Radiative Forcing Associated with Particulate Carbon Emissions

Article pubs.acs.org/est

Radiative Forcing Associated with Particulate Carbon Emissions Resulting from the Use of Mercury Control Technology Guangxing Lin,*,† Joyce E. Penner,† and Herek L. Clack‡ †

Department of Atmospheric, Oceanic, and Space Sciences, University of Michigan, Ann Arbor, Michigan, United States Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, Michigan, United States

‡

ABSTRACT: Injection of powdered activated carbon (PAC) adsorbents into the flue gas of coal fired power plants with electrostatic precipitators (ESPs) is the most mature technology to control mercury emissions for coal combustion. However, the PAC itself can penetrate ESPs to emit into the atmosphere. These emitted PACs have similar size and optical properties to submicron black carbon (BC) and thus could increase BC radiative forcing unintentionally. The present paper estimates, for the first time, the potential emission of PAC together with their climate forcing. The global average maximum potential emissions of PAC is 98.4 Gg/yr for the year 2030, arising from the assumed adoption of the maximum potential PAC injection technology, the minimum collection efficiency, and the maximum PAC injection rate. These emissions cause a global warming of 2.10 mW m−2 at the top of atmosphere and a cooling of −2.96 mW m−2 at the surface. This warming represents about 2% of the warming that is caused by BC from direct fossil fuel burning and 0.86% of the warming associated with CO2 emissions from coal burning in power plants. Its warming is 8 times more efficient than the emitted CO2 as measured by the 20-year-integrated radiative forcing per unit of carbon input (the 20-year Global Warming Potential).

■

INTRODUCTION Mercury is presently one of the most important environmental contaminants because of its cyclic transport between air, water, and soil; its tendency to bioaccumulate in the environment; the connection between its emission and the use of coal, a primary global energy source; and its toxic effects on human health. Coal-fired power plants (CFPPs) are estimated to account for 35% of global anthropogenic mercury emission to the atmosphere.1 In 2005, the U.S. EPA proposed to regulate mercury emissions from CFPPs for the first time and issued the Clean Air Mercury Rule (CAMR). In 2008, however, the DC Circuit Court vacated the CAMR. In 2012, the EPA further tightened mercury emissions via its Mercury and Air Toxics Standard (MATS), which requires existing CFPPs in the U.S. to reduce their emissions by 90% starting in 2015. Outside the U.S., the Governing Council of the United Nations Environment Programme (UNEP) agreed on a legally binding global mercury treaty (Minamata Convention on Mercury) in January 2013 that compels countries to prevent and reduce mercury emissions into the environment from all sources. In 2011, China adopted a national mercury emissions standard for CFPPs. The standard, which took effect in 2012 and whose full implementation is expected by 2015, limits mercury concentrations from CFPPs to 0.03 mg/m3.2 Among the most mature technologies for controlling mercury emissions from CFPPs is the injection into the flue gas of powdered activated carbon (PAC) adsorbents that adsorb mercury while in suspension in the flue gas. Chemically © 2014 American Chemical Society

treated PAC adsorbents have been shown to effectively reduce mercury emissions for most CFPP configurations and types of coal, both of which are important to overall reductions in mercury emissions.3−5 A typical implementation of this control technology would entail the injection of PAC sorbent upstream of an electrostatic precipitator (ESP), a particulate matter (PM) control device widely use in CFPPs. However, the electrical resistivity of PAC (104 ohm-cm6) is far below the accepted norms for optimal ESP removal efficiency (108−1012 ohmcm7,8). Dombrowski et al.9 reported results of PAC injection upstream of an ESP in which downstream particulate filters show clear evidence of darkening as upstream PAC emission rates increase, indicating increasing PAC emissions from the ESP into the atmosphere. The black-colored PAC likely behaves like submicron black carbon (BC) emitted directly from fossil fuel burning in terms of its radiative forcing and thus could increase BC forcing unintentionally. The role of BC in climate science is extremely complex and significantly less well understood than that of CO2. The recent Assessment Report of the Intergovernmental Panel on Climate Change (IPCC)10 estimates the direct climate forcing effects of CO2 and BC to be 1.82 ± 0.19 and 0.05−0.80 Wm−2, respectively, with the range resulting from predictions from Received: Revised: Accepted: Published: 10519

