A Wisconsin coal-fired power plant operated by We Energies is scheduled to launch a pilot project to capture a portion of the carbon dioxide produced as the coal is burned. It will be the first time a U.S. power plant has corralled CO2, the main greenhouse gas, before it floats out of the smokestack.

Power plants produce nearly 40% of U.S. carbon emissions; the bulk of that is from coal plants.

The project is a small step on a long road. Alstom, the technology provider, will capture just 3% of the carbon and will immediately release it rather than storing it underground. Carbon storage is widely deemed the biggest hurdle in the worldwide effort to reduce power plant CO2 emissions.

Yet, the pilot program shows that even though the Bush administration recently canceled the clean coal plant called FutureGen, industry is forging ahead, if in a more scattershot style, to strike at the single biggest source of carbon discharges. The Pleasant Prairie, Wis., trial is one of a series of carbon-capture projects Alstom and others are planning at power plants around the nation in the next decade.

The year-long effort, estimated to cost at least $10 million, is being funded by We Energies, Alstom, the Electric Power Research Institute and 35 companies.

"It's a necessary first step," says Robert Hilton, head of business development for Alstom's global environmental business.

Clean coal plants are viewed as vital to fighting global warming. Gas-fired plants emit far less carbon than coal, but natural gas prices are volatile. Wind and solar power are intermittent. Nuclear reactors are emissions-free but pricey and could take many years to build. Despite recent price increases, coal is fairly cheap and abundant.

At the Wisconsin plant, Alstom has built a 90-foot-high addition criss-crossed by huge pipes and heat exchangers to capture the carbon, using a process called chilled ammonia. After coal is burned in a boiler, ammonium carbonate absorbs about 90% of the resulting CO2 to form ammonium bicarbonate, a solid and liquid. The carbon will then be separated under high pressure and released into the air as a gas.

Chilling the carbon and other flue gases eliminates contaminants, such as sulfur dioxide, and permits a much greater amount of carbon to be absorbed, Hilton says. That means the carbon capture uses far less electricity, freeing the power for the grid.

One concern about ammonia is its volatility. "You don't want it coming up the stack," says Howard Herzog, principal research engineer for the MIT Energy and Environment lab.

Hilton says scrubbers will prevent any ammonia from escaping.

The Wisconsin pilot program will be followed by similar but larger trials by Alstom at American Electric Power plants in West Virginia and Oklahoma. Those projects will store CO2 underground or pump it to oil fields to boost output.

Alstom has said carbon capture and storage should be widely available by 2019.

By capturing CO2 after it is produced, Alstom's technology can be used with hundreds of today's traditional pulverized coal plants, Hilton says. General Electric and Siemens are developing technology for a new type of plant that turns coal into synthetic gases, filtering out the CO2 before the gases are burned, a simpler process.

Such plants are 20% cheaper than traditional coal plants, assuming both types add carbon capture and storage, says MIT professor John Deutch.

The FutureGen plant, scheduled to be built in Mattoon, Ill., would have used gasification technology. The Department of Energy canceled it last month, citing construction costs that would have pushed its price tag to nearly $2 billion, most of which the DOE would have funded. Instead, the DOE says it will help fund several smaller gasification plants spearheaded by the power industry around the country.

Deutch says only the federal government can oversee the challenging task of burying carbon in rock formations. Researchers must ensure that the carbon doesn't contaminate water supplies, and officials must determine who is liable if the CO2 leaks to neighboring properties, he says.

Hilton, however, thinks Deutch is underestimating private efforts. "Sometimes government programs prolong a product coming to market," he says.