate carbon emissions: The most optimistic scenario (no collection efficiency differential, lowest PAC injection rate) produces an estimated 211% increase in submicrometer particulate carbon emissions, that is, roughly a tripling of emissions. Newer mercury sorbent formulations, particularly bromineimpregnated PACs, have repeatedly shown much improved performance over conventional PACs, achieving mercury removal efficiency targets under the most challenging conditions (e.g., high concentrations of SO3) and at much

Figure 1. Estimated percentage increases in submicrometer particulate carbon emissions from ESPs during injection of conventional PAC. Upper: Assumes 40% carbon content in submicrometer fly ash.35 Lower: Assumes 0.6% carbon content in submicrometer fly ash (ref 34 and references therein).

submicrometer particulate carbon emissions from ESPs, subject to the assumptions stated above, as a function of both PAC injection rate and the differential ESP collection efficiency between submicrometer PAC and submicrometer FA. Both panels in Figure 1 assume injection of conventional PAC, which particle size distribution measurements have indicated has roughly 3.5% of its mass in particles smaller than 1 μm.16 Because conventional PAC is neither finely ground nor chemically impregnated, relatively high injection rates are reflected in Figure 1, representative of those used in early fullscale trials. The results for conventional PAC in Figure 1 are primarily of historical interest, as conventional PAC has been rendered largely obsolete as a result of continued developments in the formulation of mercury sorbents. Results for more advanced PAC formulations are shown in Figures 2 and 3. The BC content of the submicrometer fraction of the FA is not wellknown, and thus the upper and lower panels in Figure 1 represent results obtained for two bounding values, 40%35 and 0.6% (ref 34 and references therein), of submicrometer fly ash 7329

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

Article

Figure 4. Estimated U.S. BC emissions, including PAC emissions, by sector. Increases in stationary power sector emissions reflect PAC emissions resulting from differential ESP submicrometer PM collection. Assumed PAC injection market shares of U.S. stationary power sector of 10%, 50%, and 90% shown. Except for “No ACI”, all cases assume PAC injection rate of 2 lbs/MMacf. Left: Derived from emissions inventory of Bond et al.34 Right: BC emissions from transport sector reduced by 40% from figure at left.

during full-scale testing.37,38 Although these effects have so far limited the performance of finely ground sorbents, it is possible that sorbent formulations or sorbent feeding methods could progress such that such fine powders could be successfully injected into the flue gas, thereby providing mass transfer advantages for mercury capture. Figure 3 presents estimates of submicrometer particulate carbon emissions increases associated with the injection of a finely ground brominated PAC, assuming a 40% carbon content in the submicrometer FA as was similarly assumed in Figure 1 (upper panel) and Figure 2. As compared to conventional PAC (Figure 1) and brominated PAC (Figure 2) in which the submicrometer fractions represented 3.5% of the total mass, based on measured PSDs for a finely ground PAC with a mean particle size of 6 μm,38 the finely ground brominated PAC in Figure 3 is assumed to contain 12.5% of its total mass in the submicrometer fraction. Figure 3 shows that for finely ground brominated PAC under the most optimistic scenario of no collection efficiency differential and 40% carbon content in the submicrometer FA, submicrometer particulate carbon emissions increased from 5 to 23% as compared to the 1 to 6% increases estimated for the more coarse brominated PAC in Figure 2. In the worst-case scenario of 5 lb/MMacf injection rate and −50% ESP collection efficiency differential, submicrometer particulate carbon emissions increased by 78%. Assuming a lower carbon content for the submicrometer FA (0.6%) yields estimates of submicrometer particulate carbon emissions increases that are generally 2 orders of magnitude larger than those shown in Figure 3. It is important to place these estimates of increased submicrometer particulate carbon emissions into context. As a climate forcing agent, the effects of black carbon (BC) are complex and generally less well understood than those of CO2 Though the 2007 Assessment Report of the Intergovernmental Panel on Climate Change (IPCC)39 estimates the direct forcing effects of CO2, BC, and CH4 to be 1.66, 0.05−0.55, and 0.48 W/m2, respectively, Ramanathan and Carmichael (ref 40 and references therein) arrive at a BC direct forcing value of 0.9 W/ m2 (ranging between 0.4 and 1.2 W/m2), which they assert

