|
 |
THE RIGHT DEGREE
Dick Wilson is former assistant administrator for air and radiation at the Environmental Protection Agency and now a consultant with National Environmental Strategies in Washington, DC.
It's not whether mercury emissions from powerplants will be regulated. It's how.
The Environmental Protection Agency (EPA) has already developed emission limits for hazardous waste combustors, medical waste incinerators, and municipal waste combustors that will reduce emissions by an average of 93 percent. Total mercury emissions from these sources will go from about 100 tons per year to about 7 tons. An emission limit for chlor-alkali plants (which produce chlorine and caustic sodas, among other things) will soon be proposed.
In addition, U.S. industrial demand for mercury has dropped by more than 75 percent from 1988 levels as a result of manufacturer efforts to reduce mercury in batteries, federal bans on mercury additives in paint and pesticides, and voluntary actions by industry. State regulation of mercury emissions and products has also had an impact. Governor Jesse Ventura recently signed legislation that bans the sale and free distribution of mercury thermometers in Minnesota. At least 10 other states are also legislating or implementing rules to ban mercury-containing products, require recycling programs, cut mercury emissions, and impose labeling requirements on products sold within the state.
Coal-fired powerplants are the last major source of manmade mercury emissions that are not already regulated—mercury emissions from electric utilities total about 43 tons per year.
As a result, coal-fired plants and their mercury emissions have gained a lot of attention. Congress is considering legislation that would establish mercury control programs for such plants, and President Bush has supported the idea of multipollutant controls on coal-fired plants that include the reduction of mercury emissions (along with nitrogen oxide and sulfur dioxide emissions). And in a significant move with wide-ranging implications, EPA has proposed to regulate mercury emissions under the toxic emissions provisions of the Clean Air Act.
The issue of the moment thus does not seem to be whether mercury emissions from coal-fired powerplants will be regulated—that seems inevitable. The real issue is whether all those involved (EPA, industry, states, etc.) can find a sensible way to accomplish it—a regulatory regimen that achieves the environmental benefits in a way that is health-based, cost-effective, and consistent with evolving energy policy and the other air emission requirements facing the utility industry.
Mercury in the Environment
Mercury is a naturally occuring element, so it cannot be created or destroyed—the same amount of mercury has existed since the Earth was formed. Mercury can, however, be moved around—from land to air to water—by both natural and human activities. It can get into land and water as a result of the release of industrial wastewater or from the disposal of waste products (like batteries, fluorescent bulbs, and thermometers). But most mercury dispersed through the environment is a result of air emissions. Some comes from volcanoes. The human activities most responsible for mercury emissions, however, are burning materials such as batteries, burning fuels such as coal, and certain industrial processes. Mercury deposition can occur very close to the source, or the element can be transported great distances. The highest deposition rates in the United States are in the southern Great Lakes, the Ohio Valley, the Northeast, and some areas in the Southeast.
About 60 percent of the current mercury deposition in the United States comes from manmade sources, and the remaining 40 percent comes from sources located outside the country, as well as natural sources. The United States contributes only 1 percent of the global atmospheric pool of mercury, so like many other issues, lowering overall mercury loadings requires international cooperation. In that regard, the United States and Canada have a joint program to reduce mercury and other toxic emissions into the Great Lakes. And the United States is involved in an international United Nations program to control emissions of mercury and other heavy metals.
Mercury tends to build up in bodies of water either directly or through air deposition. It then can bioaccumulate in fish and animal tissue in the form of methylmercury. Bioaccumulation (e.g., predatory fish eating other fish) means that the concentration of mercury in some fish can be thousands of times greater than the concentrations in the water. And for the general population in the United States, exposure to mercury occurs primarily through eating contaminated fish.
When methylmercury gets into the body, it is almost completely absorbed into the blood and distributed to all tissues. A developing fetus is considered the most sensitive human to the adverse effects of mercury. Neurotoxic effects may include cerebral palsy, reduced neurological test scores, and delayed onset of walking and talking, as well as deficits in learning abilities.
