FEATURE ARTICLE
April 1, 1998 / Volume 32, Issue 7 / pp. 176 A-179 A
Copyright © 1998 American Chemical Society

Uncertainties about the roles of different natural and synthetic sources are fueling the debate on how to regulate emissions.

CAROLA HANISCH

     Government measures limiting mercury emissions to the environment may follow, but the benefits could prove difficult to assess because of remaining scientific questions about mercury. Knowledge of where mercury settles in the environment is incomplete; its source attribution, the deposited mercury's origin, has proven difficult to quantify. Moreover, not all emissions are produced by human activity, and lack of reliable data about the speciation of mercury in source emissions further contributes to assessment difficulties.

     These uncertainties have resulted in controversy about how to control mercury emissions. Although EPA accepts a plausible relationship between emissions from industrial sources and deposition, the need continues for more quantitative certainty about amounts of mercury that are locally deposited rather than globally dispersed (1). "When local contributions are high, then the state of Florida might have to do something. If they are minor, that means that it is out of our control," said Thomas Atkeson, mercury coordinator with the Florida Department of Environmental Protection. Commenting generally, Michael Oppenheimer, chief scientist at the Environmental Defense Fund said, "Of course, there is quantitative uncertainty, but it makes sense to try to regulate the biggest sources and to limit mercury emissions." According to Leonard Levin, project manager for air toxics at the Electric Power Research Institute (EPRI) in Palo Alto, Calif., compared with total anthropogenic mercury emissions released worldwide, "The United States is only a tiny player in the global pool."

Incinerators and coal-fired boilers emit more mercury to the atmosphere than all other point sources combined. Coal-fired utility boilers are the largest point source of unregulated mercury emissions in the United States.

     Efforts are under way to fill data gaps and resolve uncertainties about where mercury deposition originates. Technology required for reliable measurement of individual mercury species is being developed and validated. Area-wide geographic networks for monitoring mercury deposition at locations across the United States, such as the nationwide Mercury Deposition Network coordinated by the Illinois State Water Survey, are being established, and preliminary data have been collected that hint at deposition patterns. Studies recently completed or underway are probing local, regional, and global deposition impacts from point sources, and computer models that simulate mercury transport and deposition are providing clues about source-receptor relationships.

A mix of atmospheric sources
For decades, large amounts of mercury were directly discharged into lakes and rivers. Local discharges led to high mercury concentrations in fish, notably in the Great Lakes. Discharges of mercury into bodies of water were subsequently prohibited by environmental laws, and the use of methylmercury in fungicides was stopped, but mercury continued to be a problem. High concentrations of mercury have been found in fish taken from remote lakes throughout the world that receive no direct pollutant discharge (1, 2). These findings implicate atmospheric deposition as a contaminant source. Airborne mercury is eventually deposited to the ground and may ultimately make its way into bodies of water, where some fraction of it is transformed to methylmercury, which can bioaccumulate in the food chain.

     Contamination of water bodies by atmospheric deposition of mercury prompted research to determine the origins of the airborne pollutant. Mercury is emitted through human activities and from natural sources such as volcanic eruptions and degassing or vaporization from the Earth's crust, of which mercury is a natural constituent (2). Land and water surfaces emit mercury released from mercury-containing ores. Mercury emitted and deposited in the past can evaporate and be re-emitted. Soils and vegetation near former industrial sources of atmospheric mercury emissions are re-emission sources (3). A recent review of mercury contamination arising from natural and anthropogenic sources indicates, despite uncertainties, that across large regions of the world, anthropogenic emissions have increased relative to natural sources since the onset of the industrial period (2).

     On the basis of reported emissions inventories, prevention and control of anthropogenic mercury emissions are issues of global dimension. According to Don Porcella, manager for ecological studies at EPRI, global anthropogenic emissions of mercury are estimated to range between 2000 and 6000 metric tons (t) per year. China alone is believed to emit about 1000 t of mercury annually (4). U.S. anthropogenic mercury emissions are estimated to be about 158 t per year (1). Electric utilities, municipal waste combustors, commercial and industrial boilers, and medical waste incinerators account for approximately 80% of the total amount (Figure 1).

Figure 1: U.S. anthropogenic mercury emissions sources

Speciation data lacking
The area (local, regional, or global) to which mercury is deposited from the atmosphere is strongly influenced by mercury's speciation in emission plumes and the extent to which species are transformed into each other, for example, once they exit the stack of an incinerator or a coal-fired power plant. Reduction of the oxidized, inorganic mercury (Hg²⁺) fraction in an emissions source can lessen localized deposition, but not global deposition. Emissions of mercury in its elemental form can remain airborne for about a year and can be transported over thousands of miles before being oxidized and finally deposited.