lower injection rates. Figure 2 shows estimates of submicrometer particulate carbon emissions increases for brominated PAC injection. The underlying assumptions applied for brominated PAC injection were (1) identical particle size distribution as conventional PACs,16 (2) lower injection rates than conventional PACs, and (3) no impact of bromination on ESP collection efficiency differential (based on lab-scale results16). Estimated increases in submicrometer particulate carbon emissions for brominated PAC injection under the most optimistic scenario (no collection efficiency differential, 40% carbon content in submicrometer FA) vary from 1 to 6%, with a maximum 22% increase in submicrometer particulate carbon emissions at the highest injection rate (5 lb/MMacf) and for the highest collection efficiency differential (−50%) between submicrometer FA and submicrometer PAC. As noted earlier, assuming a lower carbon content in the submicrometer FA greatly increases the impact of PAC injection on submicrometer particulate carbon emissions. For a carbon content of 0.6% in the submicrometer FA, the most optimistic scenario (no collection efficiency differential) for brominated PAC injection yields increases in submicrometer particulate carbon emissions from 84% to 421%, i.e. a nearly two- to 5-fold increase, for injection rates ranging from 1 to 5 lb/MMacf. In addition to chemical impregnation, more finely ground powdered sorbents have been developed, or produced on-site, in order to increase available external sorbent surface area to reduce mass transfer resistance. However, finely ground PACs have generally not shown any improved performance over more coarse products of the same chemical formulation.13 One potential explanation that has been put forth has been that powdered sorbents tend to agglomerate to various degrees during the feeding and injection process, depending on the sorbent and the length and materials of construction of the feed line. Agglomeration can be promoted by either by high particle mass loading or by triboelectric charging during pneumatic feeding, effectively shifting the particle size distribution (PSD) to a more coarse state. The effects of such phenomena on the sorbent PSD during pneumatic feeding have been demonstrated at lab scale36 and have subsequently been observed 7330

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

Article

value of 7.2 Gg/y as the 1996 baseline for U.S. BC emissions from coal-based stationary power generation. As ref 34 does not provide subdivisions of BC emissions by country, constructing the U.S. BC emissions inventory in Figure 4 required values for the transport, residential, and industrial combustion sectors, as well as open biomass burning. For BC emissions from U.S. transport, residential, and industrial combustion sectors, the respective North American values were apportioned to each country based on stationary power generation (U.S. representing 94% of N. American total) based on the rationale that contained combustion in North America should, to a first approximation, scale uniformly with economic activity. For BC emissions from U.S. open biomass burning, the North American value (116 Gg/y) is apportioned to each country according to land area (U.S. representing 45% of North American total). The resulting estimates for U.S. BC emissions from transportation (262.8 Gg/y), residential (73.4 Gg/y), and industrial (16.6 Gg/y) combustion, when combined with the estimates for U.S. stationary power combustion (7.2 Gg/y) and open biomass combustion (52.3 Gg/y), lead to a 1996 baseline for total BC emissions in the U.S. of 412 Gg/y. Starting with this total of U.S. BC emissions from all sectors (for 1996), Figure 4 presents results illustrating both the magnitude of U.S. BC emissions from coal combustion relative to other sectors, as well as the potential increases resulting from emitted PAC. Figure 4 (left or right) presents and compares fifteen BC inventories: four representing different values of differential ESP collection efficiency between submicrometer fly ash and submicrometer PAC (0%, −10%, −25%, −50%), plus the case of no PAC injection, and three representing different PAC injection market share scenarios of 10%, 50%, and 90% of U.S. coal-fired electric generating units employing PAC injection. No market analyses or projections are implied for these market share scenarios; rather, they are provided only to illustrate their influence on particulate carbon emissions resulting from emitted PAC. For all cases in Figure 4, the carbon content of the emitted fly ash is assumed to be 0.6% (in agreement with 34) and the PAC injection rate is 2 lb/MMacf. In Figure 4 (left), for a small 10% market share employing PAC injection, the baseline BC emissions rate of 412 Gg/y increases to 413.5 Gg/y in the best-case scenario of no difference in ESP collection efficiency between submicrometer fly ash and submicrometer PAC. As the magnitude of the differential ESP collection efficiency increases, total U.S. BC emissions increase only slightly, limited by the small assumed market share: 414 Gg/y at −10% differential ESP collection efficiency up to 416.5 Gg/y at −50%. Thus, for a 10% PAC injection market share, total U.S. BC emissions would increase by less than 1% even for the largest differential ESP collection efficiency value considered here. Conversely, for a large PAC injection market share (90%), total U.S. BC emissions increase from the baseline of 412 Gg/y to 423 Gg/y in the best-case scenario of no difference in ESP collection efficiency between submicrometer fly ash and submicrometer PAC. With increasing differential ESP collection efficiency, total U.S. BC emissions increase: 428.5 Gg/y at −10% differential ESP collection efficiency, 436.5 Gg/y at −25%, up to 450 Gg/y at −50%. The corresponding percentage increases in total U.S. BC emissions as a result of emissions increases from stationary power combustion are 2.7% (at 0% coll. eff. differential), 4% (at −10%), 6% (at −25%), and 9.2% (at −50%). The results in Figure 4 (left) represent the impacts of emitted PAC on total U.S. BC emissions determined using

