Mercury emerged as a serious health threat in the late 1950s when a devastating nervous system disease left thousands of people living near the Japanese city Minamata dead or severely disabled. Scientists traced the cause of the disease to consumption of seafood contaminated with methylmercury. Unlike many pollutants, mercury is not just a local problem. Once it is released, mercury persists in the environment and is capable of traveling thousands of miles. Ultimately, when mercury enters waterways it changes to toxic methylmercury, which becomes more concentrated as it goes up the food chain. Human exposure occurs largely by eating contaminated fish. Because of its long-range mobility and persistence, however, addressing only local releases cannot prevent mercury from reaching local environments. A global response is necessary to reduce the threat posed by mercury.
In February 2009, the United States and 140 other nations agreed to begin negotiations on a legally binding instrument to reduce mercury in the global environment. After four years of intense negotiations, these nations agreed to the Minamata Convention and, in October 2013, the Convention opened for signature. To date, 97 nations have signed on to the Convention. Notably, China—considered the largest emitting nation—is a signatory, though India—the second largest—is not. That the Convention enjoys such wide participation is remarkable in and of itself, particularly since so many nations will need to take significant steps to implement the Convention’s control measures.
On November 6, 2013, the United States not only signed the Convention, but also filed its “instrument of acceptance,” thereby becoming the first party to join the Minamata Convention. The United States’ action is expected to spur ratification by other parties and increase the Convention’s potential for success. The Convention will become binding and enter into force after 50 nations become parties. Read the official text of the Convention.
Once in force, the Convention will require parties to implement domestic measures to (1) inventory and control mercury air emissions, including from coal-fired power plants and boilers, certain nonferrous metals production operations, waste incineration and cement production, and releases to land and water. See Articles 8 & 9; (2) phase out certain mercury-containing products, such as batteries, switches, lights, cosmetics, pesticides and measuring devices, by 2020 and create initiatives to reduce the use of mercury in dental amalgam. See Article 4; (3) phase out mercury from manufacturing processes used to produce acetaldehyde (a catalyst) and chlor-alkali (chlorine and caustic soda) by 2018 and 2025, respectively, and take measures to reduce mercury emissions from manufacturing of polyurethane, vinyl chloride monomer production, and sodium or potassium methylate or ethylate. See Article 5; and (4) ban the creation of new primary mercury mines and phase out existing mercury mining within fifteen years. See Article 3.
The Convention also requires certain steps toward reduction or elimination where feasible of the use and release of mercury from artisanal and small-scale gold mining (ASGM), which the United Nations Environment Progamme (UNEP) estimates contributes to 37 percent of global man-made mercury emissions. See UNEP, Global Mercury Assessment 2013: Sources, Emissions, Releases and Environmental Transport (Global Mercury Assessment). In an acknowledgment that constraints on ASGM would be difficult to enforce, no reduction targets or dates for ASGM are included. The Convention also requires parties to develop strategies to address the supply and trade of mercury, safer storage and disposal, and identification and management of contaminated sites. The Convention includes provisions for technical assistance, capacity building, information exchange, public awareness, research, and monitoring and requires parties to report on measures taken to implement certain provisions. Periodically, the Convention will be evaluated to assess its effectiveness at meeting its objective of protecting human health and the environment from mercury and a conference of the parties will be established and will meet at regular intervals following the Convention’s entry into force.
Among the signatories to the Convention, the United States is a leader in the regulation of mercury, so much so that new legislation to implement the Convention in the United States does not appear necessary. In announcing the joining of the Convention, the State Department indicated that the United States has already taken significant steps to reduce the amount of mercury generated and released to the environment and “can implement Convention obligations under existing legislative and regulatory authority.” U.S. Dept. of State, United States Joins Minamata Convention (Nov. 6, 2013). Presumably this assertion refers to laws within the United States that address mercury specifically, including the Mercury Export Ban Act of 2008 and Mercury Containing and Rechargeable Battery Management Act of 1996, as well as general environmental and health laws, such as the Clean Air Act (CAA), Clean Water Act, Resource Conservation and Recovery Act (RCRA), and Federal Food, Drug and Cosmetic Act, which provide authority that has been or could be used to regulate sources of mercury exposure.
