The US Mercury Emission Inventory for the Arctic Council Action Plan
Karen Rackley and Anne Pope

PDF Version: http://www.epa.gov/ttnchie1/conference/ei13/toxics/rackley.pdf

Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency, Research Triangle
Park, NC 27711
David Mobley
Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park,
NC 27711
Stan Durkee
Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC 20460
Marilyn Engle
Office of International Affairs, U.S. Environmental Protection Agency, Washington, DC 20460

The Arctic Council, having agreed to act to reduce exposures to a number of priority pollutants
in the Arctic region, has initiated a mercury project via the Arctic Council Action Plan (ACAP).
The project is led by the Danish EPA with a Steering Group from all eight Arctic countries–
Canada, Denmark, Finland, Iceland, Norway, Russia, Sweden, and United States. The overall
project objective is to contribute to a decrease of mercury releases from Arctic countries. This
will be accomplished partly by contributing to the development of a common regional
framework for an action plan or strategy for the decrease of mercury emissions, and partly by
evaluating and selecting one or a few specific point sources for implementation of control
measures. It is felt that the decrease of mercury releases from key sources should serve as a
demonstration of existing possibilities, giving inspiration to other control measures in the region.
One of the first steps in the action plan is the development of an inventory of mercury releases to
the land, air, and water. Characterization of mercury usage and its disposition will provide the
framework for an action plan and strategy for decreasing the amount of mercury in the
environment. A detailed questionnaire was developed to collect consistent data from the
involved countries, including key information on modeling parameters (e.g., latitude/longitude,
stack parameters, chemical composition, and emissions control technology).

EPA completed the U.S. portion of the questionnaire to provide data and information to the
project. All data sources are publicly available and most are from EPA inventories, e.g.,
National Emissions Inventory (NEI) for air emissions and Toxics Release Inventory (TRI) for
solid waste disposal and water discharges. The results characterize the mobilization of mercury
in the US to the land, air, and water. The overwhelming mobilization action is land disposal
associated with gold mining. The most significant air source category is coal combustion. Other
sources of air emissions include gold mining, chlor-alkali plants, municipal waste combustors,
medical waste incinerators, and industrial boilers. There were minimal discharges to water
bodies noted from the data available.

The ACAP project should result in availability of data to enable assessment of mercury issues in
the Arctic and is expected to be a model for international data exchange on mercury and other
pollutants. The overall project is intended to identify research opportunities for engineering
demonstrations that provide scientific information on mercury control options in the Arctic and
around the world.
Introduction
Reductions of exposures to a number if priority pollutants, including mercury, have been the
focus of many organizations like the Arctic Council. Comprised of representatives from eight
countries – Canada, Denmark, Finland, Iceland, Norway, Russia, Sweden, and United States, a
mercury project has been initiated through the Arctic Council Action Plan (ACAP). It is the goal
of this project to reduce the mercury emissions from the Arctic countries by the development of a
regional action plan.

Mercury is a metal that can be found in three main forms – elemental gaseous, gaseous divalent,
and particulate divalent. It is known that mercury enters the environment after being mobilized
and released from various sources such as utility coal boilers and municipal waste
combustors/incinerators. Because mercury cycles, it can be carried over long distances away
from its initial source location. Mercury is deposited into the aquatic environments through
precipitation and generally is found in its organic form known as methylmercury.

Methylmercury has many health issues associated with it, including tremors and neurological
damage. Also, methylmercury bio-accumulates, or increases in intensity as it is carried up the
food chain. Because most of the Arctic region is dependant on aquatic environments for food,
the accumulation of mercury in larger Arctic species is quite high. The concern about mercury
levels found in species such as polar bears and fish within the Arctic region spurred urgency in
developing a plan to reduce emissions for countries that may greatly affect the region.
The objective of the ACAP mercury project is to contribute to a decrease in mercury releases
from Arctic countries. To accomplish this, a regional action plan will be developed and specific
sources will be evaluated for the implementation of control measures. It is the hope of the Arctic
Council that this project will serve as a lead to other regions in developing similar regional plans.
In order to properly identify the current emissions for each country, a regional questionnaire was
developed by the Denmark EPA. The questionnaire, in the form of an Excel workbook, was
distributed to and completed by all eight participating countries. This questionnaire serves to
characterize mercury usage and disposition within each country, which will provide the basis for
an action plan. By including data for land, air, and water mercury releases and key information
on modeling parameters (e.g., latitude/longitude, stack parameters, chemical composition, and
emissions control technology), the questionnaire provides a concise inventory of mercury
releases for the Arctic countries. The first step in developing the regional action plan is the
completion of the questionnaire by each country.

