Many long-term strategies for combating climate change count heavily on the ability to capture huge amounts of carbon dioxide from the burning of fossil fuels and store it permanently in deep underground rock formations. But high costs and lingering technical uncertainties mean the technology, so-called carbon capture and storage (CCS), might not be able to play a significant role in cutting carbon emissions.

A recent report from the International Energy Agency warns that the development and deployment of CCS is "seriously off pace" as a way to prevent the average global temperature from rising more than 2 °C—a widely used target in climate strategy. The window to begin applying CCS toward consequential emissions reduction is "shrinking fast," says the agency, which has declared that CCS must supply over a fifth of the emissions reductions needed by 2050 to keep the temperature rise below 2 °C.

Today, there are eight large-scale (capturing, transporting, and injecting at least 400,000 metric tons annually) CCS projects in operation, according to the Global CCS Institute, none of which are at power plants. These projects, in all, bury nearly 20 million metric tons of carbon dioxide per year. By comparison, coal burning in the United States and China emits about 2.1 billion and 6.95 billion metric tons, respectively, each year.

To meet the 2 °C goal, says the IEA, a minimum of 110 additional projects at power plants and industrial facilities should be brought on line by 2020—enough to capture and store 269 million metric tons of carbon dioxide that year. Although 67 large-scale projects are in planning or construction phases, it can take more than a decade to build a new CCS project.

A big hindrance to CCS is its price tag. Chemically separating carbon dioxide from plant exhaust or natural gas streams is expensive. Before the gas can be buried, it must be compressed to a supercritical state and transported via pipeline to the injection site—two processes that are also expensive. Without incentives, CCS adds too much to the price of power production from existing plants to be cost-effective.

Information about the global storage capacity is limited, but a 2012 study by MIT researchers found that in the United States, underground rock formations called deep saline aquifers could hold at least a century's worth of carbon dioxide emissions from the nation's coal-fired power plants.

Many of the candidate reservoirs, however, are untested, and it's not clear how they might respond to large volumes of injected carbon dioxide. "The problem in a lot of these places (within deep saline aquifers) is that the permeability is very low," making it harder to get the fluid into the rock, says Mark Zoback, a professor of geophysics at Stanford University. Inserting fluid causes pressure changes that can induce seismicity, and "you can't inject these super-large volumes without the potential for triggering earthquakes." Even small quakes that might occur on faults "easily missed" during EPA-required site characterization studies could let the greenhouse gas escape, he says.

Earthquake risks do not entirely disqualify the technology, says Zoback. But they make it unlikely that CCS can "significantly" reduce greenhouse-gas emissions, he and Stanford hydrogeologist Steven Gorelick argue in a paper published last week.

Captured carbon dioxide is already widely used in a technique called enhanced oil recovery, in which the gas is pushed into an oil reservoir to chemically mobilize hard-to-get hydrocarbons, making them easier to pump out. The injected gas is then either trapped in the oil reservoir or is "produced" while extracting oil and re-injected. This technically qualifies as CCS.

John Litynski, the carbon storage technology manager for the National Energy Technology Laboratory's office of coal and power, sees enhanced oil recovery as a way to kick-start the CCS industry, since the NETL estimates that around 20 billion metric tons of carbon dioxide could be economically stored this way. "You've got a market driver with the oil production and you've got really well understood reservoirs for storage," he says. "It's probably going to be the first mover."