Efforts to Control Air Emissions
One of the biggest wins in the treaty for public health and the environment is the strong controls it imposes on air emissions of mercury. Because the US Environmental Protection Agency (EPA) has already taken substantial steps toward controlling major sources of mercury air emissions before the Convention comes into effect, the American public has much to gain from widespread adoption of the treaty, while domestic industry will face no new burdens as a result.
Coal is the single biggest industrial source of mercury emissions, comprising about 24 percent of global atmospheric emissions and accounting for over 50 percent of man-made emissions in the United States. Asia, which UNEP estimates contributes almost half of annual global anthropogenic mercury air emissions, continues to burn coal at a voracious pace. Mercury’s persistence in the environment and its susceptibility to long-range transport means that Asia’s demand for coal is a global problem. Some of these air emissions eventually land in Americans’ backyards and on their dinner plates (through seafood). Indeed, UNEP cites one study concluding that as much as 68 percent of mercury deposits in the Northeast and 91 percent of deposits in the west-central United States may originate from mercury emissions outside the country. See Mercury Global Assessment at 23. While transboundary pollutants have been the subject of regional efforts, the Minamata Convention is the first truly global agreement of its kind to attempt to tackle the threat of long-range transboundary air pollution. The great promise of the treaty, however, will only be realized if a critical mass of source countries implements the treaty.
The treaty curbs atmospheric mercury emissions via two main provisions. Article 8.4 of the Convention mandates that each party require new sources of mercury emissions to adopt “best available technology” and “best environmental practices.” These requirements must be in place “as soon as practicable” and no later than five years after the treaty enters into force. Article 8.5 addresses measures to control emissions from existing sources. Both provisions for new and existing sources cover five of the most significant source categories: coal power plants; coal-fired industrial boilers; smelting of nonferrous metals; waste incineration facilities; and cement clinker production facilities. Any party may propose an amendment to the Convention to add new source categories; however, the adoption of amendments is a cumbersome process. Parties must first exhaust efforts to obtain full consensus and only as a last resort may turn to a three-fourths majority vote to adopt a proposed amendment. Moreover, the United States has exercised its right under the treaty to require that any amendment be ratified separately before it becomes binding. This additional step would allow the United States to determine whether it has existing authority to control emissions from new source categories and whether Congress is likely to fill any gaps in the statutory regime, before new requirements become international obligations.
EPA has already exercised its authority under the CAA to regulate mercury emissions from source categories covered by the Convention, though these rules are in various stages of implementation. EPA regulation of mercury emissions began in earnest after the 1990 amendments to the CAA when Congress, frustrated by the lack of progress in controlling the most hazardous air pollutants, rewrote the statute with stricter and more specific requirements. In the amendments, Congress called upon EPA to control mercury emissions in three distinct ways. First, under section 112(b)(1) of the Act, Congress identified mercury among its initial list of over one hundred hazardous air pollutants that EPA must regulate. In general, under section 112(d) of the Act, EPA must impose emissions standards on categories of stationary sources that emit sufficiently large quantities of mercury (or any other hazardous air pollutant on the list). For the largest sources of such emissions, known as “major sources,” Congress set the strictest emission standards (maximum available control technology or MACT). Under section 112(d)(3) of the Act, a new major source must meet “the emission control that is achieved in practice by the best controlled similar source”—in other words, the source has to adopt pollution controls that are as good as the top performer in its class. Existing major sources have it a little easier, but not much. The statute requires these sources to meet, at minimum, the “average emission limitation achieved by the best performing 12 percent of the existing sources” (the “MACT floor”). 42 U.S.C. § 7412(d)(3), (A). EPA has some discretion to subdivide sources into sensible categories (for example, so that coal-fired boilers are not competing with gas-fired ones); but, in large part, the CAA itself demands that the toughest emissions standards apply to mercury emissions and does not allow cost considerations to soften those controls. EPA may determine that controls beyond the “MACT floor” are necessary but must take into account cost and other countervailing factors in setting those additional limits.