would be second only to CO2 in magnitude and greater than all other GHGs. In terms of BC source apportionment, global BC emissions are dominated by open biomass combustion which accounts for about 37%.34,41 Diesel-fueled transportation (approximately 13%) and residential wood combustion (approximately 11%) are also large sources globally.34,41 The contribution to global BC emissions associated with fly ash emitted from coal-fired power plants (CFPPs), as noted earlier, is uncertain, with estimates ranging from 10%41 to less than one-tenth of one percent.34 To complement Figures 1 through 3 and provide some perspective of the absolute scale of emitted submicrometer PAC as a contributor to submicrometer particulate carbon emissions associated with coal combustion, the results in Figure 4 represent the application of the estimated percentage increases from Figure 2 to BC emissions from U.S. coal combustion as reported in the BC emissions inventory of Bond et al.34 This BC emissions inventory is particularly compatible with the present analysis because both restrict their considerations of particulate carbon to submicrometer PM, a stricter standard than the PM2.5 size standard more generally used (e.g., ref 39). To this BC emissions inventory are applied the percentage increases from Figure 2, representing estimates for brominated PAC. In their global BC inventory, Bond et al.34 reported total North American (i.e., Canada, Mexico, and the U.S.) BC emissions based on 1996 fuel usage of 498 gigagrams per year (Gg/y).34 Of this total, only 1.5% is attributed to stationary power generation, with much larger percentages attributed to transport (56%), residential combustion (15%) and open biomass burning (23%); industrial combustion accounted for 4% of the total.34 Of the 498 Gg/y, open biomass burning accounts for 116 Gg/y,34 leaving 382 Gg/y as the baseline for all so-called “contained combustion” (i.e., residential, industrial, transportation, and stationary power combustion) in North America. Stationary power generation in North America in 1996 was dominated by U.S. generation: 94% (3 444 187 621 MWh)17 as compared to 3% each for Canada (110 320 346 MWh)42 and Mexico (100 256 410 MWh).43 Of U.S. stationary power generation in 1996, 52% originated from combustion of coal (1 795 195 593 MWh) as compared to 12% from combustion of natural gas (455 055 576 MWh).17 Although both fuels would seemingly contribute to these BC emissions estimates, Bond et al.34 cite and use an overall BC emission factor (g of BC/kg dry fuel) of 0 for natural gas-fuelled stationary power generation. This is likely a combined effect of the assumed PM emission factors (EFPM, g of PM/kg of fuel) for the two fuels (1.3−33 for coal, 0.002 for natural gas), the assumed fractions of submicrometer PM produced (F1.0) (1.0 for natural gas, 0.09−0.33 for coal), the assumed fraction of submicrometer PM present as BC (FBC) (0.06 for natural gas, 0.006−0.2 for coal), and the assumed fraction of submicrometer PM penetrating any PM control(s) present (Fcont) (1.0 for natural gas, 0.03−1.0 for coal).34 As a consequence, virtually all U.S. BC emissions from stationary power generation are attributable to coal combustion in this inventory. To the baseline N. American BC emissions from contained combustion (382 Gg/y) we apply a) the percentage originating from stationary power generation, b) the percentage of stationary power generation based in the U.S., and c) the percentage of combustion-based stationary power generation in the U.S. derived from coal (effectively 100% for U.S. BC emissions in this inventory, as discussed above). The result is a 7331