US Data Sources for Questionnaire
The questionnaire is comprised of 33 individual worksheets concentrating on release trends,
trade statistics, source categories, and point sources. The sheets are designed so that emissions
are categorized by industry type. US EPA data sources used in completing these sheets are listed
below:

· Release trends - Global Mercury Assessment Report, UNEP 1.
· Trade statistics - US Geological Survey, Mineral Yearbook2.
· Source categories - 1999 National Emissions Inventory for HAPs (NEI)3 and 2001
Toxic Release Inventory (TRI)4.
· Point sources - 1999 National Emissions Inventory for HAPs (NEI).
· TRI data can be found at the following website:
http://www.epa.gov/tri/tridata/tri01/data/index.htm
· NEI data can be found at the following website:
http://www.epa.gov/ttn/chief/net/1999inventory.html#final3haps
Toxics Release Inventory (TRI) -
The Toxics Release Inventory (TRI) is a Community Right-to-Know Inventory. Facilities
meeting certain criteria are required to report annually. These include facilities that manufacture
or process 25,000 pounds or more of listed substances or use 10,000 pounds or more of listed
substances, that are in the manufacturing sector (Standard Industrial Classification (SIC) codes
20-39), electric utility sector, mining sector, solvent recovery sector, petroleum bulk storage
sector, chemical wholesale sector and the treatment, storage and disposal sector, and that have 10
or more full-time employees. (Beginning in reporting year 2000, certain persistent, bioaccumulative
toxic chemicals are subject to lower reporting thresholds).

Under TRI, for each chemical that meets the threshold, a facility reports the quantity that is
released to the air, water, land, or managed as waste on-site or off-site (disposal, treatment,
energy recovery, recycling). There are over 640 chemicals and chemical categories on the TRI
list. The TRI facilities use the best available information to make their estimates, including
monitoring data, emission factors, mass balance and engineering calculations.

While TRI provides information on point sources, it does not address many other important
sources of toxics emissions, including mobile sources, combustion sources such as certain
incinerators, and agricultural sources. TRI is not a modeling inventory and does not contain
individual stack emissions for facilities nor stack parameters necessary for modeling. TRI is not
comprehensive for all point sources in the US.
The 1999 National Emissions Inventory (NEI) -
The 1999 National Emissions Inventory (NEI) is a comprehensive inventory covering criteria
pollutants and hazardous air pollutants (HAPs). The EPA’s Emission Factor and Inventory
Group (EFIG) in Research Triangle Park, North Carolina created the NEI. This database contains
information on stationary and mobile sources that emit hazardous air pollutants (HAPs). The
1 United Nations Environment Programme Chemicals. Global Mercury Assessment. UNEP Chemicals, Geneva,
Switzerland. December 2002.

2 United States Geological Survey. Minerals Yearbook. Mercury 1994-2001.
3 1999 National Emission Inventory for HAPs, Final version, U.S. EPA, Office of Air Quality Planning &
Standards. RTP, NC. July 21, 2003.

4 2001 Toxic Release Inventory, U.S. EPA, Office of Environmental Information. Washington, DC. June 16, 2003.
database includes estimates of annual emissions, by source, of air pollutants in each area of the
country, on an annual basis. The NEI includes emission estimates for all 50 States, the District of
Columbia, Puerto Rico, and the Virgin Islands. Included are emission estimates for individual
point sources (facilities), as well as county level estimates for nonpoint, mobile and other
sources.
Point sources in the NEI are sources for which the specific location is known; they may be either
major or area sources. Major sources are defined in the Clean Air Act (CAA) as stationary
sources that:

· Have the potential to emit 10 tons per year (tpy) or more of one HAP; or
· Have the potential to emit 25 tpy or more of any combination of HAPs.
Smaller point sources with annual emissions below these thresholds are defined as area sources.
Nonpoint sources in the NEI include area sources that are not identified as point sources because
their specific locations are not known. Nonpoint sources also include other sources such as
wildfires and prescribed burning whose emissions are estimated at the county level.
Data from the NEI are used for air dispersion modeling, regional strategy development,
regulation setting; air toxics risk assessment, and tracking trends in emissions over time. The
NEI is a modeling inventory and thus contains emissions, stack parameters, control device
information, and location data for individual stacks. The NEI does not have any reporting
thresholds for point sources.

