US Limits Pollution from Power Plants

US Limits Pollution from Power Plants

April 01, 2005

The US Environmental Protection Agency issued two new air emission rules in March that will require spending on costly new pollution control equipment retrofits at many existing power plants over the next five to 13 years.

The first rule — called the “clean air mercury rule”— would require reductions in mercury emissions from existing and new coal-fired power plants in a two-phased “cap-and-trade” approach starting with the first phase in 2010 and the second phase in 2018. “Cap and trade” means that power plants have a choice of reducing pollution or buying emission credits or allowances from other plant owners who have extra allowances. In addition to meeting the mercury emission caps, new coal-fired power plants that commence construction on or after January 30, 2004 will have to meet stringent “new source performance standards” for mercury emissions.

The second rule — called the “clean air interstate rule” or CAIR — would require 28 states and the District of Columbia to adopt rules that require substantial reductions in nitrogen oxide, or NOx, and sulfur dioxide, or SO2, emitted from power plants and other pollution sources. Each state has been told how much NOx and SO2 it is allowed. It must achieve the required NOx and SO2 reductions through one of two compliance methods. It can participate in an EPA-administered cap-and-trade regime that caps power plant emissions in two phases starting with an initial cap in 2009 and then a lower cap in 2015. Alternatively, it can ask EPA to approve another method for achieving the reductions from in-state sources.

The two rules are intended to work in tandem and are expected to prompt significant investments in air pollution control equipment to meet the air emission reduction requirements. Both rules are closely modeled after a bill called the “Clear Skies Act” that the Bush administration has been trying to put through Congress for the past three years. The administration suffered a significant setback in March when the Senate environment committee failed to pass the measure out of committee by a 9-9 vote. Committee chairman, James Inhofe (R-Oklahoma), said the Clear Skies Act will not be reconsidered by the committee this year. The administration had threatened to take action administratively if it could not get the legislation through Congress. The main difference between the regulatory action EPA took in March and what the administration wanted from Congress is the clean air interstate rule only applies in 28 states and the District of Columbia. Any legislation that Congress passed would have applied to the entire country and would have been better insulated from the inevitable litigation that follows any EPA rules.

The administration is bracing for protracted litigation over the legality of the mercury rule. Although less controversial, the Bush administration also anticipates that lawsuits will be filed challenging certain aspects of the clean air interstate rule. Any legal challenge to the two rules must be filed within 60 days after the final rules are published in the Federal Register.

Environmental groups complain that the clean air mercury rule should have had stricter standards and a shorter compliance timeline. They are also unhappy with the “cap-and-trade” approach since companies have a choice of reducing mercury emissions or buying allowances, meaning that “hot spots” of mercury may remain where companies choose to purchase allowances rather than invest in new pollution control.

During phase I of the mercury rule, the limit on mercury emissions from coal-fired plants is set at 38 tons annually for the entire country. EPA expects that coal-fired plants will be able to meet the phase I reductions as a byproduct of reducing NOx and SO2 emissions as required by the clean air interstate rule. Certain pollution control technologies used to reduce NOx (such as selective catalytic reduction systems) and to limit SO2 (such as flu gas desulfurization systems or scrubbers) achieve some limited reductions in mercury emissions. EPA predicts that most coal-fired power plants will not have to take additional steps to reduce mercury until the phase II mercury cap of 15 tons annually takes effect in 2018. Once the mercury rule is fully implemented, EPA predicts that mercury emissions from power plants will be reduced by almost 70% from the 1999 baseline level of 48 tons.


Mercury is widely recognized as a highly toxic, persistent pollutant that accumulates in the environment. Mercury can transform into methylmercury and build up in the food chain. Humans may consume mercury by eating contaminated fish. Coal-fired power plants are some of the more significant mercury emitters.

EPA has already imposed mercury emission limits on certain municipal solid waste combustors and medical waste incinerators. In regulating mercury from coal-fired power plants, EPA has taken a laborious legal and highly-charged political path. The agency was sued in 1992 by the Natural Resources Defense Council, or NRDC, for not including power plants in the initial list of major stationary sources that were regulated under the hazardous air pollutants or HAPs section of the Clean Air Act — section 112. The Sierra Club then sued EPA in 1994 for failing to complete a study required by the 1990 Clean Air Act to determine whether it is “appropriate and necessary” to regulate power plants under section 112. Both lawsuits were settled in 1994, and EPA agreed to complete a “Utility Air Toxics Study” by February 1998. Based on the findings of the study, EPA concluded in December 2000 that hazardous air pollutants from coal-fired and oil-fired power plants present a public health concern and that regulating such emissions is “necessary and appropriate.” As a result, coal- and oil-fired power plants were added to the Clean Air Act section 112 source category list.

