Clean Air and the Politics of Coal

Air quality policy–technically complex and always contentious–has become the focus of bitter controversy. The public debate is about an Environmental Protection Agency (EPA) program called New Source Review (NSR), which regulates emissions from industrial facilities. There’s a surprising consensus about how to fix NSR: emissions caps and a trading system. But underlying the debate and preventing compromise is deep disagreement about the future of the powerful coal industry. The administration would protect coal, whereas others give precedence to public health.

The EPA has issued a series of highly complicated new rules that will dramatically weaken NSR. Of particular concern are coal-fired electric plants that have not complied with air quality standards that are 25 years old. Environmentalists cite the new rules as a vivid example of how the Bush administration is dismantling environmental protection.

The administration responds that the old rules are too costly and prevent modernization of facilities in ways that would efficiently reduce emissions. It proposes a “Clear Skies” bill that would establish a cap-and-trade system for mercury and nitrogen oxides but would relax somewhat the schedule for reductions of sulfur dioxide, which already operates under a highly successful cap-and-trade system. Clear Skies would not include emissions of carbon dioxide, the main greenhouse gas implicated in human-induced climate change.

Congress has been considering bills to address air quality, climate change, and changes to NSR. But for now it cannot agree on what to do.

It is time to reform NSR, which exemplifies command-and-control regulations that set forth in excruciating detail how firms must reduce emissions. A cap-and-trade system would replace bureaucratic guidance by setting limits on emissions from major sources and allowing facilities flexibility to install pollution controls, switch to cleaner fuels, redesign processes and products to avoid pollution, or pay other plants to do these things.

But the Bush administration’s approach would allow old coal-fired plants to keep polluting while the industry develops a technological fix to reduce emissions of greenhouse gases. Until that fix is achieved–possibly decades from now–aging plants using outdated technology would be able to continue to burn cheap, dirty coal.

If we are serious about reducing current levels of air pollution–and we should be, given the clear evidence about the effects of coal-fired electricity on human health and the environment–we can’t afford to wait. We favor requiring old plants to clean up now and then installing a flexible, multitiered cap-and-trade system. This would create market incentives for coal and utility industries to develop clean technology rather than continue hiding behind government rules.

Costly grandfathers

The Clean Air Act of 1970, passed soon after the first Earth Day, set the pattern for U.S. environmental law. The statutes adopted during the next several years required that major industrial plants install pollution control equipment to reduce emissions at the tail end of production processes–from smokestacks, tailpipes, and drainage pipes. These requirements ushered in terms such as “best available control technology” and “lowest achievable emissions rate.” Although varying in their stringency and workability, these regulations all rely on detailed industry-by-industry specifications that firms must adopt.

This approach exerts some pressure for innovation. When better control technologies come on the market, presumably they are certified and raise standards. But the system gives firms no incentive to custom-design more effective technologies for special situations, to reexamine their entire production process, or to improve beyond required levels.

When Congress wrote the Clean Air Act, it exempted old plants, believing that the most economical time to add pollution controls was when plants were expanding production or modifying processes. This makes sense for cars, which typically are replaced every few years. But it soon became clear that many grandfathered industrial facilities, especially coal-fired power plants, were contributing a large share of the nation’s air pollution and were not going to clean themselves up any time soon.

To force cleanup, Congress in 1977 required facilities to install state-of-the art technology as part of modifications that include any physical change or any change in their method of operation that increases the amount of any air pollutant. Taken literally, this was a very high standard–too high to be practical. Presumably, Congress did not mean to trigger the NSR process when a plant installed a new light bulb. So a few years later, when Democrats controlled both the presidency and Congress, EPA excluded “routine maintenance, repair, and replacement” from the definition of modification.

This opened a battle about the difference between routine maintenance and other modifications. Consultants to the electric utility industry began promoting multimillion-dollar “life extension projects” that would allow companies to “repair” coal-fired power plants without triggering the NSR process. The EPA took one company to court, asserting that changes in its plant were not routine maintenance or repair. The judge ruled in 1988 in favor of the EPA, saying that, “the statutory scheme intends to grandfather existing industries; but the provisions concerning modifications indicate that this is not to constitute a perpetual immunity from all standards.”

