The Rock That Ate CO2
When Peter Kelemen started his fieldwork in the Omani desert several years ago, he wasn't looking for a solution to global warming. His research, which concerned the geology of the earth's mantle-the 1,800-mile-thick layer beneath the crust-was something only his rock-obsessed colleagues could appreciate.
To study ancient volcanism, Kelemen frequently collected a common mantle rock called peridotite, large tracts of which can be found in Oman, forced to the surface over many ages by tectonic collisions. When peridotite is exposed to the air, it reacts with carbon dioxide, and its outer layers are transformed into carbonate rock. For Kelemen, a geologist at Columbia University, this seemed to be bad news. "When the rock is all weathered and turned into carbonate, that obscures the high-temperature history," he says. "So my main response over the years when I would see these carbonate deposits was to run the other way."
Recently his response has undergone a metamorphosis. Kelemen now thinks that with some modest technical investment, peridotite formations-which also exist on or near the surface in California and New Guinea, on the Aegean coast, and on some Pacific islands-could be used to slow global warming by absorbing billions of tons of carbon dioxide from the atmosphere every year.
Kelemen's epiphany came about a year ago, when he and a Columbia colleague, Juerg Matter, decided to date Oman's carbonate rocks. Geologists knew that the chemical reaction between the surface layers of peridotite and carbon dioxide occurs rapidly. "If a leaf or a pebble falls on these rocks and you come back a few days later, it's all covered up with carbonate," Kelemen says. But underground veins of carbonate were thought to take millions of years to form. Kelemen and Matter found otherwise.
"We took a bunch of samples and sent them to Woods Hole for carbon-14 dating," Kelemen says, referring to the oceanographic institute in Massachusetts. "When all the dates came back, every single sample was less than 45,000 years old, and I really started to get excited."
The rocks' young age meant that they were still forming, probably fed by carbon dioxide dissolved in groundwater. Kelemen and Matter calculated that Oman's peridotite deposits naturally soak up about 100,000 tons of carbon annually. Although that is only a small fraction of the 30 billion tons of CO2 we throw into the atmosphere every year, it still dwarfed previous estimates of peridotite's appetite for carbon. More important, Kelemen and Matter realized that the carbon-storing potential of peridotite beneath the earth’s surface remained largely untapped.
In a paper published last November in Proceedings of the National Academy of Sciences, Kelemen and Matter proposed that peridotite's CO2 absorption rate could be ramped up by a factor of 100,000 using conventional oil-drilling technology. Pressurized CO2 and hot water could be pumped into peridotite formations. Heat from the water would accelerate the reaction that forms carbonate rock. After this kick-start, the reaction, which releases its own heat, would become self-sustaining. The formation of new carbonate would cause rocks to buckle and shift imperceptibly. "Not really earthquakes," Kelemen says. "Just little stress releases in the subsurface, but nothing serious.
"In principle-and when a scientist says that, you're going to hear some science fiction," Kelemen warns-"if human output remains the same, and the Omanis could somehow convert every single kilogram of peridotite in that country into solid carbonate, they could take up all the human output for a thousand years. Rocks are very dense; the air is not."
Realistically, Kelemen estimates that the peridotite in Oman could be used to lock up roughly one-eighth of the CO2 produced every year by the burning of fossil fuels. He and Matter envision two complementary strategies for storing CO2 in the world's peridotite formations. The most straightforward approach would be to build power plants near peridotite beds, where carbon emissions could be captured on-site and pumped directly into the rocks. But that wouldn't make a real dent in global emissions.
The second approach would take advantage of an enormous, natural carbon-capture system: the oceans. Surface ocean water absorbs carbon dioxide from the air, so the CO2 content of the surface water is in equilibrium with the atmosphere. Kelemen and Matter propose that the seawater overlying shallow peridotite formations off the coast of Oman, the western United States, and elsewhere could be heated and injected beneath the seabed. The peridotite would react with the seawater, removing carbon dioxide to form carbonate rock. The CO2-depleted seawater would be returned to the surface to absorb more of the gas from the air, and the cycle would start again.
Given the right combination of peridotite and shallow seafloor, this could work anywhere. "Seawater is free," Kelemen says, "and the air transports CO2 all over the world for nothing."
Kelemen and Matter are now looking for a site in Oman for a pilot project, which they hope to launch within five years, probably by capturing carbon dioxide from the flue gases of a power plant and pumping it into peridotite formations on land.
Their idea seems sound, says Wally Broecker, a geophysicist at Columbia and one of the world's leading advocates of harnessing natural processes to reduce greenhouse gases. "When we've done everything we can with the other stuff, with alternative energy and carbon taxes, we're going to find that we're still dumping CO2 into the air at an alarming rate," Broecker says. "It's a desperate situation, and we've got to push everything we can."