updated 3/15/2004 10:34:35 AM ET 2004-03-15T15:34:35

Larry Monroe pointed to a set of eight manhole cover-size plates mounted on the exhaust vent to limit mercury emissions from Gaston 3, a coal-burning power plant that feeds electricity to a half-dozen southern states.

Gaston 3 and plants like it, the backbone of the U.S. power industry, are the focus of a furious debate over mercury pollution -- how much and how fast the nation should move to regulate a toxic metal capable of causing severe neurological damage, especially to  fetuses and  young children.

Each of the plates at Gaston 3 houses an injector that squirts activated carbon dust into Gaston 3's flue gas. Particles of mercury cling to the carbon, which is then trapped by filters and discarded as toxic waste.

"The equipment is really small and simple, but we still don't know everything about the unintended consequences," said Monroe, emissions control research manager for Southern Co., the owner of Gaston 3. "Using it is like taking your car in for repairs. You're always worried [that your problems] will cascade."

Activated Carbon Injection (ACI) is state-of-the-art technology for mercury control, and at Gaston 3 it removes, on average, 80 percent of the mercury from the plant's stacks. Plant managers say the system would do better with more elaborate equipment.

What's best timeline?
But the Bush administration and its industry allies say research gaps and equipment shortages make it virtually impossible for plants using ACI or anything else to achieve the targets mandated by the Clean Air Act -- a 90 to 95 percent reduction in mercury emissions by 2007.

Instead, the Bush administration has used a different section of the act to propose a new target of a 29 percent reduction by 2007 and an industry-wide reduction of 70 percent by 2018. That prescription is scoffed at by environmentalists, who say ongoing tests at Gaston 3 and elsewhere show the Bush targets are far too modest.

At the Environmental Protection Agency, however, officials say new technologies will not be ready to approach 90 percent removal for years: "We don't think ACI is going to be commercially available" by 2007, said William Wehrum, of EPA's Office of Air and Radiation, which opened hearings on the administration proposal late last month. Only after 2010 do EPA targets envision some use for carbon injection systems, Wehrum added.

Coal power by the numbers
The United States has 1,100 to 1,200 coal-fired power plants, most of them several decades old. Gaston 3 was built in the 1960s, one of five generating units in Southern Co.'s E.C. Gaston complex, on the Coosa River in woodland 45 miles southeast of Birmingham.

These venerable and reliable plants emit five principal pollutants: sulfur dioxide, or "SOx," the leading cause of acid rain; nitrogen oxides, or "NOx," which contribute to smog-forming ground-level ozone; carbon dioxide, a greenhouse gas implicated in global warming; lung-damaging soot particles, known as "fly ash"; and mercury -- 48 tons of it annually, the country's single largest source of mercury pollution.

Technologies to control these emissions run the gamut from the proven to the nonexistent. For almost a century, power companies have used electrostatic precipitators to remove fly ash by attracting it to electrically charged metal wires. By contrast, there is no technology available to cope with carbon dioxide.

When the impetus to regulate mercury arose in 2000, the EPA eventually concluded that every power plant should be required to reduce mercury emissions to the level achieved by the cleanest 12 percent. This would require overall reductions of 90 to 95 percent by 2007.

Problem with one-size-fits-all
Experts at first were optimistic about making the deadline, because the EPA believed that equipment used to limit NOx and SOx emissions would also remove mercury. "The idea was that if you had this stuff, you had it made," said Michael D. Durham, president of Denver-based ADA Environmental Solutions and a leading expert on mercury control technologies. "We've learned a lot since then."

By the end of 2002, however, a first round of experiments sponsored by the Department of Energy and the industry had shown that a one-size-fits-all approach would not work. Technology to remove SOx and NOx extracted one type of mercury -- the ionized form in compounds such as mercury chloride or mercury oxide, but was useless in removing elemental mercury -- the vaporized metal itself. Power plants emit both pollutants in varying proportions depending on coal type, flue-gas temperature and plant configuration.

Experts considered several fixes, but each carried complications. Plants could switch from western coal, which emits relatively high levels of elemental mercury, to Appalachian bituminous coal, with higher proportions of ionized mercury. But many plants were built to burn particular coals, and few had the on-site ability to mix coal types. Furthermore, Appalachian coal has more sulfur, which would require plants burning it to install expensive "scrubbers" to avoid making acid rain worse.

