Can we hide carbon dioxide underground? Algeria site offers note of caution.

A facility in Algeria that captured carbon dioxide on an industrial scale – and locked it up deep underground – is yielding this lesson for researchers exploring ways to deal with global warming: Select a site with care, because the unexpected can happen.

A new study that aims to explain why sequestered CO2 was moving surprisingly quickly through rock formations beneath In Salah, a natural-gas extraction site in central Algeria. In Salah hosted the second-largest industrial-scale sequestration demonstration project after Norway's Statoil, which has been conducting a sequestration demonstration at the Sleipner field in the North Sea since 1996.

The new study of In Salah's effort identifies the injected CO2 itself as a key culprit. The facility was injecting the unwanted greenhouse gas at a rate that boosted the pressure of the CO2 stored in a sandstone formation more than 6,000 feet below the surface.

The rock either fractured under the pressure or the pressure widened existing fractures, allowing the CO2 to migrate into the first of two layers of denser "cap rock," selected to prevent the gas from leaking to the surface and entering the atmosphere, explains Joshua White, a researcher at the Lawrence Livermore National Laboratory in Livermore, Calif., who studies the flow of fluid through porous or fracture rock.

Dr. White and his team say the data so far indicate that the CO2 remains in check. The pressures at depth haven't dropped, as they would if the gas were escaping the reservoir. A shallow aquifer above the capstone layers shows no evidence of changes in water chemistry. And ground- and space-based sensors that have been monitoring the site have detected no signs of additional CO2 migration.

Still, White says, "It's pretty clear that the pressures were high enough to create new fractures."

At the same time, he adds, researchers know that all geologic formations are fractured to some degree. So the pressurized gas would have pried open existing fractures, providing additional pathways for the gas to migrate, he adds.

The results appear in the current issue of the Proceedings of the National Academy of Sciences.

Carbon dioxide emissions from coal-fired power plants represent about 25 percent of all US CO2 emissions and 44 percent of global emissions. The notion of capturing them has been seen as a way to continue using coal as fuel while reducing its effect on climate.

The Intergovernmental Panel on Climate Change noted that carbon capture and storage are key tools in containing the problem. If they are not used extensively globally by 2030, the odds are slim that greenhouse-gas emissions can be held at a level that by century's end would limit global warming to about 2 degrees Celsius above preindustrial levels, it said. 

Ironically, the mechanism involved in the migration of CO2 at In Salah is the same one used to tap oil and gas reserves locked up in shale formations. There, fluids are injected at high pressure to break up rock to release the gas – a process called "fracking." In addition, CO2 has been pumped into marginally producing wells to enhance their yield.

In Salah was extracting natural gas for export to Europe. But the gas has a relatively high CO2 content, which needs to be removed before shipping. In Salah's industrial partners, as well as a large group of researchers and the US Department of Energy, saw an opportunity to have the plant serve as a demonstration project for sequestering the CO2 in nearby underground formations.

As a result, the site has been intensely studied. Beginning in 2004, In Salah injected the CO2 extracted from the natural gas deep underground. Injection stopped in 2011.

As the underground formation accepted the gas, the surface above the formation began to bulge, as expected. But the bulge appeared in an unusual pattern in the satellite-borne radar data, White says. It suggested that the pressure was forming or widening vertical fractures in the rock.

Until now, much of the research has focused on trying to figure out if the CO2 indeed was migrating into the lower cap rock, or whether the readings indicating migration were spurious, White explains. The general conclusion: Yes, the CO2 had migrated.

White and six colleagues at Lawrence Livermore sought to figure out how. Several lines of evidence pointed to inadvertent fracking as a key player.

The radar evidence showed a subtle trough in the surface that one would expect to see from the stress of fracturing underground. In addition, CO2 began to appear at an abandoned well not far from one of the injection wells. The CO2 appeared far sooner than predicted by models of CO2 flowing through the rocky reservoir below. The alignment of the two wells was similar to the trough's alignment. Finally, seismic data gathered in 1997 and again in 2009 indicated that the brine that once filled the spaces between the rocks in the reservoir had been replaced by CO2.

The site was selected for its multiple rock layers that don't conduct fluids very efficiently – layers that collectively would act as a barrier to CO2's upward mobility. Yet even these individual layers can permit some fluid flow.

"There was always some expectation that CO2 would migrate from the reservoir a little ways up into the cap rock," White says, adding that what he and others found somewhat surprising was the degree of migration.

Indeed, he and his colleagues note that one "large and important uncertainty" in siting is the influence existing fractures have on the behavior of cap rock.

In a summary of lessons learned at In Salah, a team led by Statoil's Philip Ringrose noted last year that in addition to improved modeling of a site based on detailed geological studies, as well as continuous monitoring with a full suite of instruments, operators need to update their risk assessments on a regular basis to ensure their operations aren't threatening to compromise the natural barriers they rely on to hold the CO2 in place.