     For most emissions sources, only total emitted amounts of mercury are reported. Total mercury, however, is not generally correlated to regional or local deposition. Only Hg²⁺ readily deposits locally. The fraction of Hg²⁺ that occurs in mercury emitted from sources such as coal-fired power plants is not well established. According to Levin, emissions from utilities are well characterized, at least in terms of total mercury, but "measurements of mercury ionic species are not yet reliable." EPA relayed similar concerns about measurement methods, noting that "development and validation of a stack-test protocol for speciated mercury emissions are needed" (1).

     Dennis Laudal of the University of North Dakota's Energy and Environmental Research Center and colleagues recently compared different speciation measurement methods for use in coal combustion systems (5). They concluded that the Ontario Hydro mercury test method provided the best results. The technique involves the separation of mercury species in flue gases by adsorption in sodium chloride and sodium permanganate solutions. Laudal believes the procedure can serve as a reliable reference method but concedes that "wet chemistry methods such as the Ontario Hydro method are expensive and do not give real-time results. Continuous emission monitors for mercury would be more desirable."

     Results from a speciation study that used available mercury-monitoring methods suggest that mercury emissions in flue gases are divided roughly equally between elemental mercury and Hg²⁺ (6). However, Eric Prestbo, an atmospheric scientist at Frontier Geosciences in Seattle, Wash., recently reported preliminary results that are quite different (7). In direct measurements performed at a coal-fired power plant, Prestbo observed that mercury in the exhaust plume was dominated by its elemental form, which is not likely to be locally deposited from the atmosphere. Stressing the need for further research, Prestbo cautioned, "This experiment has only been conducted at a single coal-fired power plant and thus should not be extrapolated to other coal power plants or the industry as a whole." Colleague Nicholas Bloom added, "If it turns out that a lot of the presumed Hg²⁺ is in fact elemental mercury, then mercury deposition is not locally and regionally an issue; it is part of the global problem. It does not matter then whether an emission source is in Ohio or China."

     To further characterize the nature of mercury deposition and its connection to emissions sources, the transport of mercury once it exits the stack is being studied. There is controversy about whether mercury emissions from sources such as power plants have a local impact. Few data are available about mercury concentrations in the vicinity of emissions point sources. Christian Seigneur, a scientist with Atmospheric and Environmental Research, Inc., in San Ramon, Calif., thinks that mercury is more a global or regional problem than one of local concern. Seigneur recently reported that computer models have shown that the majority of power plant emissions are transported over distances exceeding 100 kilometers (km) (8).

     Some field measurements appear to contradict this conclusion. Recently, Steve Lindberg, an atmospheric scientist at the Oak Ridge National Laboratory in Tennessee, and Wilmer Stratton, a researcher with Earlham College in Richmond, Ind., developed a method to measure trace amounts of Hg²⁺ in ambient air. The species normally represents only about 3% of total gaseous mercury but is expected to account for a major portion of mercury dry deposition. On the basis of measurements of atmospheric Hg²⁺ near the ground in close vicinity to power plants, Lindberg commented, "I am convinced that cutting emissions of a particular anthropogenic source of Hg²⁺ can result in some local reduction of deposition" (9).

Global versus local sources
The effect of local or regional measures to limit mercury emissions could be minor if deposition is dominated by global atmospheric source contributions. "Establishing source-receptor relationships is important, because you want to know what happens if you try to control point sources," explains Lindberg.

     In the United States, a program is under way to analyze regional deposition. The Mercury Deposition Network, a nationwide network for mercury sampling in precipitation, currently consists of about three dozen monitoring sites and is a cooperative effort among several groups, including state agricultural experiment stations, the U.S. Geological Survey, and the U.S. Department of Agriculture. The network, a subnetwork of the National Atmospheric Deposition Program/National Trends Network, collects data to compare regions and looks for trends in deposition and correlations between deposition and emissions (Figure 2). According to Stephen Vermette, National Atmospheric Deposition Program/National Trends Network chairman, available data are not yet sufficient to describe anything on the scale of national trends. He cautioned, "We need to see 1997 data to know whether these gradients are real."

Figure 2: Regional mercury deposition patterns suggested

     Two studies in Florida are also measuring deposition. The Florida Atmospheric Mercury Study was initiated as a five-year study in 1992 to develop field collections and laboratory analytical protocols for characterizing mercury in rainfall, atmospheric aerosols, and total gaseous mercury (10). Preliminary results obtained from data collected in 1992 and 1993 show that a relatively uniform rainfall mercury deposition occurred across South Florida, and there was no strong correlation between mercury and chemical tracers. One possible explanation is that it is driven by large-scale regional or hemispheric processes rather than by local emission and deposition processes in South Florida. Bill Landing and Jane Guentzel from the Department of Oceanography at Florida State University recently reanalyzed these data using a computer model. Unpublished results show that only about 30% of the mercury deposition originates from local urban sources.