As these dilemmas materialized, the industry argued it was impossible to meet the 90 percent mercury deadline by 2007. "We haven't tested [control technologies] in the long term," said Michael Rossler, manager of environmental programs for the Edison Electric Institute, an industry umbrella group. "There is no commercially available control technology for mercury right now."

Cap-and-trade proposed
After consulting with industry, the EPA in December backed away from the strict standard, instead using another section of the Clean Air Act to propose a "cap-and-trade" program whereby companies could either control their own mercury emissions or buy "credits" from other companies that have already done so. That proposal requires that overall emissions be trimmed by 70 percent by 2018, with each plant cutting mercury by 29 percent by 2007.

Industry representatives argue that market forces make cap-and-trade a superior method of achieving greater mercury reductions faster. Advocates do not necessarily disagree, but say cap-and-trade is meaningless with such low targets.

"There is no market with these phony control levels," said David Hawkins, director of the Natural Resources Defense Council's Climate Center. "They've rigged the game by setting the levels so low that most plants won't have to do anything to reach compliance in 2007. Cap-and-trade is an excuse to do nothing."

But research has led many experts working for power companies and the Energy Department to conclude that industry is probably correct in saying that a 90 percent emissions reduction is not feasible by 2007.

"We're further ahead than we were six months ago," said George R. Offen, of the industry-funded Electric Power Research Institute (EPRI), but results from a second round of experiments that began last year "will only be understood in 2006 . . . quite late to make the deadline."

Too much time?
Still, when experts talk about needing extra time, almost all of those interviewed spoke of a grace period of a few years -- never of stretching the deadline to 2018. The Edison Electric Institute's Rossler predicted that research into carbon injection would produce "broad, industry-wide application technologies" by 2009.

"The huge challenge before us is the capture of elemental mercury," said Steve Benson, of the University of North Dakota's Energy and Environmental Research Center.

Researchers are testing two strategies -- direct capture and converting elemental mercury to the ionized form. Benson is supervising experiments to determine whether spiking flue gas with salt or other halide compounds will encourage mercury oxide formation. Other experiments are investigating ideas including catalysts mounted in vents and gold films designed to absorb elemental mercury as it flies by.

But five of the 11 experimental projects, including Gaston 3, involve carbon injection along with combinations of eastern and western coals, vent configurations and equipment. "We need to determine exactly what removal we can get as you go from one plant, or one coal, to another," ADA's Durham said. "When's the right time to pass the regulation? That's the $2 billion question."

At Gaston 3, flue gas passes through the metal wires of an electrostatic precipitator before getting to the injectors. After the injectors, the mercury-impregnated carbon heads for a second filter, a set of stocking-like receptacles known as a "baghouse," which collects it.

This combination, which exists at only two plants nationwide, performs better than other configurations. But doctoring the flue gas in an eight-story coal-devouring monster such as Gaston 3 is not a task to be undertaken lightly. The vents alone are 30 feet high, the precipitator is as big as a four-story Georgetown rowhouse and the baghouse is more than half that size. And at 270 megawatts, Gaston 3 is a smallish plant.

Cost scenarios
According to EPRI's Offen, environmental controls account for 25 percent to 30 percent of a power plant's cost. A full suite of controls, including precipitator, baghouse and SOx and NOx equipment, could cost more than $80 million at Gaston 3, which does not have SOx or NOx controls.

From this perspective, an ACI system is a potential industry darling -- small, easily installable and costing a relative pittance at $500,000. Gaston 3's experiment involves a 40-foot carbon silo about the diameter of a large pine tree and a four-inch rubber hose that carries the carbon to the injector nozzles.

But carbon injection has its constraints. Baghouses are expensive, and no one is building them in large numbers. Carbon for the system costs $500,000 to $700,000 a year, and there is no supplier big enough to handle the entire industry. Plants without baghouses will no longer be able to sell fly ash from their precipitators, because carbon makes it unsuitable as a cement additive.

Although Durham acknowledged that carbon injection still needs to be tested for different kinds of coal, results at Gaston 3 and elsewhere suggest it can be ready, at least as an inexpensive interim solution, almost immediately -- and certainly long before 2018.

© 2012 The Washington Post Company


Discussion comments


Most active discussions

  1. votes comments
  2. votes comments
  3. votes comments
  4. votes comments