May 14, 2014 July 30, 2014 August 5, 2014 August 5, 2014 dx.doi.org/10.1021/es502382h | Environ. Sci. Technol. 2014, 48, 10519−10523

Environmental Science & Technology

Article

Table 1. Coal Consumption in Power Plants, PAC Emission Rate, and Fossil Fuel BC Emission Rate coal consumed in power plants (quadrillion Btu)

PAC emission rate (Gg/yr)

fossil fuel BC emission rate (Gg/yr)

Global

U.S.A

China

India

Global

U.S.A

China

India

Global

145.0

19.2

68.1

19.4

98.4

13.0

46.2

13.2

4411.0

the Introduction. Second, we assume that all coal-fired power plants will use the PAC injection technology to control mercury. It is acknowledged that PAC injection is but one of several options available to reduce to mercury emissions, and the assumption of the technology being used uniformly worldwide is primarily for the purposes of establishing the potential scale of PAC emissions and radiative forcing. The country-level emissions are then distributed on a 1° × 1° grid by using the local population as a surrogate for the PAC emission rate within each 1° × 1° subarea of a country, following the same method used by Ito and Penner16 to estimate the geographical emissions from fossil fuel burning. The national and provincial populations were mapped to the 1° × 1° grid as follows: area associated with each country or its political subdivisions was identified and placed on the 1° × 1° cells. Then the populations of cities with population greater than 50 000 were subtracted from the total national and provincial population resulting in the rural population. The rural population for each country or its political subdivisions was distributed with a uniform density among all populated cells, excluding regions that are largely unpopulated or of no human use. Finally, for each 1° × 1° grid the total population is the sum of population in cities within this grid and the rural population assigned to this grid. Then we used the PAC emissions together with the Integrated Massively Parallel Atmospheric Chemical Transport (IMPACT) model run at a 4° latitude × 5° longitude horizontal resolution with 46 vertical layers to predict PAC concentrations in the atmosphere. The model includes the microphysics of sulfate aerosol including the formation of internal mixtures of sulfate and nonsulfate aerosols (including fossil fuel BC, biomass burning BC, secondary organic aerosol, sea salt, and dust)17−19 and was run using 1997 meteorological fields from the NASA Data Assimilation Office (DAO) GEOSSTRAT model.20 For the future simulation, the PAC emissions are predicted as above, and all other species’ emissions (e.g., BC and sulfate) for the year 2030 are adopted from Representative Concentration Pathway (RCP) 2.6 scenario.21 The treatments of dry and wet deposition used here for both gas and aerosol phase species are the same as those used in Lin et al.18 We use an off-line radiative transfer model22 to calculate the optical properties of aerosols and their resulting radiative forcing.23−25 The size distribution of fossil fuel BC is prescribed with a superposition of two fixed log-normal distributions (Table 1 in Wang et al.24). Specifically, 43% of the fossil fuel BC number concentration is represented by a log-normal distribution with a median radius rg = 0.005 μm and a geometric standard deviation sg = 1.5, while 57% is represented by a log-normal distribution with rg = 0.08 μm and sg = 1.7. A refractive index of 1.8 + i0.74 at 550 nm for fossil fuel BC is adopted from Bond and Bergstrom.26 Fossil fuel BC is assumed to internally mix with fossil fuel OA and coated sulfate. The size distribution, the optic properties, and the mixing state of PAC are calculated in the same manner as that for fossil fuel BC (i.e., internally mixed with coemitted organics and coated sulfate). The forcing associated with fossil fuel BC or PAC was

different models. The report shows that warming by BC could be second only to CO2 in magnitude and greater than all other GHGs. Additional particulate carbon emissions resulting from efforts to comply with regulatory limits for mercury emissions would increase further this climate forcing potential. Recent analysis by Clack11 projects that increases in submicron (PM1) particulate carbon emissions of several orders of magnitude are possible in cases where PAC injection is applied to a CFPP that generates low-carbon PM1 fly ash. In this paper, we aim to estimate the potential increase in secondary BC emissions in the future, those not produced from combustion but arising instead from widespread adoption of PAC injection to meet environmental regulations. We also calculate the radiative forcing associated with these secondary BC emissions by using a global atmospheric chemical transport model coupled with a radiative transfer model.

