Reactive transport modelling of CO2 storage in saline aquifers to elucidate fundamental processes, trapping mechanisms and sequestration partitioning
James W. Johnson, John J. Nitao, Kevin G. Knauss, 2004. "Reactive transport modelling of CO2 storage in saline aquifers to elucidate fundamental processes, trapping mechanisms and sequestration partitioning", Geological Storage of Carbon Dioxide, Shelagh J. Baines, Richard H. Worden
Download citation file:
The ultimate fate of CO2 injected into saline aquifers for environmental isolation is governed by three interdependent yet conceptually distinct processes: CO2 migration as a buoyant immiscible fluid phase, direct chemical interaction of this rising plume with ambient saline waters, and its indirect chemical interaction with aquifer and caprock minerals through the aqueous wetting phase. Each process is directly linked to a corresponding trapping mechanism: immiscible plume migration to hydrodynamic trapping, plume-water interaction to solubility trapping, and plume-mineral interaction to mineral trapping. In this study, reactive transport modelling of CO2 storage in a shele-capped sandstone aquifer at Sleipner has elucidated and established key parametric dependencies of these fundamental processes, the associated trapping mechanisms, and sequestration partitioning among them during consecutive ten-year prograde (active-injection) and retrograde (post-injection) regimes.
Intra-aquifer permeability structure controls the path of immiscible CO2 migration, thereby establishing the spatial framework of plume-aquifer interaction and the potential effectiveness of solubility and mineral trapping. Inter-bedded thin shales, which occur at Sleipner, retard vertical and promote lateral plume migration, thereby significantly expanding this framework and enhancing this potential. Actual efficacy of these trapping mechanisms is determined by compositional characteristics of the aquifer and caprock: the degree of solubility trapping decreases with increasing formation-water salinity, whereas that of mineral trapping is proportional to the bulk concentration of carbonate-forming elements, principally Fe, Mg, Ca, Na and Al. In the near-field environment of Sleipner-like settings, 80–85% by mass of injected CO2 remains and migrates as an immiscible fluid phase, 15–20% dissolves into formation waters, and less than 1% precipitates as carbonate minerals. This partitioning defines the relative effectiveness of hydrodynamic, solubility, and mineral trapping on a mass basis.
Seemingly inconsequential, mineral trapping has enormous strategic significance: it maintains injectivity, delineates the storage volume, and improves caprock integrity. Four distinct mechanisms have been identified: dawsonite [NaAlCO3(OH)2] cementation occurs throughout the intra-aquifer plume, while calcite-group carbonates [principally (Fe, Mg, Ca)CO3] precipitate via disparate processes along lateral and upper plume margins, and by yet another process within inter-bedded and caprock shales. The coupled mineral dissolution/precipitation reaction associated with each mechanism reduces local porosity and permeability. For Sleipner-like settings, the magnitude of such reduction for dawsonite cementation is near negligible; hence, this process effectively maintains initial CO2 injectivity. Of similarly small magnitude is the reduction associated with formation of carbonate rind along upper and lateral plume boundaries; these processes effectively delineate the CO2 storage volume, and for saline aquifers anomalously rich in Fe-Mg-Ca may partially self-seal the plume. Porosity and permeability reduction is most extreme within shales, because their clay-rich mineralogy defines bulk Fe-Mg concentrations much greater than those of saline aquifers. In the basal caprock shale of our models, these reductions amount to 4.5 and 13%, respectively, after the prograde regime.
During the retrograde phase, residual saturation of immiscible CO2 maintains the prograde extent of solubility trapping while continuously enhancing that of mineral trapping. At the close of our 20-year simulations, initial porosity and permeability of the basal caprock shale have been reduced by 8 and 22%, respectively. Extrapolating to hypothetical complete consumption of Fe-Mg-bearing shale minerals (here 10 vol.% Mg-chlorite) yields an ultimate reduction of about 52 and 90%, respectively, after 130 years. Hence, the most crucial strategic impact of mineral trapping in Sleipner-like settings: it continuously improves hydrodynamic seal integrity of the caprock and, therefore, containment of the immiscible plume and solubility-trapped CO2.