Even after this decision, litigation and lobbying continued as the EPA labored to rewrite voluminous guidance to its staff and to states that manage NSR under EPA oversight. Industry claimed that uncertainty about definitions and procedures kept many firms from investing in repairs and expansions. For example, the National Coal Council, an industry-funded adviser to the Department of Energy (DOE), claimed in 2000 that the utility industry was delaying projects that could expand generating capacity by 10 percent, while at the same time reducing emissions. The EPA encouraged companies to seek advice about whether proposed modifications were routine maintenance. But there is no requirement that firms report routine maintenance projects to the EPA or to state agencies. Many firms went ahead without asking for advice or permits, and “don’t ask, don’t tell” became the operative approach.

In 1993, the Clinton administration reorganized the EPA, establishing a separate Office of Enforcement and Compliance Assistance. Seeking to target its resources to get the biggest environmental gains, this office analyzed emissions and the number of permits and enforcement actions for different industries. This effort showed that many coal-fired power plants, refineries, steel mini-mills, chemical manufacturers, and pulp and paper mills had significantly expanded production without going through the NSR process. The EPA began investigating electric utilities in 1996, and three years later the Justice Department filed the first of 51 enforcement cases related to NSR, including 14 against electric utilities.

Trade associations for electric utilities and many individual companies vigorously protested the litigation. They claimed that the EPA was using a new interpretation of routine maintenance and had not given “fair notice.” According to their argument, EPA and state inspectors had been visiting their plants for years and must have seen life extension projects, but had never insisted on the need for NSR. Nevertheless, seven electric utilities and 14 refineries and other factories have settled out of court. They agreed to pay $79 million in penalties and to spend $4.6 billion cleaning up their facilities and $93 million on other environmental improvements.

In August 2003 came the first federal court decisions about the enforcement cases. FirstEnergy, the defendant, claimed that the changes made to its power plants were “the kinds of routine maintenance that every single coal-fired power plant does.” The court, however, found that the company’s interpretation of routine maintenance would be “in direct conflict with the superceding and controlling language of the Clean Air Act.” Later that month, a different federal court advised lawyers that “routine maintenance” should be interpreted industrywide, so that one plant could not be faulted if it followed the same practices as other firms. Still, the Justice Department and the EPA have not lost a single case.

Under the Bush administration, the EPA initially promised to continue pushing these types of enforcement cases. But in August 2003, the agency revised its definition of routine maintenance. The new definition exempts any change that does not exceed 20 percent of the replacement value of the entire process unit, any replacement of components of a process unit with identical or functionally equivalent components, any replacement that does not change the basic design parameters of the process, and any replacement that does not cause the unit to exceed emissions limits. This new definition will enable plants to conduct extensive modernization without installing best available control technologies. The “replacement value” of a 30-year-old power plant may be more than its original cost. By rebuilding a unit in stages over several years, each stage costing less than 20 percent of the replacement value, a plant might be completely rebuilt without incorporating state-of-the-art pollution controls.

Four months before the new rules, the National Academy of Public Administration (NAPA), a nonprofit organization chartered by Congress to advise agencies facing difficult problems of management and governance, issued a report on the NSR process. The fact that many old plants had not upgraded pollution control equipment was seen as a clear failure. The report concluded that the legislative history of the Clean Air Act, along with subsequent regulations and judicial decisions, supports a strict interpretation of routine maintenance. It calls for Congress to end grandfathering and for the EPA and the Justice Department to aggressively pursue enforcement cases. Within the next 10 years, all major sources of pollution that had not obtained an NSR permit since 1977 would have to lower emissions to levels that could be achieved by best available control technology.

The report also recommended that the EPA require firms to monitor and disclose their emissions publicly. This would end the “don’t ask, don’t tell” approach.

High stakes

The stakes are very high in the struggle to regulate grandfathered plants, especially coal-fired electric utilities. End-of-the-pipe controls for air emissions are often costly. An industry-financed report to the Energy Information Agency said in 2000 that complying with NSR at old coal-fired electric power plants would cost an estimated $65 billion. For individual plants, installing best available control technology may cost hundreds of millions of dollars.

