DOE OSTI2010
In this study, reactive transport simulations were used to assess the mobilization and transport of organics with supercritical CO{sub 2} (SCC), and the co-injection and transport of H{sub 2}S with SCC. These processes were evaluated at conditions of typical storage reservoirs, and for cases of hypothetical leakage from a reservoir to an overlying shallower fresh water aquifer. Modeling capabilities were developed to allow the simulation of multiphase flow and transport of H{sub 2}O, CO{sub 2}, H{sub 2}S, as well as specific organic compounds (benzene), coupled with multicomponent geochemical reaction and transport. This included the development of a new simulator, TMVOC-REACT, starting from existing modules of the TOUGH2 family of codes. This work also included an extensive literature review, calculation, and testing of phase-partitioning properties for mixtures of the phases considered. The reactive transport simulations presented in this report are primarily intended to illustrate the capabilities of the new simulator. They are also intended to help evaluate and understand various processes at play, in a more qualitative than quantitative manner, and only for hypothetical scenarios. Therefore, model results are not intended as realistic assessments of groundwater quality changes for specific locations, and they certainly do not provide an exhaustive evaluation of all possible site conditions, especially given the large variability and uncertainty in hydrogeologic and geochemical parameter input into simulations. The first step in evaluating the potential mobilization and transport of organics was the identification of compounds likely to be present in deep storage formations, and likely to negatively impact freshwater aquifers if mobilized by SCC. On the basis of a literature review related to the occurrence of these organic compounds, their solubility in water and SCC, and their toxicity (as reflected by their maximum contaminant levels MCL), benzene was selected as a key compound for inclusion into numerical simulations. Note that considering additional organic compounds and/or mixtures of such compounds in the simulations was beyond the scope of this study, because of the effort required to research, calculate, and validate the phase-partitioning data necessary for simulations. The injection of CO{sub 2} into a deep saline aquifer was simulated, followed by modeling the leaching of benzene by SCC and transport of benzene to an overlying aquifer along a hypothetical leakage pathway. One- and two-dimensional models were set up for this purpose. The target storage formation was assumed to initially contain about 10{sup -4} ppm benzene. Model results indicate that: (1) SCC efficiently extracts benzene from the storage formation. (2) Assuming equilibrium, the content of benzene in SCC is roportional to the concentration of benzene in the aqueous and solid phases. (3) Benzene may co-migrate with CO{sub 2} into overlying aquifers if a leakage pathway is present. Because the aqueous solubility of benzene in contact with CO{sub 2} is lower than the aqueous solubility of CO{sub 2}, benzene is actually enriched in the CO{sub 2} phase as the plume advances. (4) For the case studied here, the resulting aqueous benzene concentration in the overlying aquifer is on the same order of magnitude as the initial concentration in the storage formation. This generic modeling study illustrates, in a semi-quantitative manner, the possible mobilization of benzene by SCC. The extent to which the mobilization of this organic compound evolves temporally and spatially depends on a large number of controlling parameters and is largely site specific. Therefore, for more 'truly' predictive work, further sensitivity studies should be conducted, and further modeling should be integrated with site-specific laboratory and/or field experimental data. The co-injection of H{sub 2}S with CO{sub 2} into a deep saline aquifer was also simulated. In addition, the model considered leakage of the supercritical CO{sub 2}+H{sub 2}S mixture along a preferential pathway to an overlying fresh-water aquifer, followed by reaction of the CO{sub 2}+H{sub 2}S mixture with that aquifer. A simple 2-D model that included a storage formation and a sealing aquitard was developed to simulate the movement of H{sub 2}S in a typical CO{sub 2} storage formation. Model results indicate that H{sub 2}S is stripped off at the edge of the advancing supercritical plume, because of the H{sub 2}S preferential solubility in water compared to CO{sub 2}. The magnitude of H{sub 2}S preferential dissolution, however, decreases with decreasing temperature and pressure. To capture this behavior and evaluate the breakthrough of H{sub 2}S through a leakage pathway (from the deep storage formation to a shallower aquifer), another model was constructed, considering a storage formation, an overlying aquifer, and a vertical leakage pathway between them.
54 ENVIRONMENTAL SCIENCES↗