Carbon Sequestration

Mineral carbonation in peridotite for CO2 capture and storage (CCS)
 

 

 

 

 

Recently, my colleagues and I turned our attention to in situ carbonation of mantle rocks rich in olivine (Kelemen & Matter, Proc. Nat. Acad. Sci. 2008). By speeding up natural processes, scientists propose that rocks from the Earth’s deep interior, exposed on the surface by plate tectonics and erosion, may be able to capture and store billions of tons of CO2 per year. If you compare this to the total human output of CO2 to the atmosphere, currently about 30 billion tons of CO2 per year, you can see that this could make a significant difference in the overall CO2 budget of the planet until alternative energy sources replace global fossil fuel use. Given concern about CO2-driven global warming and ocean acidification, CO2 capture and storage ideas are receiving increasing attention.

Since 1990, scientists and engineers considered using naturally occurring minerals that react with CO2 to form carbonate minerals, as a means to capture CO2 from the atmosphere and store it in solid form where it will remain stable for thousands or millions of years (Seifritz, Nature 1990). The mineral olivine (Mg2SiO4) has been the focus of the most research. Olivine forms the carbonate mineral magnesite via reactions such as 

Mg2SiO4 (olivine) + 2 CO2 (from gas or fluid) = 2MgCO3 (magnesite) + SiO2 (quartz)

Olivine is abundant, has more magnesium than silicon, and reacts readily because it is far from equilibrium with the atmosphere and surface waters. 

Olivine forms 60 to 80% of the Earth’s upper mantle, which in turn comprises about ¼ of the Earth. Collisions of tectonic plates cause mantle rocks to be thrust onto the continents, where they are later exposed by erosion. Initially, engineers focused on using olivine to capture and store CO2 via “ex situ” methods. Olivine-rich rocks were to be quarried, transported to power plants, ground to a fine powder, and mixed with water plus purified CO2 in reaction vessels at high pressure and temperature. These methods for olivine carbonation “at the smokestack” have proved to be relatively expensive, in both financial and energy terms, though engineers continue to seek methods to improve their efficiency (Mazzotti et al., Chapter 7, in Metz et al., IPCC Special Report on Carbon Dioxide Capture and Storage, 2005). Meanwhile, most proposed CO2 capture and storage methods rely on injection of pure CO2 in dense, supercritical liquid form into pore space in sub-surface rocks, or even into “puddles” of dense CO2 on the seafloor. There, scientists hope, the supercritical CO2 will stay.

There is an incentive to look for alternatives, since storage of CO2 in inert, solid carbonate minerals could be safer, and easier to monitor and verify, than storage of CO2 liquid in pore space. Some scientists have focused on “in situ” mineral carbonation, leaving rocks in the ground and using methods such as injection of CO2 to increase reaction rates. Much of this work has been on carbonation of an abundant type of lava, basalt, which contains minor amounts of olivine together with large proportions of alumino-silicate minerals such as plagioclase feldspar. Basalt carbonation is slower than carbonation of olivine, but basalt is abundant near many large power plants (e.g., Goldberg et al Proc. Nat. Acad. Sci. 2008). The idea is to pump CO2 into pore space in basaltic lava, with the expectation that over decades the CO2 will combine with olivine and feldspar to produce solid carbonate minerals

Study of in situ carbonation of mantle rocks is relatively new. We’ve made two key observations. First, based on 14C ages, natural carbonation of olivine proceeds much more rapidly than geologists had previously guessed. We sampled a large peridotite exposure in Oman. Our samples of solid carbonate terraces (“travertine”) forming from spring water emanating from mantle rocks, and of carbonate veins that formed within mantle rocks beneath the surface and were later exposed by erosion, have an average 14C age of about 25 thousand years, whereas scientists had guessed that the veins were 95 to 40 million years old. 

Second, we found that the olivine carbonation process can be “self-heating”. Carbonation of olivine gives off heat energy – basically because it involves condensation of CO2 gas or liquid to form CO2-bearing solids. If the carbonation reaction is fast enough, the evolved heat can offset cooling due to diffusion from hot rocks into their cold surroundings, and due to flow of cold fluid through the rocks. At elevated pressure, perhaps one kilometer below the Earth’s surface, the carbonation reaction goes fastest at about 185°C. Under such conditions, the rock volume could become “self-heating”: we calculated that cold CO2-bearing fluid can be pumped into the rocks at about 4 centimeters per second without cooling the rocks and slowing the reaction. Thus, we propose that one method of speeding up olivine carbonation would be to “jump start” the process by drilling, fracturing, and heating a rock volume at depth to about 185°C, and then pumping purified CO2 plus water into that rock volume. we calculate that such a process would convert billions of tons of CO2 into solid carbonate minerals per cubic kilometer of rock per year.  

Such an in situ process might be less expensive than ex situ olivine carbonation, “at the smokestack”, for several reasons. First, it avoids the cost of quarrying, transporting and grinding rock reactants. Once a rock volume is heated, the energy to sustain high temperature at depth would be provided by the carbonation reaction itself. Sustained pressure would be provided, to some extent, by the presence of overlying rocks. However, the cost of most CO2 capture and storage methods is dominated by the process of CO2 capture at power plants, and this proposal is no exception. Further, a truly globally important process of this type would involve transportation of really large amounts of CO2, comparable to the amounts of oil, gas and coal currently being consumed. 

Thus, we are also considering an alternative possibility. In an end-member scenario, seawater, rather than purified CO2, could be used as a fluid to transport CO2. Shallow seawater maintains CO2 exchange equilibrium with the atmosphere, so using seawater could avoid the costs of industrial CO2 capture and transport. Deeper in the Earth, rocks are hotter, so one could drill into rocks at about 185°C. Thermal convection might drive cold water to circulate down one hole as hot water emerges from another. This would be thousands of times less efficient, in terms of kilograms of olivine transformed to solid carbonate per cubic kilometer of rock per year, than using purified CO2. However, in addition to CO2 capture and transport, this end-member scenario would avoid the costs of pre-heating a rock volume, and pumping fluid at high pressure. If drilling holes and fracturing rocks at depth is thousands of times less expensive than CO2 capture and transport, then this idea could be useful. Certainly, this epitomizes our approach to this problem: Understand the processes of natural olivine carbonation, and then do as little as possible to accelerate these processe sin order to consume globally significant quantities of CO2.

A key problem with our proposed methods for accelerating carbonation of mantle rocks is that pore space might become clogged with newly formed carbonate minerals, and remaining olivine could be armored by a layer of carbonate, isolated from continued reaction with circulating fluid. However, geologic examples show that this self-limiting situation is not inevitable, since some rocks – known as listwanites – become fully carbonated, with virtually every Mg and Ca atom combined with CO2. We believe this is because, when reaction rates are rapid, the volume changes associated with the carbonation reactions create large stresses that fracture the rock, providing new, unobstructed pathways for fluid transport and creating fresh, unreacted olivine surfaces. Study of listwanites, and of reactive cracking processes, are now a focus of our research.