Mercury Rising

How will the EPA's rulemaking affect U.S. energy markets?
Fortnightly Magazine - January 2004

How will the EPA's rulemaking affect U.S. energy markets?

With President Bush's Clear Skies program stalled in Congress, it is increasingly unlikely that a multi-pollutant regulatory package will receive congressional approval in the near future. In addition to providing another source of frustration for the Bush administration, the delay also forces the Environmental Protection Agency (EPA) to propose regulations controlling mercury emissions.

Constrained by an April 1998 court decision forcing the establishment of mercury emissions for utilities, the EPA was required to propose regulation by Dec. 15, 2003, setting emissions limits for mercury for all coal-fired and oil-fired units. Until recently, the EPA was expected to propose regulations that set emissions limits at a level that can be achieved through the installation of Maximum Achievable Control Technology (MACT). As such, compliance under a MACT command-and-control standard likely would be considerably more difficult and expensive for some than it would be under Clear Skies, which would allow for trading of mercury emissions credits between generating units. However, on Monday Dec. 15, right at the deadline, EPA Administrator Mike Leavitt issued a proposed rule that features a cap-and-trade provision similar to that proposed in Clear Skies.

By selecting a less stringent, market-based regulatory standard, EPA is certain to cause environmental groups and various state agencies to litigate the decision. Even if EPA should revert back to the MACT standard originally proposed, lawsuits brought by industry groups could be expected. (Some had guessed that EPA might propose both solutions and receive public comment before issuing final regulations.) Under any of those scenarios, the issue is likely to be left undecided for some time, adding even more uncertainty to the regulatory future. Furthermore, even if a MACT standard were issued and upheld in court, passage of Clear Skies or another multi-pollutant regulatory package that contains mercury limits prior to the MACT implementation deadline almost would certainly replace the MACT standard.

For the industry, the impacts associated with the various regulatory outcomes may be considerably different. Our analysis suggests that the effects on the electricity and emissions markets could be significant. Under a MACT standard rather than any of the discussed cap-and-trade programs, the lack of a trading market would have forced generating units to either retrofit emissions control equipment or shut down. This may benefit the units that remain from the forced retirements through increased prices as regional markets return to equilibrium more quickly. But, near-term margins for units that choose to comply are likely to be squeezed by the increase in capital expenditures associated with retrofitting the required emissions control technology. Potentially offsetting this decline for some will be a reduction in SO2 costs if compliance with the mercury standard reduces SO2 emissions to the point that the SO2 markets collapse.

The question before owners of generating units is what the various regulations might mean to them. There will be winners and losers, but the breakdown is not entirely apparent. Given the long lead times associated with installing emissions control technologies and the importance of influencing the policy debate, generation owners need to act now to understand the risks associated with the various options being considered.


Current regulation of air emissions from power plants is based on the Clean Air Act (CAA). Generating units in states that are in the NOX State Implementation Plan (SIP) Call region must comply with NOX regulations during the ozone season (May through September). The CAA also established national, year-round regulation for SO2 emissions. Both programs as currently implemented are cap-and-trade programs that allow for trading of emissions between generators.

More recently, global concerns over climate change and greenhouse gases, as well as local concerns regarding the health hazards of mercury, have created a push for a new "multi-pollutant" regulatory package that would modify the current CAA. President Bush's Clear Skies Act is the current frontrunner. Clear Skies would establish a national cap for mercury, reduce the SO2 cap initially by approximately 50 percent relative to the current law's 2010 target, and establish two regional (East and West), year-round caps for NOX. Clear Skies would allow for trading of allowances for all three pollutants under a similar cap-and-trade regime as currently exists for SO2 and NOX.

Yet, due to increasing concerns over carbon emissions and details of the proposed mercury regulations, more and more legislators are starting to think that Clear Skies may not be stringent enough. These legislators have promoted other multi-pollutant legislative packages, some of which contain CO2 regulations in addition to regulations covering the other three pollutants. As some push for more stringent controls that include carbon restrictions, others are working to loosen the standards proposed by Clear Skies. The recently released Chairman's mark put forth by Senate Environment and Public Works Chairman James Inhofe, R-Okla., and Clear Air Subcommittee Chairman George Voinovich, R-Ohio, raises the proposed cap for mercury and changes restrictions on new units.

Given this debate and the ongoing dispute of the energy bill in general, some experts believe it is unlikely that Clear Skies or any other multi-pollutant bill will pass this year. If this prediction turns out to be true, then 2005 would be the earliest Congress would pass a multi-pollutant package.

Mercury MACT

A 1998 congressional study found evidence that mercury was a hazardous air pollutant (HAP) that should be studied further by the EPA. In April of the same year, a suit filed by the Natural Resources Defense Council (and others) to force the EPA to regulate utility mercury emissions was settled. Under terms of the settlement, the EPA is required to propose regulation that controls mercury by Dec. 15, 2003. Final standards must be issued a year later, with implementation of the regulation set for 2007.

