Mineralization of carbon dioxide sequestered in volcanogenic sandstone reservoir rocks

• We proposed to use volcanogenic sandstone formations for CO2 sequestration.
• Such sandstones have high content of reactive minerals for CO2 mineralization.
• We determined porosity and permeability from mineralogy data.
• The reactive transport code TOUGHREACT was used to calculate CO2 mineralization.
• Results show that around 10–20% reactive minerals are optimal for CO2 mineralization.

Geological storage of carbon dioxide in deep saline formations can decrease the accumulation of CO2 in the atmosphere, and thus slow down global warming. Most CO2 injected into subsurface rock formations is expected to remain for a long time as either a separate supercritical phase or in solution in brine; both forms present the possibility of leakage back to the surface or other environmental impacts. Mineralogical trapping of injected CO2 is more secure but usually thought to be too slow to add significantly to sequestration security. For quartz-rich sandstones (quartzarenite and arkose), only ca. 5% CO2 mineralization is achieved over 1000–10,000 years (Audigane et al., 2007). However, if volcanogenic and other sandstones that have larger amounts of reactive minerals were used for storage, there could be a larger fraction of CO2 mineralized in a shorter time. The limitation is that porosity and permeability tend to decrease with increase of volcanic rock fragments (VRF), which limits the rate at which CO2 can be injected. We evaluate these tradeoffs to assess the feasibility of using volcanogenic sandstone to achieve secure CO2 storage. Using relationships between VRF percent, porosity and permeability from available geological data, the reactive transport code TOUGHREACT was used to model the flow, transport, mineral reactions, changes in fluid chemistry, and the rate and extent of CO2 mineralization over 1000 years during and after CO2 injection into a sandstone reservoir. We use the models specifically to evaluate the expected trade-off between higher reactivity and lower porosity and permeability. A model volcanic fragment mineralogy is used (pyroxene and feldspar mainly for which kinetic data are available) along with conservative estimates for silicate and oxide mineral dissolution kinetics and reactive surface area. Substitution of other more common reactive minerals such as chlorite and amphibole would not significantly change the results. The simulations show that in rocks with 10–20% reactive minerals, as much as 80% CO2 mineralization could occur in 1000 years and still allow sufficient injectivity so that 1 Mt of CO2 could be injected per year per well. The calculated mineralized fraction depends on several factors, most notably the kinetics and reactive surface area of dissolving silicates and the detailed relationship of reactive mineral content to effective permeability and injectivity.