■

MATERIALS AND METHODS To estimate the amount of PAC (E) emitted into the atmosphere for each country, we used the following equation: E = Ccoal × Vgas × R inj × F1.0 × (1 − f )

(1)

where Ccoal is the amount of coal consumed in each country (GJ), Vgas is the volume of wet exhaust gas per GJ of coal burned, Rinj is the PAC injection rates into the flue gas; F1.0 is the mass fraction of submicron PAC, and f is the fraction of submicron PAC that is collected (collection efficiency) by ESP. The coal consumption data in power plants for each country for year 2030 is based on the World Energy Outlook 2011 from International Energy Agency (IEA).12 Vgas is set to 323.1 m3/ GJ, a typical value adopted from Beychok.13 We assume that the PAC particles smaller than 1.0 μm account for about 3.5% of total PAC mass, which is based on particle size measurements.14 The submicron PAC collection efficiency for ESP is currently unknown,11 because (1) the collection efficiency of submicron PM, which includes both PAC and fly ash, is highly variable and uncertain and (2) the collection efficiency difference between submicron PAC and fly ash is unknown. Clack11 assumed a submicron PM collection efficiency of 79.5% and a collection efficiency difference between submicron PAC and fly ash ranging from 0 to −50%, yielding a range of possible submicrometer PAC collection efficiencies from 79.5 to 29.5%. In this study, we assumed a collection efficiency of 29.5%, so that most submicrometer PAC escapes the ESP and is emitted into the atmosphere. The PAC injection rate is determined by the coal type burned and the Hg removal efficiency that one wants to achieve.15 Here we set it to be 5 lbs/MMacf, the high end of the range of injection rates of finely ground brominated PAC considered in an analysis by Clack.11 This high injection rate together with the low submicron collection efficiency for ESP allows us to estimate the largest potential impacts of PAC emissions in the atmosphere. There are two implicit assumptions for eq 1. The first is that all coal-fired power plants will be equipped with ESP for PM collection. This reflects the dominance of ESPs in use as particulate control devices at coal-fired power plants as stated in 10520

dx.doi.org/10.1021/es502382h | Environ. Sci. Technol. 2014, 48, 10519−10523

Environmental Science & Technology

Article

Figure 1. Global distribution of potential PAC emission rates (a) in units of 10−12 kg m−2 s−1 and burden (b) in units of mg m−2 for year 2030.

Figure 2. Global distribution of direct radiatve forcing caused by PAC emissions for year 2030 at the top of atmosphere (TOA) and surface (units: mW m−2).

calculated from the difference in the radiation flux with all aerosols and that with all aerosols except fossil fuel BC or PAC.

determined by the pattern of PAC burden and by the surface albedo distribution, because PAC absorbs not only the directly emitted sunlight but also the sunlight reflected by the Earth’s surface. Over surfaces with high albedo (e.g., the arctic region), PAC can absorb abundant sunlight reflected from the surface and cause a warming. On a global basis, the PAC emissions can cause a warming of 2.10 mW m−2 at the top of atmosphere (TOA), a cooling of −2.96 mW m−2 at the surface, and an atmospheric absorption of 5.06 mW m−2 for all-sky conditions (i.e., in the presence of clouds), while for clear-sky conditions the forcing is 1.42 mW m−2 at the TOA and −4.01 mW m−2 at the surface (Table 2). The stronger warming in all-sky conditions is due to the existence of the low level clouds with a high albedo, behaving like the regions of high albedo at the Earth’s surface. The strongest warming at the TOA is over East Asia, where the warming can be as high as over 60 mW m−2.