Second, Congress again targeted mercury in a separate list of seven hazardous air pollutants subject to additional controls. In section 112(c)(6) of the Act, Congress required EPA to sweep into its regulatory reach sources emitting at least 90 percent of total mercury emissions. That is, if applying MACT to the biggest sources still leaves more than 10 percent of mercury emissions unregulated, EPA must recast its net to cover additional categories of smaller sources (known as area sources) until 90 percent of emissions are covered. The area source categories identified by EPA under section 112(c)(6) must then apply the strict MACT for mercury emissions. Gold mines came to be regulated in this manner, though they would otherwise not be subject to MACT as a major source category.
Finally, in section 129 of the Act, Congress directed EPA to establish MACT for a list of nine pollutants, again including mercury, emitted by solid waste incineration units. In contrast to section 112(d), which only imposes the strictest controls on big (major) sources, and section 112(c)(6), which may require some smaller (area) sources to attain MACT, this provision requires all incinerators to adopt emissions controls comparable to the best in their class.
Air emissions of mercury in the United States declined under this legal regime from approximately 264 tons in 1990 to 105 tons in 2005. Some of the largest mercury emitters, including municipal waste combustors and medical waste incinerators, have seen their emissions drop more than 95 percent from the 1990 baseline. Though not without some significant business impacts (in the medical waste sector, for example, thousands of hospitals that used to operate their own small incinerators shifted to a few specialized incinerators), imposing MACT curbed mercury pollution. Moreover, the CAA, in section 112(d)(6), provides that EPA must review those standards at least every eight years and revise them as necessary to take into account developments in practices, processes, and control technologies. So long as methods to reduce mercury emissions advance, EPA’s standards under section 112 of the Act will force deeper cuts in pollution over time. EPA also has authority, under section 112(a) of the Act, to redefine which sources are “major” and therefore subject to MACT, where warranted by the characteristics of a particular pollutant. Thus, EPA can expand its reach to smaller sources as needed to respond to the risks posed by mercury emissions.
EPA’s regulation of mercury emissions under the CAA remains ongoing. In 2013, EPA finalized a rule (generally known as the Mercury Air Toxics Standards or MATS rule) imposing mercury emissions limits on power plants, the last very large, unregulated category of mercury emissions. While the history of the rule is over two decades long and fraught with administrative reversals and an adverse court ruling, the MATS rule was upheld in the D.C. Circuit. See White Stallion Energy Center v. EPA, No. 12-1100 (D.C. Cir. Apr. 15, 2014). EPA contends the rule will control 90 percent of mercury emissions from coal-fired power plants.
The existing regulation of mercury air emissions under the CAA, in large part, meets and exceeds obligations under Article 8 of the Minamata Convention. Once the treaty enters force, the United States commits to require its “new” sources to use “best available techniques” and “best environmental practices” to control mercury emissions. The only sources that are “new” are those built or substantially modified one year or more after the treaty enters into force. Given that fifty countries must join the Convention before it enters into force, there is a significant window before sources will be considered “new.” “Available” is defined within the Convention to allow for consideration of costs and benefits, while the phrase “best environmental practices” encompasses the “most appropriate” combination of control measures and strategies.
Section 112 of the CAA, by requiring new sources to attain emissions controls as strict as the best controlled similar source, without consideration of costs or other factors that could make a measure less than the “most appropriate,” surpasses the treaty obligation, at least with respect to major sources that must comply with MACT. Smaller emission sources are not subject to MACT. Under Article 8.2 of the Convention, however, a party need not cover every single source in a category, so long as 75 percent of the emissions from each source category are ultimately covered by a satisfactory measure. For four of the five source categories (power plants, boilers, incinerators, and cement production plants) addressed by the Minamata Convention, at least 75 percent of the category’s mercury emissions are (or soon will be) subject to “best available techniques” and “best environmental practices” under United States law within the meaning of the treaty.
The one exception is the nonferrous smelting category. While this category is a significant polluter in other regions, there are few such sources in the United States and their mercury emissions are small, totaling well below one ton in EPA’s 2011 National Emissions Inventory. By the time the treaty becomes effective, the United States may very well have no “new” smelting operations falling within this category. Further, although EPA has not required mercury emissions standards for most nonferrous smelting operations, this category (which actually comprises more than one stationary source category under domestic regulation) is subject to standards for other hazardous air pollutants under section 112(c)(6) and other related CAA authorities. It is feasible that EPA could determine that existing controls targeting other hazardous air pollutants at these operations constitute “best available techniques” and “best environmental practices” under the Convention. In any event, EPA has authority under the CAA to revisit existing standards to address the relatively small mercury emissions from these sources directly.