Once a source category is listed under section 112, the Clean Air Act requires that a hazardous air pollutant standard be proposed within three years and a final rule issued within a year thereafter. With the clock ticking, EPA was obligated to propose a rule to regulate mercury from coal-fired plants and nickel from oil-fired plants by December 2003.

As the deadline for proposing mercury standards for coal-fired power plants rapidly approached, EPA faced a dilemma over possible options to regulate mercury without crippling the utilities and independent power producers that own coal-fired power plants. If the agency adopted the traditional “command and control” approach set out in section 112, then companies would have three years from the date of the final rule to implement stringent “maximum achievable control technology” or MACT standards. For existing power plants, MACT standards must be based on the average emissions achieved by the best performing 12% of plants in a particular category or subcategory of sources.

Imposing stringent MACT limits presented several problems. There is currently no “silver bullet” technology available to achieve substantial mercury air emission reductions from power plants. Some pollution control devices hold promise, such as activated carbon injection, but these technologies are still in the development stage and are not yet widely commercially available. (EPA believes that activated carbon injection and other mercury removal technology achieving reduction levels between 60% and 90% may be commercially available in the 2010 to 2015 time frame.) Another significant complication is that there are several different types of coal, including bituminous, sub-bituminous and lignite, and each of these types of coal has different levels of mercury content. The types of mercury within these coals also differ. For example, bituminous coal contains higher levels of mercury than sub-bituminous and lignite coals, but the mercury is much harder to remove from sub-bituminous and lignite coals.

Mercury is emitted either in a particulate form, a gaseous elemental form or divalent oxidized form. The particulate form is easiest to remove by using a baghouse or similar particulate control device. However, only small amounts of mercury are emitted as a particulate. Divalent oxidized mercury reportedly remains in the atmosphere for less than two weeks due in part to its solubility in water, and it can be deposited over 50 to 500 miles away from the source through rain, snow or dry deposition. This form of mercury can be captured by wet flue gas desulfurization systems. Elemental mercury is insoluble in water and is generally more difficult to remove. Elemental mercury can reportedly remain in the atmosphere for over a year before being oxidized.

Wet flue gas desulfurization scrubbers are capable of removing mercury emissions from the burning of bituminous coal, and capture efficiencies range from about 20% to more than 80% depending on the amount of elemental mercury in the flue gas. Other conventional technologies include using a different type of coal, coal cleaning and certain particulate control devices (baghouses). Mercury removal efficiencies using conventional technologies are typically less when burning sub-bituminous or lignite.  The chlorine content of coal also affects the form of mercury in the flue gas. Bituminous coals have higher chlorine levels and generally produce a flue gas that is higher in oxidized mercury. Most eastern coals are bituminous coals.  The western coals are generally either sub-bituminous or lignite, which have a lower chlorine content, and the mercury is typically present in the harder-to-remove elemental form.

In searching for a workable solution, EPA evaluated whether a cap-and-trade regime could be applied to a mercury program. Similar cap-and-trade programs, such as the acid rain program and the NOx budget trading program in the northeastern US, have afforded the regulated community a cost effective way to achieve required emission reductions. Under this approach, each existing power plant is given the right to emit a certain amount of mercury. Anyone who wants to emit more mercury than authorized must first purchase allowances from other plants that have them. Since section 112 of the Clean Air Act does not authorize emission trading, EPA had to find other statutory authority to support a cap-and-trade program.

In January 2004, EPA proposed two alternative approaches for reducing mercury emissions from coal-fired plants and nickel emissions from oil-fired plants based on different sections of the Clean Air Act. The first alternative involved the “command-and-control” MACT standard approach under section 112. The second alterative was a cap-and-trade approach under section 111 of the Clean Air Act. That section is much less prescriptive and provides more flexibility in setting mercury and nickel emission standards.

In order to make the second alternative work, EPA also proposed in January 2004 to rescind the agency’s December 2000 conclusion that the regulation of mercury and other hazardous air pollutants from coal- and oil-fired utilities is “necessary and appropriate” under section 112.  The agency said the regulation of mercury from coal-fired plants and nickel from oil-fired plants under section 112 air toxic provisions is no longer “necessary.”