■

RESULTS Table 1 shows the predicted global coal consumption in power plants, PAC emissions, and BC emissions from fossil fuel burning in year 2030. The whole world is projected to use 145 quadrillion Btu coal in power plants, of which 47% will be consumed by China, 13% by the U.S., and 13% by India. Consequently, the maximum global PAC emissions are predicted to be 98.4 Gg. Since the amount of PAC injected into the flue gas is proportional to the amount of coal burned in the power plants, the distribution of PAC emissions among the countries is the same as that of the coal consumption. Not surprisingly, China would be the country with the largest PAC emissions, corresponding to the largest coal consumption in power plants. For a comparison, Table 1 also lists the global BC emission from fossil fuel burning in year 2030. The PAC emissions rate is about 2.2% of the fossil fuel BC emission rate. Figure 1a depicts the global distribution of the PAC emission rate. As expected, the largest emissions occur in South and East Asia, North America, and Europe. The local emissions can be transported by wind within the atmosphere to remote regions (e.g., ocean or arctic regions), which is illustrated by the spread of local plume in the PAC burden plot (Figure 1b). The global distribution of direct radiative forcing resulting from PAC emissions is shown in Figure 2. Its spatial pattern is

Table 2. Global PAC Burden and Its Resulting Direct Radiative Forcing

PAC burden (Gg) 0.80 10521

clear-sky forcing (mW m−2)

all-sky forcing (mW m−2)

atmospheric absorption (mW m−2)

TOA

surface

TOA

surface

clear-sky

all-sky

1.42

−4.01

2.10

−2.96

5.43

5.06

dx.doi.org/10.1021/es502382h | Environ. Sci. Technol. 2014, 48, 10519−10523

Environmental Science & Technology

Article

scale injection rates of up to 32 lbs/MMacf have been reported in the past using early versions of PAC products. Most power plants may be capable of meeting the current regulations by using injection rates less than 5 lbs/MMacf or in conjunction with mercury removal within other air pollution control devices. However, this is not always possible even for similar plant/fuel configurations. In addition, more demanding regulation in the future might require a higher injection rate. Finally, the more advanced current-day PAC products have enabled the sharp reductions in PAC injection rates in the U.S. However, there are multiple ongoing patent infringement lawsuits relating to this technology, which raise the likelihood that non-U.S. markets may resort to cheaper, less advanced PAC products that present little or no legal liability, but which require higher injection rates. In addition to the above uncertainties, one limitation of this work is the use of the population density as a proxy to allocate the country level PAC emissions to a 1° × 1° grid. This does not account for the exact location of coal-fired power plant units, while it generally represents the distribution of coal-fired power plants in China and India,28 which would account for around 60% of coal-fired power plants in the world in year 2030. In the future, a higher-level PAC emission distribution (e.g., a state or province-level) might be needed to better represent the distribution of coal-fired power plants and thus the global PAC emission distribution, especially on smaller scales. Finally, the radiative effect of PAC is sensitive to its size. Due to a lack of knowledge of PAC size information, we assume the same size distribution for PAC as fossil fuel BC. The PAC size distribution needs to be updated when the information is available in the future.

The global average radiative forcing associated with fossil fuel BC is 0.13 W/m2. The global mean radiative forcing resulting from PAC emission is about 1.6% of that from fossil fuel burning BC. We also calculated the CO2 emissions resulting from coal burning in power plants. We use the CO2 emission factor of 228.3 lb/MMBtu for anthracite, 205.9 lb/MMBtu for bituminous coal, 213.9 lb/MMBtu for sub-bituminous coal, and 212.5 lb/MMBtu for lignite.27 The global total CO2 emission rate is estimated as 1.38 × 107 Gg for year 2030, which is 1.4 × 105 times the PAC emission rate. The ratio of warming caused by PAC emissions to the warming caused by CO2 emissions can be estimated roughly by multiplying the global warming potential (GWP) of BC with the ratio of the PAC emission rate to the CO2 emission rate, because the GWP is defined as “the time-integrated radiative forcing due to a pulse emission of a given component, relative to a pulse emission of an equal mass of CO2.”10 The global warming potential (GWP) of BC due to only the direct aerosol effect for a time period of 20 years is estimated to be around 1200.10 Given the GWP and the ratio of emission rates, we estimate that the warming caused by PAC emissions accounts for 0.86% of the warming resulting from CO2 emissions. While the radiative forcing of PAC and CO2 gives us their warming potential, an evaluation of their radiative forcing per unit of carbon input provides a measure of their relative efficiency of carbon use as a process feedstock. Dividing their radiative forcing by their carbon input (i.e., the carbon in the coal consumed vs the carbon in the PAC injected), we estimate that the radiative forcing per unit of carbon input for PAC injection is about 8.1 times the forcing per unit of carbon input for coal burning.