The public also has a big stake. If power plants reduced emissions of sulfur dioxide and nitrogen oxides by 70 percent from their 1997 levels–a reasonable assumption if all old power plants had to meet the standards for new facilities–recent EPA estimates are that cleanup could prevent 14,000 premature deaths from asthma, other respiratory diseases, and cancer each year.

Do these figures suggest that forcing old plants to clean up would be an efficient use of resources? Caution is necessary; both sides have commissioned numerous studies that make different assumptions and yield different estimates. One simple evaluation strongly suggests that cleanup would be justified on cost/benefit grounds. The federal government generally uses estimates of $3 million to $6 million for the value of a life saved. At this rate, 14,000 deaths prevented in one year would justify more than half of the total cost of installing best available control technologies, and deaths prevented in future years would come free. There would be other benefits as well, such as fewer short-term health problems and less acid rain and haze. But are there better, cheaper ways to save lives and clean the air?

In 1990, Congress amended several sections of the Clean Air Act. It decided not to tangle with NSR but did establish a precedent for later reforms. The amendments set national limits on emissions of sulfur dioxide, allocated shares of this cap to existing plants, and allowed plants either to use their full share or to buy shares from other companies. Under the cap-and-trade system, cleanup costs proved to be far lower than expected. Many plants switched from coal to natural gas; others switched from the high-sulfur coal produced in the Midwest to low-sulfur coal from Wyoming and Montana. Competition spurred the manufacturers of pollution-control equipment to innovate and cut prices. Utilities found more efficient ways of operating power plants. And some firms bought emissions rights, which sold at lower prices than expected.

The 1990 amendments left open the possibility that EPA regulations might allow individual firms to conduct a scaled-down version of cap-and-trade, a process that became known as “plantwide applicable limits” (PALs). Many large industrial facilities have dozens of operating units that discharge pollutants, and the plants must obtain a permit for each unit. PALs allow a plant to obtain one permit for the whole facility. This gives companies an incentive to reduce emissions at the cheapest point, to develop new technologies, or to change processes in ways that reduce pollution cheaply.

Industries that change product design frequently–for example, semiconductors and pharmaceuticals–have been particularly interested in PALs. For them, the complexities and uncertainties of NSR are a major problem, because it can take a year or more to apply for a permit and get a decision. PALs would allow firms to change their production without getting a new permit as long as total emissions from the plant do not exceed the cap. PALs are less critical for electric power plants, because most of their emissions come from a few stacks.

The NAPA report endorsed PALs as well as cap-and-trade programs after plants clean up. It proposed a three-tier system that would include statutory changes and provide more flexibility for plants that can install monitoring equipment to ensure that trading is based on solid information.

Tier 1, Cap and Trade, would feature a national or regional multipollutant system for all fossil fuel­fired power plants, industrial boilers, and similar facilities that can monitor emissions continuously or model their emissions reliably.

Tier 2, Cap and Net, would kick in where continuous monitoring is not economically or technologically feasible but where emissions can nevertheless be reliably monitored or modeled. Emission limits for a facility would be established initially based on the reductions that could be achieved by state-of-the art equipment. These caps would cover all sources of emissions from the facility and thus function as PALs. The facility could then modify any part of its operation as long as its permit lasts (presumably 10 years) without an NSR permit, as long as the emissions cap is not exceeded.

Tier 3, Unit Cap, would set limits for sources not included in the first two tiers.

During the Clinton administration, the EPA experimented with PALs but was not able to negotiate an agreement with industry and environmentalists concerning their widespread use. The regulations adopted by Bush’s EPA authorize PALs but relax NSR requirements.

Some analysts have suggested that caps should be based on meeting current state-of-the-art emissions standards and then should decline over time to capture at least some of the emissions reductions that NSR would normally require. The new regulations are far more liberal. They allow facilities to set emissions baselines at the level of emissions during any two-year period during the previous 10 years. NSR baselines had been based on emissions during the previous two years of operation. The new rules raise the possibility that a PAL cap might be significantly higher than recent emissions. Further, facilities are not required to reduce emissions over the life of the PAL. This allows facilities to continue operating at high emissions levels for several years.