     The second Florida study provided estimates that the impact of local mercury emission sources is larger. The South Florida Atmospheric Mercury Monitoring Study, conducted in 1995, examined the potential impacts of local emission sources on the atmospheric wet and dry deposition of mercury. Preliminary findings indicate that local emission sources and deposition of mercury in South Florida and the Everglades may be more substantial than previous estimates (11).

     In a more recent computer model analysis of this data, Jerry Keeler, with the University of Michigan Air Quality Laboratory, and colleagues produced results that show about 60% of the deposited mercury originates from local urban sources. "A factor of 2 is not much for this kind of environmental problem," said Atkeson. "We are now planning for additional measurements and modeling to narrow this difference," he added.

     Other researchers have also used computer models to analyze source-receptor relationships. Data that describe mercury emissions from all known sources in the continental United States were incorporated into a computer model used by Russ Bullock, a meteorologist with the National Oceanic and Atmospheric Administration, to calculate a nationwide deposition pattern (1). Results indicate that roughly one-third, about 47 t, of U.S. atmospheric emissions is deposited within the country. The remaining two-thirds are transported beyond U.S. borders and diffused into the global atmospheric reservoir. The model also predicted that 32 t of mercury from the global atmospheric reservoir are deposited into the United States. Regional effects can be seen in the model outcomes: Most deposition occurs in the East, where most emission sources are located. According to Bullock, "Twenty-five to thirty percent of the deposition in the United States comes from global background, the rest from sources in the United States."

     Lake sediments also provide clues about the origins of deposited mercury. Daniel Engstrom, of the St. Croix Watershed Research Station in Marine on St. Croix, Minn., and Ed Swain, of the Minnesota Pollution Control Agency, cored and dated sediments taken from four lakes in northern Minnesota (12). Results indicate that the mercury accumulation rate had increased rapidly since the beginning of the 19th century but has decreased over the past 20 years. In comparison, lake sediment cores from coastal Alaska had a different deposition profile. "There was much less of an increase and no decline," said Swain, who believes that the Alaska cores reflect a global deposition trend. According to Swain, the declining mercury accumulation in the Minnesota cores reflects regional emissions that recently began decreasing because of emissions control, taller stacks, and, in general, less consumption of mercury by industry. On the basis of a comparison of the core profiles, Swain attributes 60% of the mercury deposition in lakes to global sources (natural and anthropogenic) and 40% to regional anthropogenic sources.

     Other attempts to determine the mercury's source have focused on larger-scale deposition measurements. A monitoring program, the Nordic Network for Atmospheric Mercury, was conducted in Sweden from 1985 to 1989 (13). Air and precipitation samples were collected at different sites in Scandinavia and analyzed for mercury content. The study confirmed decreasing mercury deposition levels from south to north, explained by regional transport of mercury from Europe.

     A more complete understanding of where mercury deposition to the environment originates will continue to emerge as these studies progress. Already, however, most researchers agree with William Fitzgerald of the Department of Marine Sciences at the University of Connecticut: "It is neither regional nor global. It is always a combination. In some rural locations, the regional input might not be large versus the global input, but in industrialized areas, the regional component might be strong."

References

(1) Mercury Study Report to Congress; EPA-452/R-97-003; U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards, Office of Research and Development, U.S. Government Printing Office: Washington, DC, June 1998.

(2)  Fitzgerald, W. F.; Engstrom, D. R.; Mason, R. P.; Nater, E. A.Environ. Sci. Technol. 1998 , 32 (1), 1-7.

(3)  Kim, K-H.; Lindberg, S.; Meyers, T. Atmos. Environ. 1995,29 , 267-282.

(4)  Lidqvist, O. Presented at the Fourth International Mercury Conference, Goeteborg, Sweden, Aug. 1996.

(5)  Laudal, D.; Brown, D.; Heidt, M. Presented at the Fourth International Conference on Managing Hazardous Air Pollutants, Washington, DC, Nov. 12-14, 1997.

(6)  Galbreath, K. C.; Zygarlicke, C. J. Environ. Sci. Technol.1996, 30, 2421-2426.

(7)  Prestbo, E. Presented at the Fourth International Conference on Managing Hazardous Air Pollutants, Washington, DC, Nov. 12-14, 1997.

(8)  Constantinou, E.; Wu, X. A.; Seigneur, C. Water, Air, Soil Pollut. 1995, 80, 325-335.

(9)  Lindberg, S. E.; Stratton, W. J. Environ. Sci. Technol. 1998,32 , 49-57.

(10)  Pollman, C. T. et al. Water, Air, Soil Pollut. 1995, 80, 285-290.

(11)  Dvonch, J. T. et al. Sci. Total Environ., in press.

(12)  Engstrom, D. R.; Swain, E. B. Environ. Sci. Technol. 1997,31 , 960-967.

(13)  Iverfeldt, . Water, Air, Soil Pollut. 1991, 56, 251-265.

Carola Hanisch is a freelance writer based in Nashville, Tenn.