Most certainly, EPA regulation of existing sources under the CAA exceeds any international commitment under the Convention. The Convention’s provisions addressing existing sources are much looser than the restrictions on new sources, allowing Parties to take into account “national circumstances,” economic and technical feasibility, and affordability in selecting from one of five measures to control mercury from existing sources. Parties’ options include adopting an overall quantified emissions goal (e.g., cap mercury emissions at 100 tons per year); emissions limits at the source level (e.g., limit a source to 50 pounds of mercury emitted per million pounds of product output); use of best available techniques and best environmental practices; and multipollutants strategies (e.g., set limits on particulate matter emissions with the co-benefit of reducing mercury). See Article 8.5(a)-(d). The fifth potential measure, Article 8.5(e), is a catchall, including any “alternative measures to reduce emissions from relevant sources.” Parties must implement the measure of its choice as soon as practicable, but no longer than ten years from entry into force of the treaty. Because the Convention affords parties such a broad degree of discretion in how and to what extent to control emissions from existing operations, no further legislative or regulatory action will be required domestically to comply with treaty obligations.
The lax treatment of existing sources waters down the treaty’s overall effectiveness, as mercury emissions from grandfathered sources may persist for decades, but also likely makes widespread ratification of the treaty more feasible by giving parties the policy space to select measures that balance economic development goals and limited capacity. The Minamata Convention strikes a compromise between the competing interests of effectiveness and clear compliance targets, on the one hand, and policy discretion and increased participation, on the other, in its mix of hard obligations and softer, more aspirational provisions.
Efforts to Control Releases to Water and Land
The Convention’s controls on mercury releases to water and land fall within the latter, largely aspirational category. Though air emissions are the biggest contributor to global mercury deposits, UNEP estimates that at least 1,000 tons of mercury are released into water as a result of human activity globally, with several hundred tons attributable to point sources. In addition, as air pollution controls are more widely implemented and capture mercury before it is released at the smokestack, this leads to greater accumulation of mercury-contaminated wastes that must be responsibly managed. Article 9 of the Convention addresses this aspect of the mercury problem by requiring parties to adopt measures to control releases to water or land. Like the provisions addressing existing sources of air pollution, however, parties maintain substantial discretion in how and to what extent to control such releases. An “alternative measure to reduce releases from relevant sources” suffices to meet treaty obligations.
EPA regulatory activities under the Clean Water Act and RCRA are measures to control releases within the meaning of Article 9 of the Convention. Under the Clean Water Act, discharges of pollutants from a point source are prohibited without a permit. EPA promulgates national technology-based effluent limitation guidelines for certain categories of sources and has issued limitations on mercury in effluent for at least fifteen categories of sources. Sources within such a regulated category are required, at minimum, to meet the mercury effluent limits as a condition of their discharge permit. Moreover, EPA continues to develop standards to further control mercury discharges. In 2014, EPA expects to finalize an amended guideline for steam electric power plants (including coal-fired plants) that would incorporate stricter limits on mercury, and it has announced plans to develop a pretreatment standard for dental practices that discharge dental amalgam. See Notice of Final 2010 Effluent Guidelines Program Plan, 76 Fed. Reg. 66,286 (Oct. 26, 2011).
Likewise, RCRA imposes limits on the storage, transport, and disposal of wastes containing mercury. Widely generated wastes, such as spent batteries, thermostats, various gauges, switches, and light bulbs that contain mercury are designated as “universal wastes” and subject to certain standards. Other wastes containing mercury face stricter limits, including specific treatment standards intended to reduce mercury releases that apply to any land disposal.
Efforts to Reduce Mercury-containing Products
UNEP estimates that releases from consumer product waste constitute nearly 50 percent of global mercury releases to water. See Global Mercury Assessment at 24. In contrast with its provisions addressing land and water releases, the Convention offers hard targets for the phaseout of certain products that contain mercury. Again, US regulatory initiatives are ahead of the curve, and little to no additional work will need to be done to comply with treaty obligations.