The Bush administration has also proposed an amendment to the Clean Air Act to allow a broader cap-and-trade system. Congress is now considering this Clear Skies proposal, as well as other bills that would require quicker, deeper cuts in emissions. As if this were not sufficiently difficult, the congressional debate is now further complicated by the wild-card question of how much consideration to give to climate change.

The carbon question

Within many industries, there are firms that have struggled to understand and comply with the rigid and complex NSR requirements. The new regulations adopted in 2003 will provide substantial relief. The electric power industry has the biggest stake in further changes. Electric utilities produce 66 percent of the nation’s sulfur dioxide emissions, 37 percent of the carbon dioxide, and 26 percent of the nitrogen oxides. Most of these emissions come from coal-fired plants, of which 80 percent are grandfathered. They produce 55 percent of the nation’s electricity.

At this point, most old coal-fired plants are inexpensive to run. Their owners earned enough long ago to cover the costs of construction, and coal is cheap. Thus, they can sell electricity at relatively low prices and still earn healthy profits, while at the same time helping local manufacturing plants keep costs down. But the future of old coal-fired plants is uncertain.

If the EPA were to enforce NSR requirements as it began to do under the Clinton administration, then some old coal-fired plants might install costly new pollution control devices, while others would switch to natural gas or shut down. However, additional regulations controlling carbon dioxide emissions to deal with climate change could force coal-fired utilities to invest in very different technologies than those required by NSR. Pollution controls for gases regulated today could easily become lost investments if carbon dioxide is regulated. If lenders believe that carbon eventually will be regulated, coal-fired utilities might have to pay a premium now to borrow funds for cleaning up to meet NSR requirements. Thus, if one assumes that the United States eventually will regulate emissions of carbon dioxide, it makes little sense to force coal-fired plants to clean up under current law. Instead Congress should create a cap-and-trade system now for carbon dioxide as well as for pollutants covered by current law.

During the presidential campaign, George W. Bush endorsed cap-and-trade legislation for carbon dioxide as well as for nitrogen oxides, mercury, and sulfur dioxide. Soon after the election, however, electric utilities and the coal industry made a concerted effort to change his position. In February 2001, the president announced his opposition to the regulation of greenhouses gases and to the Kyoto Protocol on climate change, an internationally proposed agreement to reduce such emissions. The administration’s Clear Skies initiative would set caps on nitrogen oxides and mercury to substitute for long-planned command-and-control regulation, and it would relax somewhat the schedule for reductions in sulfur dioxide. Environmentalists claim this would roll back environmental protections even further than the new NSR regulations. The administration argues that the changes, on balance, would keep limits tight and that allowing trades of nitrogen oxide and mercury would sharply reduce the costs of cleanup.

In taking this approach, the Bush administration is putting the United States on a very different path from that taken by Europe in combating climate change. During the 1990s, several countries in Europe reduced their use of coal sharply. France built several nuclear power stations to generate electricity, while Britain, Germany, and the Netherlands switched to newly discovered natural gas from the North Sea or from Russia. The United States has far larger deposits of coal, a quarter of the world’s reserves. In the 1990s, coal production remained high even as utilities built many new natural gas plants that compete with coal. The result is excess capacity for generating electricity in parts of the country, especially the South.

The administration’s approach to NSR, as well as to other air quality issues, seems to rely on developing a technological fix for carbon dioxide that will enable coal to remain a mainstay of energy production for decades. The proposed EPA regulations will let coal-fired plants postpone the difficult choice about which pollutants to reduce. In the meantime, these plants can continue operating indefinitely burning cheap, dirty coal and can preserve their capital for later investment in so-called “clean coal” technologies.

Indeed, the administration has invested heavily in clean-coal technology, with DOE citing the president’s “$2-billion commitment to coal” and calling coal-fired electricity generation plants the “cornerstone of America’s central power system.” DOE issued its initial solicitation for matching grants for clean-coal projects in March 2002, supported by more than $300 million, stating that the ultimate goal of the research efforts is developing “an emission-free coal plant of the future.” The research agenda contemplates the deployment of several new technologies to meet this goal, including coal gasification, advanced combustion, fuel cells, and sequestration of carbon dioxide.