The NEI is compiled from the following sources of data: state and local agencies, tribes, EPA
Maximum Achievable Control Technology (MACT) and EPA “residual risk” data collected
during regulatory development, industry data, TRI, mobile source data ge nerated using EPA
mobile source onroad and nonroad models, and EPA generated data for approximately 30
nonpoint source categories. NEI mercury estimates are of high quality for the most reported
sources. Utility coal boilers, municipal waste combustors, medical waste incinerators, and
hazardous waste incinerators emissions are based on source testing at all facilities.

Methodology
Crosswalk of the Questionnaire with NEI and TRI Data
Because of the large amount of data available from both the TRI and NEI databases, proper
classification was essential for use in the questionnaire. In order to properly group the data
provided by both databases into the source categories provided by the Danes, several
assumptions and crosswalks were performed.
For the TRI data crosswalk with the ACAP categories, the primary identifier for source
categories was the SIC codes. Each SIC code present in the TRI data was defined and then
grouped with one of the ACAP source categories.
For the NEI data crosswalk with the ACAP categories, several identifiers were used for each:
MACT, SIC codes, and SCC codes. (SCC codes are specific EPA codes that are assigned to
describe stationary and mobile processes having associated air emissions.) Most data were
classified by MACT code if available because MACT codes were more descriptive and provided
a more accurate grouping determination.
After the NEI and TRI were grouped into the ACAP source categories, the emissions values
were categorized based on six classifications for releases and transfers. The classifications and
corresponding inventory used are listed below:

· Atmospheric releases of Hg - NEI
· Direct releases of Hg to aquatic environments - TRI
· Direct releases of Hg to soil environments (including via diffusely lost waste) - TRI
· Hg to public/municipal waste water systems - TRI
· Hg to municipal/general waste treatment - TRI
· Hg to waste collected and treated as hazardous/medical waste - TRI
The TRI data had to be further grouped to fit into the above release and transfer classifications.
The table below shows the TRI categories and how they were classified into the ACAP release
and transfer categories.

ACAP Release/Waste Transfer Category TRI Category*
Direct releases of Hg to aquatic environments Total surface water discharge
Direct releases of Hg to soil environments
(including via diffusely lost waste)
Total underground injections
Other landfills
Total land treatment
Surface impoundment
Other disposal
Hg to municipal/public waste water systems
Publicly Owned Treatment Works (POTWs)
Waste Water Treatment
Hg to municipal/general waste treatment Transfers to other off-site locations
Transfers for disposal
Hg to waste collected and treated as
hazardous/medical waste
Resource Conservation and Recovery Act
(RCRA) subtitle C landfills
*Some of the SIC codes used in the TRI program contain some ambiguities. For example, for
SIC codes 30XX, 31XX, 34XX, 35XX and 36XX contain mercury releases from coal but also
contain mercury releases from non-coal activities.

Point Sources
Another crosswalk of the data had to be performed on the NEI for specific point source
information. The same process stated above for the source categories was used for point sources.
Because the questionnaire asked for point sources based on facility and not individual stacks, a
site latitude/longitude value was assigned to facilities with multiple stacks reported. In general,
this site latitude/longitude value was based closely on the highest emitting stack at the facility.
Only the top 10 emitting facilities for each ACAP source category were entered into the
questionnaire, except for utility coal facilities that required the top 20 facilities. For each point
source provided as a top 10 emitter, the site name, latitude/longitude, stack height (m), total
mercury air emissions (kg/yr), and mercury speciation were provided. The mercury speciation
was provided by various EPA sources, including Clear Skies percentages for all source
categories except utility coal and chlor-alkali sources. Utility coal sources were already
separated by chemical composition in the NEI inventory and chlor-alkali sources were classified
using a more recent percentage value of 95/5. In March 2004, EPA’s Emissions Factor and
Inventory Group (EFIG) updated default mercury speciation profiles.