In 2000, the EPA officially classified mercury as a HAP and said it would begin the regulatory process for developing a mercury standard under Section 112 of the CAA. Under Section 112, HAPs must be regulated through implementation of MACT. Until this month, the EPA was expected to follow these regulatory guidelines and propose a MACT standard to control mercury. EPA's proposed rulemaking would create more flexibility for generating units by establishing a cap-and-trade provision similar to that proposed in Clear Skies. In choosing this direction, the EPA would rescind its Section 112 determination and regulate mercury under Section 111 of the CAA (see Figure 1).

Even though the EPA did not include a MACT standard in its proposed rulemaking, details of what the mercury MACT standard would have been were unclear. Compliance would be considerably more costly for industry because of the limits required under the MACT distinction, and because generators would not be allowed to use emission cap-and-trade programs. Instead, each unit that chooses to remain in operation must find a way to reduce mercury emissions. Furthermore, the restrictions on new units may differ considerably.

A feasible MACT level has not been easy to determine due to the inconsistency of measurement data studied by the EPA ,as well as the variability in both the mercury content and the difficulty of removing mercury from various coal types. For example, mercury in bituminous coals is considerably easier to remove than mercury in sub-bituminous and lignite coals. In addition, different combustion processes have been found to lead to different mercury removal levels. To accommodate these differences and uncertainties, the EPA has discussed the possibility of developing sub-categories of mercury limits by process and fuel type for a MACT standard.

Possible Impact

The EPA's regulatory process and the ultimate outcome of the Clear Skies bill remain uncertain. Furthermore, the final level that would be required under a mercury MACT standard, as well as the difficulty of complying with that standard, remains unknown. Nevertheless, we can postulate several impacts on U.S. energy markets of a MACT standard versus any of the discussed cap-and-trade programs. These impacts can be divided into two categories: primary impacts to coal-fired and oil-fired generating units, and secondary market impacts.

Coal-fired and oil-fired generating units that will be required to meet the specific mercury emissions level required by the statute will be primarily affected. While the final standard of any MACT regulation is uncertain, many generators would find it necessary to install additional emissions control equipment. Some EPA studies suggest that the emission level achievable with the installation of Powdered Activated Carbon Injection (PACT) technology may be used in setting the final limit.

While final emission levels and different sub-categorization schemes may drastically alter the ability of individual generators to comply with the standards, installation of PACT technology alone may not be sufficient for many generators to remain in compliance. Many may have to install PACT in combination with other emission control technologies (upgraded electrostatic precipitators, fabric filters, scrubbers, etc.) to reach the discussed limits, as demonstrated in Figure 2.

For those units that already have installed these technologies to control SO2, NOX, and particulate emissions, the marginal cost of complying with the mercury MACT standard will be reduced. For others that would have needed to install the additional control equipment, the marginal cost of compliance obviously will be greater. Initial estimates of PACT installation suggest it will cost roughly $15 to $35 per kilowatt, or about $15 million to $35 million for a 1,000-MW coal plant. Addition of a fabric filter or upgraded electrostatic precipitator could add $25 million to $40 million for the same 1,000-MW coal plant. In addition, variable costs for most units likely will increase on the order of $0.50 to $1.00/MWh with PACT technology and more if additional control equipment is required.

Still others may find that a fuel change would have been required to meet the mercury MACT limits. For example, many coal units recently have switched to Powder River Basin (PRB) coal to comply with SO2 regulations. PRB coal is an inexpensive low sulfur coal, but it has relatively high mercury content that is difficult to remove compared with many bituminous coals. Faced with mercury MACT regulation, coal generation owners burning PRB coal (or other high mercury content coals) now may have found that a switch away from higher mercury content coals is the most cost-effective way of meeting mercury limits, as shown in Figure 3.

Secondary Impacts

In addition to the direct impact on coal-fired and oil-fired units, a MACT standard likely would have a number of indirect impacts on the industry. For example, if a significant number of coal-fired and oil-fired units choose to retrofit scrubbers, then it was probable that SO2 allowance prices would decrease as units that retrofit scrubbers lower the demand for SO2 allowances. This could have a secondary effect of increasing the competitiveness of coal units and, in some cases, displacing some gas units in the market.

Counteracting the impact of lower SO2 prices, many gas units may benefit as smaller or less efficient coal-fired and oil-fired units are forced to retire. These units would be retired based on the fact that it may not be technically feasible or economically rational to install the required technology, or because the increased cost of operation may reduce their cost competitiveness in the market and, as a result, lower the dispatch frequency of the units. This would benefit gas units by improving their position within the dispatch merit order, which would tend to increase the amount that the units run.