EPA made these steps final on March 15, 2005.  The acting EPA administrator, Steve Johnson, signed a final rule reversing the agency’s December 2000 finding that the regulation of mercury from coal-fired plants is “necessary and appropriate.”  The agency said the December 2000 finding lacked foundation and more recent information demonstrates that it is not appropriate or necessary to regulate coal-fired power plants under section 112. The action paved the way for signing the clean air mercury rule on the same day that calls for a cap-and-trade regime as the mechanism to reduce mercury emissions from existing coal-fired plants. EPA is not moving under the clean air mercury rule to reduce nickel emissions from oil-fired power plants.

Nine states immediately filed a lawsuit challenging the EPA actions in the US appeals court for the District of Columbia. The states are New Jersey, California, Connecticut, Maine, Massachusetts, New Hampshire, New Mexico, New York and Vermont. The case will probably be consolidated with lawsuits being filed by other parties. A decision in the case is not expected until late 2006 or early 2007.

The mercury rule applies to coal-fired steam generating units capable of combusting more than 25 megawatts on an output basis and that sell more than 25 megawatts to the grid. These utility units typically consist of a furnace firing a boiler, which is used to produce steam that is run in turn through a steam turbine to generate electricity for sale. The mercury rule also applies to cogeneration units capable of combusting more than 25 megawatts on an output basis and that put more than a third of their capacity and more than 25 megawatts into the utility grid for sale. (A cogeneration facility is a power plant that generates two useful forms of energy from a single fuel, such as burning coal to boil water and produce steam, some of which is used as steam to heat an adjacent building and the rest of which is run through a steam turbine to generate electricity.)

EPA has adopted a 38-ton mercury emission cap for the first phase commencing in 2010 and a 15-ton cap for the second phase starting in 2018. This is the amount of mercury emissions that would be allowed each year from all coal-fired power plants nationwide. In both phase I and phase II, mercury allowances would be issued to coal-fired plants based on a unit’s share of the total heat input from existing coal units multiplied by an adjustment factor depending on the type of coal. The adjustment factors are 1.0 for bituminous, 1.25 for sub-bituminous and 3.0 for lignite coals. One allowance will correspond to one ounce of mercury.

States have the option of participating in an EPA-managed cap-and-trade program or electing to adopt their own state programs. Several environmental groups are encouraging states to adopt more stringent state rules to control mercury emissions from coal-fired plants. To date, four states — Connecticut, Massachusetts, New Jersey and Wisconsin — have adopted specific mercury emission limits that apply to older coal-fired plants in those states. For example, Connecticut moved in 2003 to require coal-fired power plants in the state to reduce mercury emissions by approximately 90% starting in 2008.  The four coal-fired power plants in Massachusetts will need to meet an 85% mercury reduction level by January 1, 2008, and a 95% reduction level starting on October 1, 2012.

EPA agreed in the final mercury rule with the regulated community’s argument that different mercury emission reduction targets should apply based on the type of coal being burned. Thus, allowances for existing plants will be weighted according to the type of coal burned. In the final rule, EPA set new source performance standards for new plants at levels that take into account the type of coal being burned. New coal-fired power plants that commence construction on or after January 30, 2004 must achieve the following output-based mercury levels: 0.0026 nanograms per joule (21 x 10-6 lb/MWh) for bituminous-fired units, 0.0055 nanograms per joule (42 x 10-6 lb/MWh) for sub-bituminous-fired units equipped with wet scrubbers, 0.0103 nanograms per joule (78 x 10-6 lb/MWh) for sub-bituminous-fired units equipped with dry scrubbers, 0.0183 nanograms per joule (145 x 10-6 lb/MWh) for lignite-fired units, 0.00017 nanograms per joule 1.4 x 10-6 lb/MWh for coal refuse-fired units, and 0.0025 nanograms per joule (20 x 10-6 lb/MWh) for integrated gasification combined cycle units. These limits are somewhat higher than originally proposed by EPA, and are based on an analysis of the “best demonstrated control technology” under section 111.

The regulated community does not need to take any immediate steps to reduce mercury emissions in anticipation of the phase I deadline, and potentially significant compliance costs will probably not be necessary until a few years before the start of the phase II portion of the rule in 2018. It is because of this long lead time that certain northeastern states and several environmental groups are lining up to challenge the mercury rule, and the fate of the rule will ultimately be decided by the US appeals court in the District of Columbia. In the meantime, it is possible that several other northeastern and mid-Atlantic states may take the initiative by adopting tougher mercury rules of their own for coal-fired power plants.

