The standard knock on carbon capture and sequestration (CCS) is that it hasn't been tested and proven on an industrial scale. That's really only true in the narrow sense in which you start with coal, produce electricity, and then collect and bury the CO2 that comes out the stack--which I imagine is what CCS evokes for most people who have even heard of the technology. Some years back, the US government set out to close that gap by building a large-scale test facility to demonstrate the coal-to-CCS cycle, with help from a consortium of industry partners. The program was called FutureGen. It died in 2008 after reported cost overruns but was revived in a different format last year. Now the reoriented effort has spawned a new project at a different location--though still in Illinois--to replace the ill-fated Mattoon project. Its basic concept differs significantly from the original FutureGen, and in ways that might improve the odds that coal could continue to contribute a substantial share of the US energy mix for many decades.

The CO2 produced by power plants is much harder to capture and dispose of than the traditional pollutants we associate with them, not least because it is the primary chemical result of the combustion of hydrocarbons, along with water vapor, rather than a byproduct resulting from a fuel impurity or imperfect combustion. That requires dealing with emissions that exceed the mass of fuel being consumed, rather than an order of magnitude or two smaller. And when fossil fuels are burned in air, the CO2 produced must be separated from all that nitrogen, which is the largest constituent of flue gas, before it can be sequestered. All this is expensive, in both energy and financial terms. The original FutureGen was designed to finesse this problem by converting coal into a hydrogen-rich gas that could be burned efficiently in a combined-cycle gas turbine (IGCC), producing emissions consisting mainly of water vapor, plus a sequestration-ready CO2 stream from the hydrogen-production process. Unfortunately, the hardware necessary to do that isn't cheap, either.

FutureGen 2.0, as announced, would take a different tack. It aims to convert an existing power plant owned by Ameren Corporation into an "oxy-coal" plant, in which pure oxygen replaces air in the boiler for combustion, resulting in flue gas consisting mainly of CO2. This approach has pluses and minuses, compared to IGCC. It requires a bigger air separation plant to support full combustion, but it eliminates all the hardware associated with hydrogen. That should entail somewhat lower capital costs, but not necessarily lower operating costs, particularly when you consider that the efficiency of IGCC exceeds that of most existing US coal power plants, though not necessarily supercritical or ultra-supercritical pulverized coal plants. (I couldn't tell how much the basic power block of Unit 4 of Ameren's Meredosia, IL plant, which formerly burned fuel oil, will be modified.) As in FutureGen 1.0, the resulting compressed CO2 would then be pipelined to a disposal site elsewhere in the state.

Although it would take some doing to convince me that oxy-coal with CCS is a better technology than IGCC with CCS, the revised approach to FutureGen looks like a good call on the part of the government. That's because the context in which FutureGen is being pursued has altered significantly since it was first devised. Instead of a scenario of continuing to build many new coal-fired power plants every year to meet steadily-growing electricity demand, the future--at least in the US--looks quite different. An article in yesterday's Washington Post pointed out that a number of new coal plants are still under development, but the rate of new construction has slowed dramatically, due to regulatory pressures, weaker electricity demand, competition from cheaper natural gas, and the growth of renewables. If we want to have an impact on the emissions from the US coal-fired power plant fleet--which accounts for 31% of total US emissions and 91% of the emissions from the electricity sector--then our best strategy probably doesn't involve building hundreds of gleaming new IGCC plants, but rather retrofitting hundreds of existing units built with older technology, for which conversion to IGCC would likely be cost-prohibitive. If FutureGen 2.0 succeeds--technically, if not economically--it would validate that retro-fitting potential.

The world hasn't stood still while the Department of Energy wrestled with all the political and technical challenges that FutureGen faced. The original siting competition between Texas and Illinois looked like a textbook case of logrolling, and FutureGen 1.0 exhibited the hallmarks of a classic government boondoggle. Meanwhile, commercial projects such as Duke Energy's Edwardsport IGCC (without CCS, but in effect CCS-ready) and the Good Spring IGCC project of Future Fuels LLC have emerged and appear to be making progress. The latter is based on technology from the Thermal Power Research Institute of China, which is a good bet to beat all of these projects to the punch with its GreenGen power plant in Tianjin. If FutureGen 2.0 is going to matter, it must be built smartly, quickly and cost-effectively. Yet technical success still won't guarantee that this technology will be taken up and deployed widely. In a market economy, rather than a centrally-planned one, it's hard to see any of this going beyond a demonstration plant or two without a substantial price on CO2 emissions to offset the inherently higher costs of generating power this way.