In addition, the retirement of some coal-fired and oil-fired units will eliminate some of the surplus capacity that exists in many overbuilt markets, forcing a quicker return to market equilibrium and the resulting higher market prices. The forced retirement of some units would likely be one of the major impacts of a MACT command-and-control standard. Under the alternative market-based regulatory approach, many of the units that would retire under a MACT standard would find it economic to purchase allowances and continue operating.

The effect of coal-fired and oil-fired units that switch to lower mercury-content fuels would further impact the dispatch and electricity market prices for gas units. The change in the fuel prices would affect the dispatch order of coal-fired and oil-fired units, which in turn also would impact the dispatch order of gas units. In addition, average energy prices would tend to rise as the marginal cost of the compliance units increases, benefitting gas units by increasing the prices they receive in the market.

But all of these impacts would vary by region. Impacts in each region would be determined by the specific interaction of supply and demand, fuel mix and the cost of emissions control compliance. As demonstrated by changes in capacity factor shown in Figure 4, the impact of a mercury MACT standard on combined-cycle plants in gas-dominated Texas would be very different than that for a similar plant in Virginia (see Figure 4).

Strategic Implications

All market participants will need to carefully consider the strategic implications of a mercury MACT standard versus a cap-and-trade approach. Coal-fired and oil-fired generation owners first will have to determine if it is economically and technologically feasible to retrofit compliance technology and possibly switch to lower mercury coal, or shutdown their units. To determine the economic feasibility of retrofitting compliance technology, generation owners need to know how the mercury MACT regulation would affect market prices and the dispatch of their units.

Further, coal-fired and oil-fired generation owners would consider the impact of the mercury regulation on environmental strategies designed to minimize other pollutant costs. It is possible that litigation might bring back the MACT standard. For example, as previously discussed, many coal units that have switched to PRB coal for the low sulfur content may decide to switch back to a bituminous coal to comply with the mercury MACT regulation. This switch will obviously affect the SO2 compliance strategy of the unit, suggesting that a fresh look at an integrated environmental strategy is now needed. Gas-fired generation owners will need to revisit their market strategies once they have a clear understanding of how their units are affected by any changes in dispatch order and market prices. Many new gas plants that are facing years of noncompensatory returns in overbuilt markets may find the economic picture improving much sooner. For some, this may mean the difference between continued postponement of unfinished projects and mothballing of completed projects versus full economic operation.

Considerable regulatory uncertainty surrounds a final EPA regulation. Furthermore, if a mercury MACT standard is proposed later as an alternative, implementation is likely to be delayed beyond the current 2007 deadline, adding more uncertainty to the mix. Many generation owners may make a different compliance decision if they're forced to comply with a mercury MACT standard versus a cap-and-trade alternative.

For coal-fired and gas-fired units, this choice (act now and retrofit with the required control technology or wait for more certainty) poses a strategic dilemma. Given the long lead time required to install compliance equipment (installation can take three or four years from design to completion), generation owners would have to start funding the installation of compliance technology in the next year or two to meet the 2007 compliance deadline. But if a generation owner starts funding based on the assumption that mercury MACT will be binding and Clear Skies or another cap-and-trade program is implemented, that same generation owner runs the risk of incurring additional costs through over-compliance.

Furthermore, proposed emission levels and different sub-categorization schemes may drastically alter the ability to comply with a MACT standard for different types of coal-fired and oil-fired plants. In addition, available cost projections are only rough estimates based on a number of studies. Given the variability in removal efficiencies of different technology options and the wide variance in removal difficulty associated with different coal types, the actual costs and benefits may vary considerably from current estimates for each utility.

All of these direct and indirect impacts (including how the standards will impact the performance and cash flows of specific units) need to be well understood to permit owners of coal-fired and oil-fired units to make informed decisions. For gas generation owners, the different regulations will affect their unit's performance and cash flows. Depending on the region and type of gas unit, the impact of different regulations could be positive or negative.

For units that have a low cost of abatement, acting now may be the best strategy. If mercury MACT is proposed and is upheld, then the units will meet regulations in time for the 2007 compliance deadline. If the EPA cap-and-trade mercury alternative is ordered or if Congress approves Clear Skies, these units likely would be able to benefit from their low cost of abatement position by selling surplus allowances in the mercury market that would develop.

For others that know they will retire rather than retrofit, playing the waiting game probably makes the most sense. If the mercury MACT standard takes effect in 2007, these units can retire then without having spent additional cash in the interim. If Clear Skies (or other cap-and-trade regulation) ultimately triumphs, these units would have the ability to purchase allowances from units with a lower cost of abatement to meet mercury standards.

For those in the middle, the best strategic direction is less clear. What is clear for all, however, is that a considered study of the various components and timing of the various alternatives is necessary to determine the best course of action. This study needs to consider all of the variables at play, including regulatory developments, compliance costs, time required for retrofitting, mercury content and removal efficiencies for different coal types, impact on market prices, and integration with other pollutant compliance strategies. Those that act now will be in a better position to manage risks and help shape the ongoing debate that continues to brew.

Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.