Article 4.1 of the Convention requires parties to “take appropriate measures” to prohibit the manufacture, import, and export of certain “mercury-added products” listed in Annex A by 2020. Among the products listed in Annex A that must be phased out are certain types of batteries, switches, and relays including mercury thermostats; compact fluorescent lamps of 30 watts or less that exceed 5 milligrams of mercury per burner; linear fluorescent lamps including halophosphates with more than 10 milligrams mercury per lamp; cosmetics including skin lightening soaps and creams, pesticides, biocides, and topical antiseptics; and nonelectronic measuring devices including barometers, thermometers, and blood-pressure measuring devices.
Importantly, several products are excluded from Annex A and are not subject to the 2020 prohibition, including certain button-cell zinc batteries, cosmetics containing less than 1 part per million (ppm) mercury content, mascara and other eye-area cosmetics where mercury is used as a preservative for which a substitute cannot be found, certain switches and relays and fluorescent lamps for which no mercury-free alternative replacement is available, and vaccines containing thimerosal (a mercury compound) as a preservative. Products essential for military use and religious or traditional ceremonies also are excluded from the Convention. Under Article 4.3, dental amalgam—a mercury alloy used in fillings—also is not subject to the 2020 prohibition, though Article 4.3 and Part II of Annex A require Parties to the Convention to take measures aimed to reduce the use of dental amalgam, including promoting mercury-free alternatives and setting national objectives for dental health promotion. In addition to the products excluded from the Convention’s reach, Article 6 of the Convention also provides for an up to five-year exemption from the 2020 phaseout date for products listed in Annex A.
In large part, current United States regulation of mercury-containing products mirrors the requirements under Article 4.1, Annex A of the Convention, with the United States already regulating or restricting those products that must be phased out while permitting many of those products excluded from Annex A. Under existing statutory authority, EPA and the Food and Drug Administration (FDA) already limit or restrict the use of elemental mercury or mercury compounds in many consumer products, medical devices, and cosmetics. For example, mercury in products such as barometers, light switches, flow meters, and certain automotive switches is subject to EPA regulation under the Toxic Substances Control Act (TSCA). Likewise, mercury compounds in cosmetics are subject to restrictions by the FDA under the Food Drug and Cosmetic (FD&C) Act.
As a specific example, the use of mercury compounds as cosmetic ingredients is permitted in eye area cosmetics at concentrations not exceeding 65 ppm of mercury and only if no other effective and safe preservative is available for use. All other cosmetics containing mercury are considered “adulterated” under the FD&C Act and subject to regulatory action by the FDA, unless mercury occurs in a trace amount of less than 1 ppm and its presence is unavoidable under conditions of good manufacturing practice. Pursuant to this authority, FDA regularly takes action against unlabeled, “adulterated” products, including creams for lightening the skin, fading freckles or treating acne that can contain up to 1000 times the 1 ppm level of mercury allowed.
Mercury-containing batteries also are regulated in the United States. Today, most batteries produced in the United States are mercury free, except for two types not prohibited by the Mercury Containing Battery Management Act of 1996, 42 U.S.C. §§ 14331–14336: large mercuric oxide batteries produced for military and medical equipment; and button cell batteries of the zinc air, alkaline, or silver oxide varieties, containing mercury of up to 25 milligrams. Zinc air button cells are used mainly in hearing aids; silver oxide batteries are used in watches and cameras and alkaline manganese button cell batteries are used in digital thermometers, calculators, toys, and many other products requiring a compact power source. The use and disposal of these button cell batteries remains largely unregulated, and they are not widely targeted for recycling. Like the eye-area cosmetics discussed above that are permitted by both FDA and the Minamata Convention, certain button cell batteries are excluded from phaseout under the Convention.
Likewise, removal of mercury from vaccines is not a requirement under the Minamata Convention. Since 1999, however, the use of mercury-containing preservatives in vaccines has declined in the United States, although thimerosal is still used as a preservative in low concentrations. A vaccine containing 0.01 percent thimerosal contains approximately 25 micrograms of mercury per 0.5 mL dose. In the United States, thimerosal has been removed from or reduced to trace amounts (1 microgram or less mercury per dose) in all vaccines routinely recommended for children 6 or younger, with the exception of inactivated influenza vaccine. The FDA is continuing efforts to remove thimerosal from all existing vaccines for older children and adults.