If the past is prologue, however, the outlook for such efforts may not be rosy. The federal government has provided substantial subsidies–research grants, tax breaks, and loans–for clean-coal projects for two decades without achieving a major breakthrough. Nonetheless, elements of the coal industry and some environmentalists are enthusiastic about the promise of integrated gasification combined-cycle (IGCC) plants, which use a technology different from other clean-coal projects. The National Coal Council reported in 2000 that new IGCC plants can compete with natural gas plants when gas costs $3.75 to $4 per million cubic feet, and prices in 2003 have generally exceeded $5 per million cubic feet.

IGCC plants heat coal, water or steam, and oxygen to high temperatures under high pressure. They produce a gas of hydrogen and carbon monoxide that can be used in today’s natural gas-fired electric plants, with minor modifications. They also can produce high-grade chemicals and diesel fuel, and they generate as waste a slag that is less voluminous and less reactive than the vast quantities of waste from conventional end-of-the-pipe technologies.

Carbon sequestration is the other part of the government’s coal-based strategy. The United States, the European Union, Russia, China, and several other countries recently agreed to create the Carbon Sequestration Forum, whose aim is to stimulate research into sequestration technologies to clean up fossil fuels by capturing carbon dioxide at the source and storing it for thousands of years deep underground.

There is plenty of room for skepticism about clean coal. Will new technologies prove adequate to reduce air and other forms of pollution from coal? Can IGCC plants match the record of coal-fired units for delivering electricity reliably and economically? Can carbon dioxide be economically recovered from combustion processes and stored for centuries in seismically stable underground reserves?

If clean coal ultimately proves economically feasible and environmentally sound, then this resource may be an essential part of a global strategy to address climate change. India and China will soon account for almost 40 percent of the world’s population. Both countries have substantial deposits of coal, but little oil and gas. Their economies are growing rapidly and their use of coal is rising sharply. In the long run, the world probably must shift from fossil fuels to alternative sources of energy, but natural gas and perhaps clean coal might be transitional fuels for two or more decades.

Politics at work

Regional interests add to this complex picture. Most of the coal in the United States is in the Midwest and northern Appalachians, the northern Great Plains, and Texas. The utilities that burn coal are concentrated in the Midwest and parts of the South. Indeed, cheap coal-fired electricity is a bulwark of the economy in these regions, and for decades it has been a key factor in attracting and retaining manufacturing plants. As politicians from the Northeast are fond of pointing out, pollution from these coal-fired plants drifts east and north into Pennsylvania, New York, New Jersey, and New England. However, the Northeast cannot blame all of its air quality problems on other states, because cars and trucks contribute substantial pollution, and several of the oldest, dirtiest coal-fired plants are located in this region.

Regional disparities in the costs and benefits of cheap coal-fired electricity lead to conflict. In 1970, Sen. Edmund Muskie, a Democrat from Maine, led the fight to write the Clean Air Act, against resistance from his committee chair, Sen. Randolph Jennings of West Virginia. Today, members of Congress from the Northeast are among the strongest critics of the administration’s NSR reforms, and many members from Ohio, West Virginia, Illinois, and Montana are supporters. The battle lines are not really partisan; they are based on the fuels burned to generate electricity in different states and thus are largely regional. New York’s Republican governor and Maine’s two Republican senators are vocal critics of NSR reforms, and Senate Democrats from West Virginia, Illinois, and Montana are quiet. Attorneys general from 15 states, mostly in the Northeast, have filed lawsuits saying that the administration’s new NSR rules violate the Clean Air Act.

These regional differences have immense political importance. Midwest states, including Ohio, Illinois, Michigan, and Pennsylvania, have historically been swing states in presidential elections. In 2000, coal-mining West Virginia went Republican for the first time in three decades, and Tennessee, where the Tennessee Valley Authority has several coal-fired plants, voted Republican as well.

Regional differences also prompt varying views among electric utilities. Utilities that produce electricity in nuclear plants or with hydropower emit very little of the pollutants covered by NSR. These utilities, along with those that have invested in new gas-fired turbines, compete with coal-fired utilities now that many states have deregulated the generation and transmission of electricity. Utilities that generate electricity with nuclear, hydro, or gas might capture new customers if coal-fired utilities are forced to clean up and raise rates.