Interstate Rule

The clean air interstate rule signed in March focuses on reducing the amount of NOx and SO2 from power plants and other sources that blow across state lines where it contributes to ozone and particulate matter pollution in downwind states. NOx and SO2 are precursors of fine particulates, or PM2.5, and NOx is a precursor of ozone. NOx, SO2 and PM2.5 can travel for hundreds of miles and these pollutants have been linked to serious respiratory illnesses as well as having an impact on sensitive ecosystems and reducing visibility.

The final rule directs 28 states and the District of Columbia to issue new regulations that will require dramatic reductions in NOx and SO2 emissions by 2015 in a two-stage approach.

EPA determined that NOx and SO2 emissions from 23 states and the District of Columbia lead to unhealthy fine particulate levels in downwind PM2.5 nonattainment areas. It also identified 25 eastern states and the District of Columbia as contributing to ozone pollution in downwind areas that are not achieving the 8-hour national ambient air quality standard for ozone.  These states are identified in the map on page 6.

The clean air interstate rule adopts a NOx and SO2 emissions budget for each state. States must comply with the allocated budgets either by participating in an EPA-administered cap-and-trade program that targets reductions from power plants in two stages or by proposing other measures, including requiring emission reductions from sources in other industrial sectors. However, the tone of the final rule strongly suggests that EPA expects states to adopt the cap-and-trade program.

Assuming most states go with cap-and-trade, EPA models suggests that the clean air interstate rule would bring about a 53% reduction or 1.7-million-ton decrease in NOx emissions from 2003 levels by 2009. In 2015, the rule is expected to reduce power plant NOx emissions by two million tons, which is a 61% reduction from 2003 levels. In 2010, the clean air interstate rule will reduce SO2 emissions by a projected 4.3 million tons, a reduction of 45% from 2003 levels. By 2015, SO2 emissions will have been reduced by 5.4 million tons or 57% from 2003 levels in the 23 affected states.

Each state must adopt its own regulations to implement the clean air interstate rule, and the regulations must be approved by EPA. For states subject to findings of significant downwind contribution for PM2.5, EPA is establishing annual budgets, and for states subject to findings of significant downwind contribution affecting compliance with the 8-hour ozone standard, the rule requires ozone season — May 1 to September 30 — emission budgets. States subject to findings for both PM2.5 and 8-hour ozone are also subject to an annual NOx budget. The annual and ozone season budget caps are listed in the following chart.

The model cap-and-trade program incorporated in the clean air interstate rule uses a 25-megawatt cut-off to define affected sources similar to how affected units are defined in the so-called NOx SIP Call rule and the acid rain program. The model cap-and-trade program also includes an exemption for small cogeneration units. Cogeneration units that generate more than 25 megawatts and supply more than one-third of their capacity and more than 219,000 megawatt-hours to the utility grid for sale would be subject to the cap-and-trade requirements.

Power plants located in states that are subject to the clean air interstate rule would be required to submit acid rain program allowances at particular retirement ratios to meet their SO2 reduction obligations under the new rule. Affected power plants would be able to use pre-2010 vintage SO2 allowances on a one-to-one basis. Vintage 2010 to 2014 SO2 allowances could be used at a two-to-one ratio, and vintage 2015 SO2 allowances and beyond would be retired at a ratio of 2.86 allowances for every one ton of SO2 emissions. One effect of this SO2 allowance retirement is that the value of post-2009 SO2 allowances would be reduced in states subject to the clean air interstate rule SO2 reduction requirements. There also may be a premium placed on achieving early SO2 emission reductions since pre-2010 vintage year SO2 allowances will be more valuable vis-à-vis 2010 and beyond allowances.

The clean air interstate rule has detractors, but it is not nearly as controversial as the mercury rule. It is much more likely to survive a legal challenge, since it is modeled after the NOx SIP Call rule, which was largely upheld by a US appeals court after a protracted legal battle.

The clean air interstate rule will trigger installation of a new round of costly pollution control measures at some power plants and other industrial facilities, including selective catalytic reduction systems or selective non-catalytic reduction systems to reduce NOx emissions to meet the 2009 deadline and flue gas desulfurization systems to limit SO2 emissions starting in 2010.