Similarly, EPA is working with the dental industry to voluntarily reduce mercury releases from dental amalgam. According to data submitted to the eight states in the Interstate Mercury Education and Reduction Clearinghouse (IMERC), mercury releases from dental amalgam have been reduced by nearly half since 2001. Yet, dental amalgam continues to remain the second-largest category of mercury-containing products reported to IMERC (behind switches and relays). View IMERC reports.
Although the United States already limits or prohibits many, if not most, of the mercury-containing products targeted for phaseout under Article 4.1, Annex A of the Convention, some products remain unregulated by the federal government, including some fluorescent bulbs. Though compact fluorescent lamps (CFLs) contain small amounts of mercury, federal energy efficiency standards promote the use of CFLs in place of low-efficiency incandescent bulbs. The Minamata Convention, however, is unlikely to alter federal policy to promote use of CFLs. Only CFLs containing 5 milligrams or more of mercury per burner will need to be phased out by 2020; but EPA reports that on average most CFLs sold domestically contain about 4 milligrams of mercury.
In addition to the mercury-containing products that will be limited or phased out under Article 4, Annex A of the Minamata Convention, the Parties to the Convention must review Annex A within five years of the Convention’s entry into force and consider potential amendments to the Annex. Thus, additional mercury-containing products could be added to Annex A of the Convention and required to be phased out by the Parties.
Other Efforts to Control Mercury
Two of the more restrictive provisions of the Convention, addressing primary mining and manufacturing processes, will have little to no regulatory impact in the United States. Though Article 5, Annex B of the Convention includes hard targets for phaseout of mercury use in certain manufacturing processes, such as chlor-alkali production, the large majority of such manufacturing operations in the United States have shut down or already switched to nonmercury processes. The few remaining mercury cell chlor-alkali producers also are expected to convert due to regulatory pressures and cost-effective nonmercury alternative processes. Similarly, primary mercury mining, prohibited by Article 3 of the Convention, has not occurred in the United States since the last such mine closed in 1990.
Likewise, Article 12’s soft language requiring parties to endeavor to develop appropriate strategies for identifying and assessing contaminated sites is not specific enough to impose any new obligation on the United States. Pursuant to Article 12.3, the Conference of the Parties, in which the United States will participate, will adopt guidance for managing such sites, including defining and characterizing them.
Future Implications for the Convention
The United States has much to gain from effective implementation of the Convention. It has already taken substantial steps to control its own mercury sources; further reducing mercury exposure to Americans and their environment will depend on the actions of other countries.
As a party to the Convention, the United States will have access to a number of mechanisms to advance the treaty objectives. International environmental treaties commonly establish treaty bodies and secretariats, institutions that house scientific and technical expertise and help promote compliance with treaty obligations. The Minamata Convention is no exception. The Minamata secretariat will gather and make available information that assists in treaty implementation, such as information about the risks and benefits of mercury-added products. The Conference of Parties will establish a financial mechanism to support developing countries to implement their obligations under the treaty and a subsidiary Implementation and Compliance Committee, which will review national reports and issue recommendations to address compliance gaps. Through its early participation in the treaty regime, the United States will have an opportunity to shape these various mechanisms and ensure their effectiveness.
It is also worth noting how the Convention differs from other environmental treaty regimes. A number of multilateral environmental agreements rely upon mechanisms by which treaty obligations can be ratcheted up as scientific understanding of a problem improves. The Montreal Protocol has been effective, for example, because the parties have been able to readily modify the agreement to identify new ozone-depleting substances and set phaseout dates for their use. While the Minamata Convention contains mechanisms to amend the treaty text and the Annexes identifying covered sources, the procedures are consensus-driven and are therefore likely to be, at best, a gradual mechanism to alter treaty obligations. Given the narrow focus of the treaty on a single pollutant and its comprehensive sweep across major known sources of the pollutant, however, the inability to easily modify the scope of covered sources seems unlikely to be a stumbling block to the Minamata Convention’s success.