A few industrial-scale projects are in operation, including in Norway, Canada and Algeria, but none tests all parts of the capture process. Heat-trapping carbon dioxide can be taken from the exhausts of a coal-fired power plant, for instance, then piped underground into porous rocks.

The IEA estimates that after the $56 billion investment in CCS globally from 2010-2020, a further $646 billion will be needed from 2021 to 2030, Tanaka told the Carbon Sequestration Leadership Forum.

U.N. studies have indicated that CCS could do more to limit greenhouse gas emissions this century than a shift to renewable energies such as wind or solar power. CCS has been limited by high costs.

That final point, the potential effectiveness of CCS in reducing CO2 emissions, has recently been underlined in a diary by nanne and it's well worth repeating:

In terms of climate change, the upside of CCS is that the technology can remove up to 80-90 percent of carbon dioxide emissions at what could be a lower cost than replacing fossil fuel plants with renewables, with the main downside being uncertainty related to sequestration. However, coal has a variety of negative effects aside of causing climate change, like emissions of heavy metals and the destruction of nature and landscape that accompanies its extraction.

What really makes CCS a misdirected research strategy, though, is that it will cause efficiency losses of at least 10% for capture, and a few more for transport and sequestration. The European Commission estimates a 35% net efficiency for a demonstration project to be launched in 2015, as opposed to a 45-47% current net efficiency for new coal plants without CCS. So, we will need a lot more coal to feed CCS plants. While coal is not yet on the brink of running out, there is less cheap coal available for extraction than has been assumed

And nanne's final point may be the elephant in the room Tanaka doesn't really (want to) notice: Peak Coal anyone?

World coal URR value has been estimated at between 700 and 1243 Gt of coal. The model projects that worldwide coal production will peak between 2010 and 2048 on a tonnage basis, and between 2011 and 2047 on an energy basis. The notion that coal is widely abundant therefore appears to be unjustified. Further work is needed to better determine the URR range of coal.

Bold mine. So by 2050, these envisioned 3400 plants of Tanaka will be run on... what? Ponies, I guess.

An additional side note, I don't think I've yet come across an analysis that actually incorporates such things as supplementary environmental damage or a honest estimate of human deaths due to silicosis, exploding methane pockets or increased risk of lung cancer due to (increased) coal mining. One wonders if it exists.