Not surprisingly, women of childbearing age and people who frequently eat contaminated fish are the most likely to be at risk. In the United States, those at risk include subsistence fishermen and some Native American populations.
States issue fish consumption advisories for contaminated waters. Such warnings have been issued in 40 states and American Samoa for thousands of water bodies, including the Great Lakes and their connecting waters.
What Are Unhealthy Levels?
But defining actually unhealthy levels for toxic chemicals is always difficult because available data are often based on unique animal or human exposures that must be extrapolated. EPA's health protective standard for methylmercury is 0.1 micrograms per kilogram of body weight per day. This reference dose represents EPA's view of the amount of methylmercury exposure that is safe. In setting this standard, EPA relied most on data from a 1971 Iraqi poisoning incident, when more than 6,500 people ate bread inadvertently made from grain treated with a mercury fungicide.
But many felt that these data were not adequate since the doses were much higher than those associated with normal diets and that, therefore, EPA's safety factors were too stringent. As a result, Congress directed EPA to contract with the National Academy of Sciences (NAS) to analyze the science regarding mercury health effects in general and the reference dose in particular. NAS concluded that EPA's guideline for protecting the public from methylmercury was justifiable, although the Academy also recommended that EPA rely on some newer studies.
NAS concluded that more research should be done to gain a better understanding of the various risk factors for the U.S. population, including regional differences and sensitive subpopulations.
Notably, in its 1999 National Health and Nutrition Examination Survey (released in March 2000), the Centers for Disease Control found that all 5,000 women and children who took part had exposure below the reference dose.
Some Controls
Coal-fired utilities actually do capture a lot of mercury, but indirectly. Much has been learned, recently about controlling mercury emissions from coal-burning utility boilers. Unfortunately, the more that is learned, the more complicated the issue becomes. It seems clear, however, that the amount of mercury that is controlled depends on
- the type of coal used (low-sulfur coals contain less mercury than high-sulfur ones);
- whether the coal has been cleaned and the cleaning method used;
- the type of burner at the plant;
- the boiler operating conditions;
- the design and operation of any particulate collection devices; and
- the design and operation of any flue gas treatment devices.
Technologies (such as scrubbers, electrostatic precipitators, and fabric filters) currently used to control sulfur dioxide (SO2), nitrogen oxides (NOX), and particulate emissions also tend to reduce mercury, but to varying degrees. And ways to increase mercury control are being tested that would augment these controls with sorbents such as powdered activated carbon.
Still, mercury can be, and, in fact, is being controlled in coal-fired powerplants. It also seems that more mercury can be controlled more cost-effectively if programs are implemented in conjunction with those for other emissions. It also appears that the ease and cost of mercury control will vary from plant to plant and that a control program should allow for such variability by not requiring each plant to meet the same control level. Finally, research on advanced mercury control approaches should be taken into account.
Toward Regulation
This brings us to the current state of affairs—a proposal to regulate mercury emissions alone, without considering the controls being required in other programs. The Clean Air Act required EPA to study the public health effects of air toxic emissions from utilities that burn fossil fuels and to determine whether it is necessary to regulate those emissions. During the winter of 1997-98, the agency sent two reports to Congress in response to those requirements—"The Mercury Study Report" and the "Utility Air Toxics Report." These identified mercury emissions as the utility toxic emission of greatest concern.
Then, last December, EPA made a regulatory finding that it is "appropriate and necessary" to regulate mercury emissions from powerplants. This was no doubt expected. But the agency also determined that utilities had to regulate those emissions under the maximum achievable control technology (MACT) provisions of the Clean Air Act. That was the big surprise—EPA, in effect had decided in its determination how to regulate the emissions, when, in fact, the decision as to which Clean Air Act authority to use did not need to be made until an actual regulation was proposed. Moreover, instead of regulation based on health effects, EPA had opted for the most stringent and costly technology-based approach, which could go well beyond what is needed to protect human health.
