The reason for the huge disparity in estimates is twofold. First, because deep saline aquifers have no commercial value, there has been little exploration to determine their extent. Second, the fluid dynamics of how concentrated, liquefied carbon dioxide would spread through such formations is very complex and hard to model. Most analyses have simply estimated the overall volume of the formations, without considering the dynamics of how the CO2 would infiltrate them.
The MIT team modeled how the carbon dioxide would percolate through the rock, accounting not only for the ultimate capacity of the formations but the rate of injection that could be sustained over time. "The key is capturing the essential physics of the problem," Szulczewski says, "but simplifying it enough so it could be applied to the entire country." That meant looking at the details of trapping mechanisms in the porous rock at a scale of microns, then applying that understanding to formations that span hundreds of miles.
"We started with the full complicated set of equations for the fluid flow, and then simplified it," MacMinn says. Other estimates have tended to oversimplify the problem, "missing some of the nuances of the physics," he says. While this analysis focused on the United States, MacMinn says similar storage capacities likely exist around the world.
Howard Herzog, a senior research engineer with the MIT Energy Initiative and a co-author of the PNAS paper, says this study "demonstrates that the rate of injection of CO2 into a reservoir is a critical parameter in making storage estimates."
When liquefied carbon dioxide is dissolved in salty water, the resulting fluid is denser than either of the constituents, so it naturally sinks. It's a slow process, but "once the carbon dioxide is dissolved, you've won the game," Juanes says, because the dense, heavy mixture wou
|Contact: Kimberly Allen|
Massachusetts Institute of Technology