■

■

UNCERTAINTIES AND LIMITATIONS There are many sources of uncertainty in estimating the radiative effect of future PAC emissions, including the emissions themselves and the optical properties of PAC. Beginning with the emissions, one uncertainty is associated with the collection efficiency of PAC by ESP. As mentioned in the section Materials and Methods, the collection efficiency of PAC is unknown and needs to be identified by laboratory studies in the future. In the absence of such essential data, a central objective of this manuscript is to assess the potential scale of such emissions. Toward this end, we have assumed a potential minimum collection efficiency of 29.5%, which would give us the maximum PAC emission rate into the atmosphere. Another uncertainty results from the projection of usage of ESP and PAC injection technology in coal-fired power plants in year 2030. We assume that all plants would use ESP to collect PM and PAC injection technology to remove mercury effectively, based on the dominance of ESP in PM control and the full development and promising future of PAC injection technology. Again, this assumption provides an estimate of the largest potential emission of PAC, in service to the objective of assessing the potential scale of such emissions. There are, of course, other technologies available to control mercury (e.g., using current existing air pollution control devices or Hg oxidation technologies). The adopting of these technologies rather than PAC injection would decrease PAC emissions. The PAC injection rate is a source of uncertainties for PAC emissions as well. Generally, this rate depends on the coal type burned, air pollution control devices used, and level of mercury control that one expects to reach. We adopt a value of 5 lbs/ MMacf, the high end of the range of values that have been tested in more recent full-scale demonstrations, although full-

AUTHOR INFORMATION

Corresponding Author

*Phone: (+1)7347640564. Fax: (+1)7349360503. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is funded by Partnerships for Innovation in Sustainable Energy Technologies (PISET) program, Energy Institute, University of Michigan, Ann Arbor, Michigan, United States of America. The figures in this paper were created with NCAR Command Language (NCL). We thank Zbigniew Klimont for providing us with the coal consumption data in power plants for the year 2030.

■

REFERENCES

(1) Pirrone, N.; Cinnirella, S.; Feng, X.; Finkelman, R. B.; Friedli, H. R.; Leaner, J.; Mason, R.; Mukherjee, A. B.; Stracher, G. B.; Streets, D. G. Global mercury emissions to the atmosphere from anthropogenic and natural sources. Atmos. Chem. Phys. 2010, 10, 5951−5964. (2) Ministry of Environmental Protection of the People’s Republic of China, Emission Standard of Air Pollutants for Thermal Power Plants, GB 13223−2011. http://www.zhb.gov.cn/gkml/hbb/qt/201109/ t20110921_217526.htm. (3) Sjostrom, S.; Durham, M.; Bustard, C. J.; Martin, C. Activated carbon injection for mercury control: Overview. Fuel 2010, 89, 1320− 1322. (4) Zhou, W.; Eggenspieler, G.; Rokanuzzaman, A.; Lissianski, V.; Moyeda, D. Prediction of Activated Carbon Injection Performance for Mercury Capture in a Full-Scale Coal-Fired Boiler. Ind. Eng. Chem. Res. 2010, 49, 3603−3610. 10522