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

Article

1996 fuel use and emissions factors,34 a period when BC emissions from the transportation sector in the U.S. had barely begun to decline. In the 15 year interval between 1990 and 2005, BC emissions in the U.S. decreased by 30%, and are projected to decrease by a further 80% (from 2005 levels) by 2030, primarily due to the implementation of PM controls for transportation diesel engines.44 Figure 4 (right) illustrates how the increases in particulate carbon emissions associated with PAC injection compare to total U.S. BC emissions where emissions from transportation have been reduced by 40% from their 1996 values (BC emissions from residential, industrial, and open biomass combustion are unchanged from Figure 4 (left)). Whereas the greatest PAC injection market share (90%) and the greatest differential ESP collection efficiency (−50%) resulted in a 9.2% increase in total U.S. BC emissions based on 1996 data, a 40% reduction in BC emissions from transportation that is more representative of conditions in the late 2000s increases the impact of PAC emissions, yielding a 12.4% increase in total U.S. BC emissions for the same conditions (Figure 4 (right)). The quantitative conclusions drawn from Figure 4 depend heavily on the U.S. BC emissions inventory employed; specifically, and as noted earlier, the inventory of Bond et al. assumes exceedingly small fractions of particulate carbon (FBC) contained within submicrometer PM. For such low-carbon PM, the addition of PAC, even in small mass concentrations, results in the very large percentage increases in BC emissions discussed earlier. As FBC of the combustion PM increases, the percentage increase in BC emissions associated with PAC emissions decreases (see Figure 1). However, baseline BC emissions increase with increasing FBC, such that the net result would be smaller percentage increases in larger BC emissions rates that would also constitute a larger portion of total U.S. BC emissions.