Data Results
Mercury is released to the atmosphere, water environments, and soil environments. There are
interactions between these environments, causing fluxes in mercury concentrations within each
media. Understanding the global mercury cycle helps illustrate this interaction between media.
All mercury releases contribute to the global pool of mercury in the biosphere. Mercury has the
ability to be continuously mobilized, deposited on land and water, and then re-mobilized.
The highest reported releases of mercury in the US are to the soil environments (94%), with air
and water environments making up a small percentage of the total releases reported.
Figure 1: Mercury Releases to Individual Media (Percentage)
4%
2%
94%
Air
Water
Soil

To better understand the total releases, the ACAP mercury questionnaire examined individual
source categories for mercury releases. The highest reported releases, directly affecting the soil
release percentage, were from gold mining operations. If gold mining releases were not taken
into consideration, the percentages would reflect a more balanced distribution (62% soil, 27%
air, and 11% water). Table 2 provides the reported mercury releases by the ACAP defined
source categories for releases to air, water, and soil.

Table 2: Mercury Releases to Individual Media (metric tpy)
ACAP Category Air Water Soil
Co-production of non-ferrous metals 0.05 0 6.00
Dental amalgam fillings 0 0 0
Mercury from cement production 2.31 0 1.16
Mercury in batteries 0.01 0 0
Mercury in chlor-alkali production 5.93 0.05 0.44
Mercury in landfills/deposits 0.17 -- --
Mercury in manometers, blood pressure gauges and in education 0.78 -- --
Mercury mobilized by coal usage - large combustion plants 43.46 0.17 33.06
Mercury mobilized by coal usage - other coal uses 9.10 0.01 0.03
Mercury mobilized by extraction and use of oil, gas and biofuels 7.80 0.05 0.15
Mercury mobilized by other primary extraction/production of materials 3.41 0.33 81.11
Mercury mobilized by primary extraction and processing of copper 0.03 0 49.55
Mercury mobilized by primary extraction and processing of gold 10.45 45 2407.26
Mercury mobilized by primary extraction and processing of lead 0 0 7.78
Mercury mobilized by primary extraction and processing of other metals 1.42 0.06 33.33
Mercury outputs from incineration/combustion of hazardous/medical waste 8.63 -- --
Mercury outputs from incineration/combustion of municipal/general waste 4.62 0 26.60
Mercury outputs from other waste treatment 2.33 -- --
Mercury outputs from waste water systems 0.82 -- --
Mercury releases from recycling of other metals and materials 0.98 0 0.05
Mercury switches, relays and contacts 0 0 0
Mercury thermometers 0 0 0
Mercury use in other products and processes 3.61 0.10 11.71
Mercury used in light sources 0.97 0 0
Other 2.30 -- --
TOTAL EMISSIONS (tpy) 109.2 46 2658.2
Releases to Soil Environments

As shown in Figure 1, the highest reported amounts of mercury releases are to soil environments.
The total reported emissions to the soil are 2232.8 metric tpy. The classification of these
emissions in the ACAP questionnaire creates fewer details about the specific industries. By
looking at the TRI source categories by SIC codes, a more detailed description of the releases is
provided. Table 2 below provides a listing to the top 10 categories by SIC code for releases to
soil environments. As observed in the ACAP categories, gold mining is at the top of the list with
over 2000 metric tpy of reported mercury releases to the soil environment. Of the top 10, 4 are
related to metal mining.

Mercury is found in most metal ores. During gold mining, a cyanide solution is used to extract
gold from ores, but also has been found to extract mercury during the process. Once the mercury
has been extracted from the ore, it is found in sediments left at the mining site. This mercury is
then remobilized each year during the winter and spring high flows and gets redistributed in
water or in the soil.

Table 3: Top 10 Soil Emissions by SIC
SIC Description tpy
1041 Gold Ores 2082.4
3274 Lime 41.6
1021 Copper Ores 35.5
1044 Silver Ores 26.9
4911 Electric Services 18.1
1031 Lead and Zinc Ores 8.5
2816 Inorganic Pigments 3.5
3339 Primary Smelting and Refining of Nonferrous Metals, Except Copper and Aluminum 3.3
3331 Primary Smelting and Refining of Copper 3.3
XXXX All other SIC codes 9.6

Other mining operations also allow for mercury extraction but have stricter controls and
regulations to guard against excessive releases. In the smelting and refining processes, mercury
can be extracted with other metal impurities and be collected as a sludge, which contributes to
the releases to the soil.