Utilities are also divided by their views on climate change. Some utilities have bitten the bullet and endorsed regulation of greenhouse gases now to avoid a second cleanup. In the late 1990s, American Electric Power, which is based in Ohio and burns more coal than any other utility, endorsed federal legislation to regulate carbon dioxide. In contrast, the Southern Company, which burns coal in Georgia, Alabama, Mississippi, and Florida, has lobbied aggressively not to regulate carbon dioxide.

Guiding principles

With so much at stake, maneuvering has been constant, will continue for some time, and may figure in the presidential campaign. But sooner or later federal lawmakers must decide whether to bet on clean coal or insist that coal-fired plants clean up now. Four principles may help guide decisions.

The first principle is to move away from command-and-control regulation as epitomized by NSR. This shift must be made carefully, as NSR is part of a tightly woven fabric of environmental statutes. The NSR process is working far more effectively for new plants than for old. Indeed, these rules play a critical role in managing local air quality. States must write and implement plans to clean the air over metropolitan regions that are home to more than 120 million people, where air quality has improved but still does not meet national standards. The public resists requirements for cleaner cars, fewer roads, or tighter limits on small businesses and backyard grills, which are generally more costly than policies aimed at cleaning up coal-fired power plants. Many states have found it easier to squeeze small increments of cleanup from big facilities, so the companies that own these facilities have a direct interest in helping states educate the public about the need for everyone to share the burden. Furthermore, the Clean Air Act provides that a new plant must not only install best available control technology but also must finance cleanup by other existing sources. If NSR were wiped away, state regulators would lose this tool. For such reasons, NSR should be retained for new plants at least for the time being.

Second, strict enforcement of law is essential. This is the foundation for insisting on cleaning up grandfathered plants. There are so many sources of pollution and so many opportunities for preventing and reducing pollution that no one–the EPA, the states, or environmental activists–can keep track of them all. Environmental regulation depends on voluntary compliance and public trust. Strong targeted enforcement that stops egregious violations, deters others from violating the law, and builds public trust in the regulatory program is essential. Congress clearly expected that grandfathered plants would eventually clean up or shut down, but many have skirted the law. Twenty-five years is enough. They should clean up. When they do, the benefits to public health and to public support for and compliance with environmental regulations will be substantial.

The third principle is to build a greater capacity to adapt to changing conditions. Cap-and-trade programs and PALs give incentives for innovation throughout the production process, not just end-of-the-pipe pollution control. The next step is to set long-term goals for climate change policy. The United States has not yet had a robust national debate about how to fuel its economy in ways that will reduce the risks of climate change. Public debate about climate has focused primarily on the Kyoto Protocol, on the growing rift with the European Union, on whether developing nations should reduce their greenhouse emissions, and on high-profile issues such as oil exploration of the Arctic National Wildlife Refuge. It is time to set 50-year goals for global concentrations and national emissions of greenhouse gases and to begin introducing caps on carbon dioxide into the Clean Air Act.

Finally, any solution must be structured to include opportunities for compensating regions and industries that lose when policies shift. This was a key to the success of the Clean Air Act. During the 1980s, Congress made repeated unsuccessful efforts to address worries about acid rain. However, when the first Bush administration endorsed a cap-and-trade system, Congress was able to pass a bill, partly because it found a way to funnel money to “losing” coal states and coal-fired plants in the form of job training for coal miners and generous emission allowances for those plants. The devil in any political compromise is always in the details, and in environmental matters these details often concern money. Perhaps this is why the battle over NSR has been so polarized; the battles are played out in court, where it is harder to compensate losers.

The debate over NSR takes the form of classic struggles: environmentalists versus industry about how much to spend to reduce pollution, and the Northeast versus the Midwest about who should pay. The disagreement about climate change is more difficult, reflecting not only differences in values and interests but also differences in judgments about how soon action must be taken to address the issue. But both are matters on which compromise should be possible if politicians have the will to think boldly and structure debate properly.

The key is to frame the issues in ways that the public can understand. Rather than a series of changes in highly technical regulations, the administration or Congress should put the central question squarely to the public. Should the federal government exempt old coal-fired plants from clean air standards in order to ensure that coal remains a low-cost option until clean coal technologies are in place? Or should the government reform NSR but stand by its statutory commitment to protect public health?”