■

(5) Espinola, A.; Miguel, P. M.; Salles, M. R.; Pinto, A. R. Electrical properties of carbonsResistance of powder materials. Carbon 1986, 24, 337−341. (6) ADA-ES. Final Site Report for Brayton Point Generating Station Unit 1: Sorbent Injection into a Cold-Side ESP for Mercury Control; 2005. (7) ADA-ES. Evaluation of Sorbent Injection for Mercury Control: Topical Report for AmerenUE’s Meramec Station Unit 2; U.S. Department of Energy, National Energy Technology Laboratory, 2005. (8) ADA-ES. Evaluation of Sorbent Injection for Mercury Control: Topical Report for DTE Energy’s Monroe Station; U.S. Department of Energy, National Energy Technology Laboratory, 2006. (9) ADA-ES. Pleasant Prairie Power Plant Unit 2 - Sorbent Injection into a Cold-side ESP for Mercury Control. Final Report; U.S. Department of Energy, May, 2003. (10) Sjostrom, S. Evaluation of Sorbent Injection for Mercury Control: Final Report for Sunflower Electric’s Holcomb Station, AmerenUE’s Meramec Station, American Electric Power’s Conesville Station, DTE Energy’s Monroe Power Plant, Missouri Basin Power Project’s Laramie River Station and AmerenUE’s Labadie Power Plant; U.S. Department of Energy, National Energy Technology Laboratory, 2008. (11) Dombrowski, K., Richardson, C.; Padilla, J.; Chang, R.; Archer, G.; Fisher, K.; Smokey, S.; Brickett, L. Evaluation of Novel Mercury Sorbents and Balance-of-Plant Impacts at Stanton Unit 1. In DOE-U.S. EPA-EPRI-AWMA Power Plant Air Pollution Control ″Mega″ Sympsoium, Baltimore, MD, 2008. (12) Dombrowski, K.; Richardson, C.; Padilla, J.; Fisher, K.; Campbell, T.; Chang, R.; Eckberg, C.; Hudspeth, J.; O’Palko, A.; Pletcher, S. Evaluation of low ash impact sorbent injection technologies for mercury control at a Texas lignite/PRB fired power plant. Fuel Process. Technol. 2009, 90, 1406−1411. (13) Nelson, S., Landreth, R.; Zhou, Q.; Miller, J.; , Mercury Sorbent Injection Test Results at the Lausche Plant. In DOE-U.S. EPA-EPRIAWMA Power Plant Air Pollution Control ″Mega″ Symposium, Washington, DC, 2003. (14) U.S. Energy Information Administration. U.S. Historical Electricity Generation by Source. http://www.eia.gov/electricity/ monthly/excel/epmxlfile1_1.xls. (15) Beychok, M. R. Fundamentals of Stack Gas Dispersion, 4th ed.; Milton R. Beychok, 2005. (16) Prabhu, V.; Kim, T.; Khakpour, Y.; Serre, S.; Clack, H. L. On the Electrostatic Precipitation of Fly Ash-Powdered Mercury Sorbent Mixtures. Fuel Process. Technol. 2012, 93, 8−12. (17) Annual Energy Review. http://www.eia.gov/totalenergy/data/ annual/showtext.cfm?t=ptb0802b (accessed Jan. 29, 2012). (18) Co., B. W. Steam: Its Generation and Use. 41st ed.; Babcock & Wilcox Co., 2005. (19) Chen, Y.; Shah, N.; Huggins, F. E.; Huffman, G. P.; Linak, W. P.; Miller, C. A. Investigation of primary fine particulate matter from coal combustion by computer-controlled scanning electron microscopy. Fuel Process. Technol. 2004, 85, 743−761. (20) Damle, A. S.; Ensor, D. S.; Ranade, M. B. Coal combustion aerosol formation mechanisms: A review. Aerosol Sci. Technol. 1982, 1 (1), 119−133. (21) Taylor, D. D.; Flagan, R. C. The influence of combustor operation on fine particles from coal combustion. Aerosol Sci. Technol. 1982, 1 (1), 103−117. (22) Neville, M.; Quann, R. J.; Haynes, B. S. Vaporization and condensation of mineral matter during pulverized coal combustion. Proc. Combust. Inst. 1981, 18, 1267−1274. (23) Buhre, B. J. P.; Hinkley, J. T.; Gupta, R. P. Fine ash formation during combustion of pulverised coal - coal property impacts. Fuel 2006, 85 (2), 185−193. (24) Zhang, L.; Ninomiya, Y.; Yamashita, T. Formation of submicron particulate matter (PM1) during coal combustion and influence of reaction temperature. Fuel 2006, 85, 1446−1457. (25) EPA. U. S., PM2.5 Filterable and PM10 Filterable emissions trends for Electric Generating Utilities for 1970 to 2005−August 2008. In Process Optimization Guidance for Reducing Mercury Emissions from

AUTHOR INFORMATION

Corresponding Author

*Current address: Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, Michigan, USA. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS I thank Dr. William Linak of the U.S. EPA National Risk Management Research Laboratory (U.S. EPA-NRMRL) and Dr. James Staudt of Andover Technology Partners for their very helpful insights and discussions.

■

REFERENCES

(1) 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. (2) Lin, G. In A Discussion about Strategy of Flue Gas Dust Removal for India Coal Fired Boiler, 11th International Conference on Electrostatic Precipitation; Yan, K., Ed.; Zhejiang University Press: Hangzhou, Zhejiang Province, China, 2008. (3) Zykov, A., Jozewicz, W. Final Report: Low-Cost Upgrades of Electrostatic Precipitators in Russia and Other Countries of the NIS, 2006. (4) Wang, S. X.; Zhang, L.; Li, G. H.; Wu, Y.; Hao, J. M.; Pirrone, N.; Sprovieri, E.; Ancora, M. P. Mercury emission and speciation of coalfired power plants in China. Atmos. Chem. Phys. 2010, 10, 1182−1192. 7332