Releases to the Atmosphere
The second highest reported mercury emissions are to the atmosphere. The total reported
releases to the atmosphere are 117.8 tpy. Table 4 shows the top 10 source categories of mercury
releases to the atmosphere. The highest reported emissions are from sources using utility coal
boilers (47.9 tpy). These sources, along with municipal waste combustors and hazardous waste
incinerators have traditionally been found to be the largest source emissions of mercury to the
atmosphere. Because of the implementation of maximum achievable control technology
(MACT) standards, both municipal waste combustors and hazardous waste incinerators have
been able to reduce their emissions considerably.

The second highest amount of mercury releases to the atmosphere comes from gold mines. As
stated in the section above, mercury is extracted when cyanide solutions are used to extract gold
from ores. Gold mining by-products have been shown to be the only large mining-related source
of mercury air emissions. Because of voluntary initiatives with Nevada gold mines, pollution
prevention strategies have been implemented in hopes of reducing mercury releases from these
sources.

Chlorine production ranks third for mercury emissions to the atmosphere. Mercury cells are
commonly used at chlor-alkali plants, causing mercury to be released to the atmosphere.
Recently, a MACT standard for these facilities has been promulgated to help reduce their
contribution of mercury.

More information regarding air emissions from specific industries can be found on the AP-42
website: http://www.epa.gov/ttn/chief/ap42/index.html.
Table 4: Top 10 Air Emissions by Source Category
Source Category tpy
Utility Boilers: Coal 47.9
Industrial/Commercial/ Institutional Boilers & Process Heaters 12.0
Gold Ores 11.5
Chlorine Production 6.5
Municipal Waste Combustors 5.1
Medical Waste Incinerators 2.8
Stationary Reciprocal Internal Combustion Engines 2.5
Commercial Hazardous Waste Incinerators 2.5
On-Site Hazardous Waste Incinerators 2.4
All other Source Categories 24.0
Releases to Aquatic Environments
The lowest mercury releases are reported to the water. These releases make up about 2% of the
total mercury releases in the U.S. Most of these releases occur as a result of runoff or mercury
contained in wastewater.

Table 5: Top 10 Water Emissions by SIC
SIC Description tpy
2631 Paperboard Mills 0.26
4911 Electric Services 0.19
1221 Bituminous Coal and Lignite Surface Mining 0.09
2816 Inorganic Pigments 0.07
3312 Steel Works, Blast Furnaces (Including Coke Ovens), and Rolling Mills 0.06
2812 Alkalies and Chlorine 0.05
2911 Petroleum Refining 0.05
2869 Industrial Organic Chemicals, NEC 0.02
2819 Industrial Inorganic Chemicals, NEC 0.01
XXXX All other SIC codes 0.05

Conclusions
Benefits of Participation
By participating in this project with the Artic Council, EPA has been able to achieve several
benefits. First, EPA has been able to demonstrate leadership in both policy and
research/assessment. As an agency, EPA has been able to provide guidance and assistance to
other countries in developing their own emission inventories and rule-making strategies.
Through taking a national emissions inventory, EPA has identified the facilities that contribute
the most to mercury emissions. Since the questionnaire that was compiled examined all release
media (soil, air, and water), EPA now has a complete modeling inventory to allow for
determination of how the U.S. mercury emissions affect the Artic region. Table 6 below shows
the results of the ACAP project by listing total atmospheric mercury emissions reported from
each country.

Table 6: Breakdown of Atmospheric Mercury Releases
From the Arctic Countries (metric tpy)*
Country tpy
Canada 8
Denmark 1.4
Finland 0.5
Iceland NA
Norway 0.6
Russian Federation 39
Sweden 0.6
USA 107
Total 10.5
*Note: Results taken from DRAFT Report, “Reduction of atmospheric mercury emissions from Arctic countries –
Regional mercury inventory.

Conclusion
Through EPA’s participation in the ACAP Mercury Project, the U.S. has a complete multimedia
modeling inventory, a result not previously obtained. By using the modeling inventory, the fate
and transport of mercury can be determined. The process used in this inventory can be used in
the future to develop multimedia inventories for other pollutants as well.
By examining the trends in mercury air emissions, it has been found that mercury air emissions
has decreased as a result of the implementation of MACT standards. With several MACT
standards having promulgated after this project’s completion, it is EPA’s hope that this
decreasing trend will continue.

Gold mining operations were identified as an industry sector needing further attention to
determine if controls are needed to reduce their emissions. This industry contributes the highest
releases to soil environments, and second highest to the atmosphere.
Finally, this project and EPA’s participation, can serve as a model to other countries and regions.
By developing complete regional emission inventories, protecting the whole environment can be
accomplished.