The long-term fate of CO2 in the subsurface: natural analogues for CO2 storage

Abstract

AbstractCO2 is a common gas in geological systems so that planned storage of CO2 in the subsurface may do no more than mimic nature. Natural CO2 has a wide number of sources that can be at least partly identified by carbon stable isotope geochemistry. Three pairs of case studies with different reservoir characteristics and CO2 contents have been examined to assess the natural impact of adding CO2 to geological systems. Carbonate minerals partially dissolve when CO2 is added simply because the CO2 dissolves in water and forms an acidic solution. Therefore, carbonate minerals in the subsurface are not capable of sequestering secondary CO2. The addition of CO2 to a pure quartz sandstone (or a sandstone in which the supply of reactive aluminosilicate minerals has been exhausted by excess natural CO2 addition) will have no consequences: the CO2 will simply saturate the water and then build up as a separate gas phase. The addition of CO2 to carbonate cemented sandstone without reactive aluminosilicate minerals will induce a degree of carbonate mineral dissolution but no solid phase sequestration of the added CO2. When CO2 is naturally added to sandstones it will induce combined aluminosilicate dissolution and carbonate cementation if the aluminosilicate minerals contain calcium or magnesium (or possibly iron). Examination of a CO2-filled porous sandstone with abundant reactive aluminosilicate minerals that received a huge CO2 charge about 8000 to 100 000 years ago reveals minimal evidence of solid phase sequestration of the added CO2. This indicates that either dissolution of reactive aluminosilicates or precipitation of carbonate minerals is relatively slow. It is very likely that the slow dissolution of reactive aluminosilicates is the rate-limiting step.Solid phase sequestration of CO2 occurs only when reactive aluminosilicates are present in a rock and when the system has had many tens to hundreds of thousands of years to equilibrate. The two critical aspects of the behaviour of CO2 when injected into the subsurface are (1) that the rock must contain reactive Ca and Mg aluminosilicates and (2) that reaction to produce carbonate minerals is extremely slow on a human timescale. The reactive minerals include anorthite, zeolite, smectite and other Fe- and Mg-clay minerals. Such minerals are absent from clean sandstones and limestones but are present in ‘dirty’ standstones (lithic arenites which are mineralogically immature) and some mudstones.The analysis of geological analogues shows that injection of CO2 into carbonate-bearing rocks that do not contain reactive minerals will induce dissolution of the carbonate, whether it is a matrix cement, rock fragment, fault seal or part of a top-sealing mudstone.