dx.doi.org/10.1021/es502382h | Environ. Sci. Technol. 2014, 48, 10519−10523

Environmental Science & Technology

Article

(5) Krishnakumar, B.; Niksa, S.; Sloss, L.; Jozewicz, W.; Futsaeter, G. Process Optimization Guidance (POG and iPOG) for Mercury Emissions Control. Energy Fuels 2012, 26, 4624−4634. (6) Espinola, A.; Miguel, P. M.; Salles, M. R.; Pinto, A. R. Electrical properties of carbonsresistance of powder materials. Carbon 1986, 24, 337−341. (7) Calvert, S.; Englund, H. M. Handbook of Air Pollution Technology; Wiley-Interscience: Hoboken, NJ, 1984. (8) Davis, W. T.; Association, A. W. M. Air Pollution Engineering Manual; Wiley-Interscience: Hoboken, NJ, 2000. (9) 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; DOE-U.S. EPA-Epri-Awma Power Plant Air Pollution Control “Mega” Sympsoium, Baltimore, MD, 2008. (10) IPCC: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P. M., Eds.; Cambridge University Press: Cambridge, United Kingdom, 2013; p 1535. (11) Clack, H. L. Estimates of increased black carbon emissions from electrostatic precipitators during powdered activated carbon injection for mercury emissions control. Environ. Sci. Technol. 2012, 46, 7327− 7333. (12) World Energy Outlook 2011; International Energy Agency: Paris, November 2011. (13) Beychok, M. R. Fundamentals of Stack Gas Dispersion, 4th ed.; BeyChok, M. R., Ed.; 2005. (14) Prabhu, V.; Kim, T.; Khakpour, Y.; Serre, S. D.; Clack, H. L. Evidence of powdered activated carbon preferential collection and enrichment on electrostatic precipitator discharge electrodes during sorbent injection for mercury emissions control. Fuel Process. Technol. 2012, 93, 8−12. (15) Technologies for Control and Measurement of Mercury Emissions from Coal-Fired Power Plants in the United States: A 2010 Status Report; Northeast States for Coordinated Air Use Management: Boston, MA, 2010. (16) Ito, A.; Penner, J. E. Historical emissions of carbonaceous aerosols from biomass and fossil fuel burning for the period 1870− 2000. Global Biogeochem. Cycles 2005, 19, GB2028. (17) Liu, X. H.; Penner, J. E.; Herzog, M. Global modeling of aerosol dynamics: Model description, evaluation, and interactions between sulfate and nonsulfate aerosols. J. Geophys. Res. 2005, 110. (18) Lin, G.; Penner, J. E.; Sillman, S.; Taraborrelli, D.; Lelieveld, J. Global modeling of SOA formation from dicarbonyls, epoxides, organic nitrates and peroxides. Atmos. Chem. Phys. 2012, 12, 4743− 4774. (19) Lin, G.; Sillman, S.; Penner, J. E.; Ito, A. Global modeling of SOA: the use of different mechanisms for aqueous phase formation. Atmos. Chem. Phys. 2014, 14, 5451−5475. (20) Coy, L.; Swinbank, R. Characteristics of stratospheric winds and temperatures produced by data assimilation. J. Geophys. Res. 1997, 102, 25763−25781. (21) van Vuuren, D. P.; Lucas, P. L.; Hilderink, H. Downscaling drivers of global environmental change: Enabling use of global SRES scenarios at the national and grid levels. Global Environ. Change 2007, 17, 114−130. (22) Collins, W. D.; Rasch, P. J.; Boville, B. A.; Hack, J. J.; McCaa, J. R.; Williamson, D. L.; Briegleb, B. P.; Bitz, C. M.; Lin, S. J.; Zhang, M. H. The formulation and atmospheric simulation of the Community Atmosphere Model version 3 (CAM3). J. Climate 2006, 19, 2144− 2161. (23) Penner, J. E.; Xu, L.; Wang, M. Satellite methods underestimate indirect climate forcing by aerosols. Proc. Natl. Acad. Sci. U.S.A. 2011, 108, 13404−13408. (24) Wang, M.; Penner, J. E. Aerosol indirect forcing in a global model with particle nucleation. Atmos. Chem. Phys. 2009, 9, 239−260.

(25) Lin, G.; Penner, J. E.; Flanner, M. G.; Sillman, S.; Xu, L.; Zhou, C. Radiative forcing of organic aerosol in the atmosphere and on snow: Effects of SOA and brown carbon. J. Geophys. Res.. 2014, 119 DOI: 10.1002/2013JD021186. (26) Bond, T. C.; Bergstrom, R. W. Light Absorption by Carbonaceous Particles: An Investigative Review. Aerosol Sci. Technol. 2006, 40, 27−67. (27) Fuel and Carbon Dioxide Emissions Savings Calculation Methodology for Combined Heat and Power Systems; Combined Heat and Power Partnership, U.S. Environmental Protection Agency: Washington, DC, August 2012. (28) Lu, Z.; Zhang, Q.; Streets, D. G. Sulfur dioxide and primary carbonaceous aerosol emissions in China and India, 1996−2010. Atmos. Chem. Phys. 2011, 11, 9839−9864.

10523

dx.doi.org/10.1021/es502382h | Environ. Sci. Technol. 2014, 48, 10519−10523