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333

Environmental Science & Technology

Article

Coal Combustion in Power Plants: A Report from the Coal Combustion Partnership Area, 2008. (26) United Nations Environment Programme, 2010. (27) Riehle, C.; Loffler, F. The effective migration rate in electrostatic precipitators: May an increase in the effective migration rate observed in electrostatic precipitators with wider plate spacings or faster gas streams really be termed “non-deutschian”? Aerosol Sci. Technol. 1992, 16, 1−14. (28) Kim, S. H.; Lee, K. W. Experimental study of electrostatic precipitators performance and comparison with existing theoretical prediction models. J. Electrost. 1999, 48, 3−25. (29) Zhuang, Y.; Kim, Y. J.; Lee, T. G.; Biswas, P. Experimental and theoretical studies of ultra-fine particle behaviour in electrostatic precipitators. J. Electrost. 2000, 48, 245−260. (30) Sung, B.-J.; Aly, A.; Lee, S-H; Takashima, K.; Kastura, S.; Mizuno, A. Fine-particle collection using an electrostatic precipitator equipped with an electrostatic flocking filter as the collection electrode. Plasma Process. Polym. 2006, 3, 661−667. (31) Brocilio, D.; Podlinski, J.; Chang, J. S.; Mizeraczyk, J.; Findlay, R. D. Electrode geometry effects on the collection efficiency of submicron and ultra-fine dust particles in spike-plate electrostatic precipitators. J. Phys.: Conf. Ser. 2008, 142, 012032. (32) Nelson, P. F.; Shah, P.; Strezov, V.; Halliburton, B.; Carras, J. N. Environmental impacts of coal combustion: A risk approach to assessment of emissions. Fuel 2010, 89, 810−816. (33) Calvert, S., Englund, H.M. Handbook of Air Pollution Technology; John Wiley & Sons: New York, 1984. (34) Bond, T. C., Streets, D.G.; Yarber, K.F.; Nelson, S.M.; Woo, J.H.; Klimont, Z.; , A technology-based global inventory of black and organic carbon emissions from combustion. J. Geophys. Res. 2004, 109. (35) Linak, W. P.; Yoo, J-I; Wasson, S. J.; Zhu, W.; Wendt, J. O. L.; Huggins, F. E.; Chen, Y.; Shah, N.; Huffman, G. P.; Gilmour, M. I. Ultrafine ash aerosols from coal combustion: Characterization and health effects. Proc. Combust. Inst. 2007, 31, 1929−1937. (36) Lee, E. M.; Clack, H. L. In situ detection of altered particle size distributions during simulated powdered sorbent injection for mercury emissions control. Energy Fuels 2010, 24, 5410−5417. (37) Dombrowski, K., McDowell, S.; Richardson, C.; Fisher, K.; Chang, R.; Eckberg, C.; Hudspeth, J.; A. O’Palko, Evaluation of lowash impact sorbent injection technologies at a Texas lignite/PRB fired power plant. In Air Quality VII; Arlington, VA, 2009. (38) DeNigris, J., Pollack, N. An assessment of finely activated carbon agglomeration using on-line laser diffraction analysis. In DOE-U.S. EPA-EPRI-AWMA Power Plant Air Pollution Control ″Mega″ Symposium, Baltimore, MD, 2010. (39) IPCC Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change; Intergovernmental Panel on Climate Change, 2007. (40) Ramanathan, V.; Carmichael, G. Global and regional climate changes due to black carbon. Nat. Geosci. 2008, 1, 221−227. (41) Cooke, W. F.; Liousse, C.; Cachier, H.; Feichter, J. Construction of a 1° × 1° fossil fuel emission data set for carbonaceous aerosol and implementation and radiative impact in the ECHAM4 model. J. Geophys. Res. 1999, 104, 2137−2162. (42) Canada Analysis; Energy Information Administration, 2011. (43) Mexico Analysis; Enegy Information Administration, 2011. (44) Rao, T., Somers, J. Black carbon as a short-lived climate forcer: a profile of emission sources and co-emitted pollutants; EPA Emissions Inventory Conference, San Antonio, TX, 2010.

7333

dx.doi.org/10.1021/es3003712 | Environ. Sci. Technol. 2012, 46, 7327−7333