Laboratory batch experiments and geoehemical modelling of water-rock-super critical CO2 reactions in Gulf of Mexico Miocene rocks: Implications for future CCS projects

摘要

Storage of CO2 in deep saline formations in a super critical liquid state has been proposed as a way to mitigate the effects of increased atmospheric CO2 levels. The ultimate fate of the CO2 after injection requires an understanding of mineral dissolution/precipitation reactions occurring between the target formation minerals and the existing formation brines at formation temperatures and pressures in the presence of supercritical CO2. In this experiment core material taken from a Miocene age Gulf of Mexico core from a depth of 2806 m was reacted with synthetic brine at varied but high temperatures and pressures in the presence of super critical CO2. XRD and SEM analyses were conducted before and after reaction to identify dissolution of existing minerals and precipitation of authigenic mineral phases. Periodic geoehemical analysis of the reaction fluid was used to quantify changes in the elemental composition of the reaction fluid which helps identify potential mineral dissolution/precipitation reactions. Reaction brine (140 ml) was loaded into a high pressure reaction vessel with 8 g of core sample. Experimental temperature was set to 70, 100 or 130°C;; pressure was set to 200 or 300 bar, and solution chemistry was changed from de-ionized (DI) water to a 1.88 M NaCl solution. After the introduction of CO2 the Ca and alkalinity concentrations showed the largest increases, Ca concentrations increased ~1000 ppm, suggesting carbonate dissolution was the dominant geoehemical reaction. Final equilibrium Ca concentrations increased with decreasing reaction temperature because of greater CO2 solubility. In addition, the reactions with the NaCl brine produced higher equilibrium Ca concentrations than the DI water experiment, likely due to the decrease in ion activity with higher ionic strength solutions. Pressure change from 200 to 300 bar did not significantly alter reaction rates. Unlike Ca, silicate dissolution reactions appear to be positively correlated with reaction temperature. Silicate dissolution rates are 2 orders of magnitude slower than carbonate dissolution rates. In this study, PHREEQC was used to simulate brine-rock-CO2 interactions in batch experiments under high pressure and high temperature. Generally, the geochemical models reproduced concentration of Ca, Mg, K and Si seen in the water rock experiments suggesting that carbonate and K-feldspar dissolution are the dominant geochemical reactions. In addition, geochemical models show that dawsonite precipitates in higher salinity (higher Na+ concentration) experiments.