The schedule provides for a proposed regulation by December 15, 2003, a final regulation by December 15, 2004, and compliance no later than December 15, 2007.
In the meantime, any new plants would be subject to a case-by-case determination of the appropriate MACT emission standard.
Mismatched MACT
Here is the problem. Unfortunately, the MACT approach in the Clean Air Act is very inflexible regarding the method used to determine the level of control and the time allowed to achieve it. The control level must be based on the best performing 12 percent of existing sources, and new facilities are required to control to the level achieved by the best controlled existing source. Compliance by existing facilities must be achieved within three years.
These constraints make it difficult to design a program that establishes a cost-effective set of controls on a mix of plants and coals as varied as exist in the electric utility industry (although EPA often tries to deal with this issue by subdividing a source category—by, for example, type of coal, boiler design, installed emission control equipment, etc.—and setting different control levels for different subcategories). The three-year compliance timetable also makes it impossible to phase in controls in conjunction with other emission control schedules.
Also, using cost-efficient market mechanisms such as trading is difficult, if not impossible, under the MACT program—an unfortunate inflexibility. The SO2 trading program, set up under the 1990 Clean Air Act Amendments, has been a great success—it allows utilities to install controls in the most cost-effective way by controlling more where it is cheapest to control and less where it is more expensive. And having a trading program has often allowed EPA to set more stringent standards, because standards can be based on what the average plant can achieve rather than on what the least able plant can achieve, since the least able plant can use trading to comply.
Many environmental groups, while comfortable with the SO2 trading program, have raised questions about the appropriateness of trading when the emission is a toxic. The concern is that some plants might not make any reductions, and the local mercury levels might not improve. These issues will have to be addressed to enable the use of trading in conjunction with a mercury control programthe point is, not to allow consideration of trading leads to inflexibility and the prospect of much higher costs.
Finding a Solution
From the standpoint of a coal-based utility, the good news is that mercury emissions have been greatly reduced through control programs put in place by other industries. The bad news is that it appears inevitable that it is the utility industry's turn to control.
The issue, though, is whether all those involved can find a sensible solution—one that requires controls that are cost-effective, provides time for controls to be installed in conjunction with other control programs, takes advantage of ongoing research on control technologies, and makes maximum use of market mechanisms such as trading to assure cost-effectiveness.
First, EPA, the Department of Energy, the utility industry, and others must determine what level of control makes sense from a cost-effectiveness standpoint. Obviously, what can be achieved depends on the coal being burned, the type of boiler, the type of controls at the plant, and so on. That means that the appropriate level may vary substantially from plant to plant. The stakeholders should then develop an averaging/trading program, since it allows the level of control not to be driven by what the worst plant can do or what the best plant can do. Rather, the program could base control levels on the average plant, with the worst plants using the trading program to comply. Environmental groups and the public will need evidence that the program will improve mercury levels everywhere. Perhaps a minimum level of control by area or plant may be necessary to assure that these improvements occur.
After deciding on the appropriate control level, stakeholders should determine the most efficient timetable for compliance. Where mercury reductions can be achieved as a result or in conjunction with other required controls, every effort should be made to coordinate those compliance schedules. Mercury levels are going down, and there is no health emergency requiring controls within a set time frame; hence, compliance schedules should be set to take maximum advantage of coordinated control efforts. President Bush's National Energy Policy, for one, promotes such an integrated approach by recommending a flexible, market-based program to reduce and cap emissions of SO2, NOX, and mercury—together—from electric power generators.
Once the appropriate level of control and the proper timing are decided, EPA and others should decide what authority to use to implement those controls. The restrictions of the MACT process may or may not limit its usefulness compared to other provisions of the Clean Air Act.
Can EPA, industry, and states devise a reasonable regulation that controls mercury emissions and is cost-effective and consistent with evolving energy policy and the other air emission requirements? Of course they can, if the agency uses common sense in setting mercury control levels and timetables, and if the industry is constructively involved in the process.
|
