DOE OSTI2014
This study uses modeling and simulation approaches to investigate the impacts on injectivity of trace amounts of mercury (Hg) in a carbon dioxide (CO 2 ) stream injected for geologic carbon sequestration in a sandstone reservoir at ~2.5 km depth. At the range of Hg concentrations expected (7-190 ppbV, or ~ 0.06-1.6 mg/std.m 3 CO 2 ), the total volumetric plugging that could occur due to complete condensation of Hg, or due to complete precipitation of Hg as cinnabar, results in a very small porosity change. In addition, Hg concentration much higher than the concentrations considered here would be required for Hg condensation to even occur. Concentration of aqueous Hg by water evaporation into CO 2 is also unlikely because the higher volatility of Hg relative to H 2 O at reservoir conditions prevents the Hg concentration from increasing in groundwater as dry CO 2 sweeps through, volatilizing both H 2 O and Hg. Using a model-derived aqueous solution to represent the formation water, batch reactive geochemical modeling show that the reaction of the formation water with the CO 2 -Hg mixture causes the pH to drop to about 4.7 and then become buffered near 5.2 upon reaction with the sediments, with a negligible net volume change from mineral dissolution and precipitation. Cinnabar (HgS(s)) is found to be thermodynamically stable as soon as the Hg-bearing CO 2 reacts with the formation water which contains small amounts of dissolved sulfide. Liquid mercury (Hg(l)) is not found to be thermodynamically stable at any point during the simulation. Two-dimensional radial reactive transport simulations of CO 2 injection at a rate of 14.8 kg/s into a 400 m-thick formation at isothermal conditions of 106°C and average pressure near 215 bar, with varying amounts of Hg and H 2 S trace gases, show generally that porosity changes only by about ±0.05% (absolute, i.e., new porosity = initial porosity ±0.0005) with Hg predicted to readily precipitate from the CO 2 as cinnabar in a zone mostly matching the single-phase CO 2 plume. The precipitation of minerals other than cinnabar, however, dominates the evolution of porosity. Main reactions include the replacement of primarily Fe-chlorite by siderite, of calcite by dolomite, and of K-feldspar by muscovite. Chalcedony is also predicted to precipitate from the dissolution of feldspars and quartz. Although the range of predicted porosity change is quite small, the amount of dissolution and precipitation predicted for these individual minerals is not negligible. These reactive transport simulations assume that Hg gas behaves ideally. To examine effects of non-ideality on these simulations, approximate calculations of the fugacity coefficient of Hg in CO 2 were made. Results suggest that Hg condensation could be significantly overestimated when assuming ideal gas behavior, making our simulation results conservative with respect to impacts on injectivity. The effect of pressure on Henry’s constant for Hg is estimated to yield Hg solubilities about 10% lower than when this effect is not considered, a change that is considered too small to affect the conclusions of this report. Although all results in this study are based on relatively mature data and modeling approaches, in the absence of experimental data and more detailed site-specific information, it is not possible to fully validate the results and conclusions.
54 ENVIRONMENTAL SCIENCES↗