Speaking of coal, what's the deal with those coal plants of the future that will supposedly capture their carbon and store it deep underground, rather than belching it into the air? And, more importantly, how much is it going to cost? Earlier this month, Harvard's Belfer Center put out a new study on "Realistic Costs of Carbon Capture" to try to answer this question.

The central finding here is that coal capture and storage (CCS) may, someday, offer a cost-effective means of curbing greenhouse-gas emissions—if it proves feasible on a large scale. The coal industry would still be ripping apart Appalachia, but at least it wouldn't be roasting the planet. In the short term, however, demonstrating that CCS can actually work is going to be extremely costly. The initial round of plants will capture carbon for around $100 to $150 per ton, which is about ten times as high as the price of carbon will likely be in the first decade under the House climate bill. (A caveat: That pricetag doesn't factor in potential revenue from pumping the captured carbon into wells to recover oil, which, according to the report, could potentially slash the cost of those initial plants by quite a bit.)

This means that if all Congress does is set up a cap-and-trade program to regulate carbon pollution, utilities won't find it worthwhile to build CCS plants for a long time. They'd look toward other, cheaper ways of reducing emissions. CCS will only become a commercial reality via additional subsidies and government-backed demonstration projects. Right now, only a handful of pilot plants exist around the world. That explains why, in the current congressional debate, coal-state Dems are working overtime to subsidize CCS in the climate bill. Without that assistance, coal's screwed.

The semi-cheery tidings for CCS is that as more and more of these plants get up and running, the technology will improve, economies of scale will kick in, engineers will learn how to tweak the designs, and the costs should, in theory, tumble down to about $25 to $50 per ton—making CCS plants competitive under a cap-and-trade regime. Two points here: First, there need to be a fairly vast number of CCS plants up and running—say, 50 to 100 gigawatts worth—for costs to sink to that level. Second, the "learning rate" for CCS plants is pretty comparable to other clean energy sources, except for solar (where costs seem to tumble even faster as more panels are produced) and nuclear (where costs don't appear to fall very much as you build more plants).

If the coal folks cross their fingers, get their lavish subsidies and demo projects, and everything works as hoped for, then here's how much the Belfer Center projects a bunch of different energy sources could cost in or around 2030:

This chart, note, doesn't factor in the effects of a carbon price—under a cap-and-trade regime, the two fuel sources on the left, coal and natural gas, would be a lot more expensive because they have to pay for the pollution they produce. In this scenario, wind power is the cheapest low-carbon source in 2030, followed by CCS, followed closely by solar thermal, followed by nuclear and solar photovoltaic panels. (Actually, photovoltaics would probably be cheaper than nuclear because they have lower transmission costs.)

So should we believe the "clean coal" hype? It's really mixed. According to this report, it's certainly possible for CCS to reach the stage where it's competitive with other low-carbon sources. But getting there will require truckloads of public support in the next two decades, and those first steps will be very costly. It's certainly one option for clean-energy policy, but whether it's the optimal route is another question entirely.

--Bradford Plumer